GEORGE McCoy
1956 [CONTRIBUTION FROM
THE
AND
ALLANR. DAY
Vol. 65
DEPARTMENT OF CHEMISTRY AND CHEMICAL ENGINEERING OF
THE
UNIVERSITY OF
PENN-
SYLVANIA]
The Reaction of ortho-Quinones and ortho-Quinonimines with Primary Amines BY GEORGE McCoy1 AND ALLANR. DAY While the reactions of retenequinone and phenanthraquinone with ammonia to form the Corresponding quinonimines have long been known,* no work has been reported on the behavior of these quinones with primary amines under siniilar conditions. We have found that certain primary amines react with retenequinone to form %substituted retenoxazoles. Only those amines which have two a-hydrogen atoms will react to form retenoxazoles. This is showii by the fact that isopropylamine and a-methylbenzylamine do not yield oxazoles, whereas 7~-butylamine, ethanolamine and benzylamine yield 3substituted retenoxa~oles.~ The following mechanism is proposed for this reaction
I \.
\
I I1
retenequinone and benzylamine in alcohol Benzaldehyde was isolated as its semicarbazone from the hydrolysis products. The hydrolysis of 111 should also yield 9-amino-10-retenol. Therefore, if aminoretenol were condensed with an aldehyde to yield an oxazole, the above evidence would be confirmed. Since aminoretenol is very unstable, 9,lO-aminophenanthrol was prepared instead and condensed with n-butyraldehyde and benzaldehyde, giving 2-propyl- and 2-phenylphenanthrox azoles, respectively. Evidence for the course of reaction involved in ring closure is strongly in favor of the one postulated. Knoevenagel,4 Mayer5 and Stein and Day6 have established the fact that active hydrogen compounds readily add to Schiff bases under suitable conditions. That the dihydrooxazole IIV) was dehydrogenated by the quinone could iiot I)e established directly due to the instability of retenehydroquinone. Satisfactory evidence for this step was obtained by adding p-benzoquinone to reaction mixtures of retenequinonimine and priniar) ,mines. Both hydroquinone and quinhytlrorie were isolated from these reactions.' If the reaction follows the proposed final step. a maximum yield of 30cI, would be expected when reteiiequitione and amine are used in equivalent amounts. The higher yields actually obtained may be explained by the ease with which retenehydroquinone undergoes air oxidation to retenequinone. The dehydrogenation tnay be represented as
\-I
(1) Present addresi, L'niveriity of Pennsylvania, Philadelphia, P a ( 2 ) Bamberger a n d H u o k e r . . 4 x u , 229, 1111' i 1 8 8 i ) , Pschorr. 3
K
0
The forin,itiori of the Schiff base type of compound 11 is mtlic,ited bv the fact that when the reaction was carried out in dry toluene water was liberated, and when the quinone was replaceti by the quinonimine ammonia was formed with subsequent formation of a 2-subsituted retenoxnzole. Evidence for the formation of 111 was obtained by adding hydrochloric acid a t the tirne to a reaction mixture of
~
ption for i t r r : ~ c triith ~ retene-7 (')/,,#!
,L''
63, l 2 h S (18931
\C'
\C'
H
c
1
11
/
+
0 -C'
'3
I
c
(I
I \.
I
'c c
A,
.c( ~
zR
L'.FI
s
~
q
0-c
f+;
-OH
-
i ( 4 ) Knoevenagel, B e ? . , 31, 2598 (1898). ii) Mayer, ~ ~ isoc i .C h i m . , 3 3 , 1 5 7 . 3 ' 3 I(:! Stein and D a y , T A I SJ O V R N A L . 64, ' . i )T h e rate of reaction between the p r i m d r y amine and P-bro7.o quinone IS greater than the rate of reaction between the a m i n e and retenequinone, making i t impractic:il to separate any 1i)droqiiinone from the mixtiire v i products. IIouever, when reteneqiiinone wa. replaced bb- i t , i m s t e r r r t r n c q i i i n o n i m i n e , t h e r . t t P a wrre l a ~ ~ o r a b l x ir\eiard
Oct., 1943
REACTIOSOF 0 \C/
\C-R I
1,
/C--iY
v
\C-OH
+ l i
/C-OH
VI
When retenequinone is replaced by retenequinonimine. the reactions proceed more rapidly and give higher yields of oxazoles. The formation of the latter rather than imidazoles indicates that the initial condensation takes place a t the imino group. The fact that higher yields are obtained may be attributed to the liberation of ammonia in place of water, thus eliminating any secondary hydrolysis reactions. The reaction of phenanthraquinone with benzylamine gave small amounts (9yo)of ?-phenylphenanthroxazole and 1-benzyl-2-phenyl-phenanthrimidazole ( 2 . i 7 c ) . Benzaldehyde was recovered in the steam distillate from this run as its phenylhydrazone and semicarbazone.8 The use of phenanthraquinonimine in place of the quinone gave better yields of 2-phenylphenanthroxazok and the reaction with butylamine gave a X C i , yield of 2-propylphenanthrosazole as contrasted with the complete failure of the reaction when the quinone was used
Experimental All of the melting points represent corrected value5 and check the literature values unless otherwise stated. Retenequinone.-This compound was prepared from retene by the method of Kreps and Day,Qm. p. 197-199". Phenanthraquinone was prepared from phenanthrene according to the directions of GracbelO and the purification mas carried out by the method of Courtot.11 It was then recrystallized from 5 0 5 acetic acid; yields 55-60c4,; m. p. 208-209.5 '. Retenequin0nimine.---Theqiiiitoiiirtiine was prepared from rctcttequinone arid ammonia according to 13amberger and Hooker2 and recrystallized from warm (50') dry alcohol saturated with ammonia; yields 50-C,0yo; m. p . 107-108".
Phenanthraquinonimine.-The i i i c t l i ~ i ~ l of I'schorr* u d for t l t i y preparatiurt '1'111. crii(l(, p r i ~ l t t c t NXCI recrystallized from \r-arm (50") d r y ak(JhlJl \atiiratetl I\ it11 ammonia; yields 70-757; m . 1). 1T,ii-137°. 9,10-Aminophenanthrol Hydrochloride: O i t c . gruiii \.as
jl),O048 m o l r ~of phiiiatitltrurjitittiiiiimiiic \ \ ; I \ arlrlctl t i l III(I cc. of dry alcohol, ivhich Iin ivlicre t h e crude product was contaminated \\-it11 gummy material, the rcaction mixture was steam di3tilled and the residue so obtained then recrystallized from dry alcohol. At rooin tcmperature the reaction proceeds very sloivly. The best yields (70-807;) were obtained by heating the reaction mixtures at 75-100' for three hours. Heating for longer periods of time or a t higher tcrnperaturc\ (lid riot increase thc yields. It:illixe the rcsitltic from alcohol unt il cr)lorlL,ys i:ry,tnly of "-prol)ylrctriieoxazole were ohtaiiirtl. l i t . 1). l i 1 0 1 1 1 1 O . l I i \ u l iiicltitig point determination-. \vi111 a ~ i r i 1 ) I t lmqxtrml . l)y Stc,in a i d Day6 shoired no d r ~ i r c . ~ i ~ ~11i \\:I. . i I i 1 ~ i 1 ~t1t:ii 1 ai 78' i i t u u i i n u i i i yields (62'; of t' j ~ r ~ i ~ i ~ l ~ - c ~ \~\ i~w~ i~i li ~~~~~ ~i t :i iici i~~ tr l i eight t u clc\.cii I r o i t i - ~\ \ l i i l c . i i i ; i \ i i i i t t t i t y i c l i l ~elf t' ~)lttwylrctciio\anIIc \ e < rt' i ~ l , l L i i i i t ~ ii ili I ltrL,c, I t i ~ t i r ~I.t ~ 1 t i ) 1 1 1 ( 1 IK Iioittte(I o i l 1 1 1 i : i t 1111. [ i i t ' . e I I < Y ' o f t l i r ~ilicnylgriiiip u i IJCIIzy1:iiniitc. \ \ ~ i i i I < l I K t ~ \ j i ~ c ~ f c I~Oi l f;ic,ililalc t l t c hydrogcii 5ltift iio1t.rI t ~ a r l r ai111 i~ I Itii-. iiicit,:tw 111c rate cii 0\iizoIc foriiial ioti. h ~ ~ c r i t r t \ct i~rei ~t i~i t ; ~ I l y c ~ c ~ r ~ i c ; i i r i'.' A n d . Calcd. iur C1 H , " X 2 S , '4.61. Study of the Course of Reaction. Step 1. T h e Reaction of Retenequinonimine with BenzqlRmine. ' iiii C > k !JC'llLyi(0.0038mole) of retenequixionitti:ii~ a i i ( I amine were dissolved in 50 c c . (11 c!rv : i ~ I ~ : x i i 'ri r i i i rcfiiiked for four hours. The evolved a : t ; ! ~ i ~ ~iirl t i t c~~~ i l l c t~ c i in i boric acid solution and titratti1 5; i t 11 11;;Ct , J < Illuric acid i l l the presence of m e t h y l r e d , uiiiiiii~~iia eviilvcll !J ( l l l : 3 niuic. The reaction mixture way \ l c a n i tli-t illcd :ind t h c re4tiuc recrystallized from alcohol, y;clrl of L'-lihviiylrct cnokamlv 1.11 g. (&$>!; m . p. 17-1 ;:ti', Step 2. Reaction of Retenequinone with Benzylamine. I irc ~v~icii~iiiic~ii~ nil\ i l i -Five grani. ( ( ( 1 1t1!t III( I 1,
niolei of berizylarnine addctl. T h e w l u t i o n v. :I> rabllii\ctl : r t i < i ~ l 1 U l i O ltit I l l l < ~ < ii i x h :Ill < \ < ' I : - Of I
R.m
y
Vol. 65
THEFLUORESCENCE OF VITAMIN A
Oct., 1943
soluble fraction (0.65 g., 9%) was recrystallized from alcohol and proved to be 2-phenylphenanthroxazole, m. p. 206207". The more soluble fraction (0.25 g., 2.7%) was 1benzyl-2-phenylphenanthrimidazole; m. p. 24 1-24 1.5'. This compound has not been previously reported. Anal. Calcd. for C Z ~ H ~ ~ N C,Z 87.46; : H, 5.24; N, 7.27. Found: C, 87.69; H, 5.30; N,7.12. (b) In Glacial Acetic Acid.-The same quantities of quinone and amine were dissolved in 25 cc. of glacial acetic acid, refluxed for three hours and filtered while hot. The brown residue (3.1 g.) was digested with hot dioxane and subsequently with hot nitrobenzene. This left a green residue of phenanthroxazine; yield 2.2 g. (48%); m. p. > 360'. Anal. Calcd. for CBHI~NO: N, 3.66. Found: N, 3.46. The original filtrate on cooling deposited 2-phenylphenanthroxazole which was recrystallized from alcohol; yield 1 g. (14%); m. p. 206-207". Water was added to the filtrate from the oxazole, precipitating a gum. The latter was dissolved in a hot mixture of acetone and ether, sepafrom which 1-benzyl-2-phenylphenanthrimidazole rated on cooling. It was recrystallized from alcohol; yield 0.4 g. (4%); m. p. 241-241.5'. Reaction of Phenanthraquinone with n-Butylamine in Toluene.-Although this reaction was carried out under various conditions, only intractable brown solids and gums were obtained. Reaction of Phenanthraquinonimine with Benzylamine. -One gram (0.0048 mole) of phenanthraquinonimine and 0.53 cc. (0.0048 mole) of benzylamine were dissolved in 50 cc. of hot toluene and refluxed for three hours. The
[CONTRIBUTION FROM
THE
1959
solution was steam distilled and the residue recrystallized from alcohol; yield of 2-phenylphenanthroxazole 0.83 g. (59%); m. p. 205-206'. No imidazole could be isolated. Reaction of Phenanthraquinonimine with n-Butylamine. -Two grams (0.0097 mole) of phenanthraquinonimine and 0.96 cc. (0.0097 mole) of n-butylamine were dissolved in 35 cc: of hot toluene and refluxed for eight hours. After steam distillation, the residue was dissolved in a mixture of dry alcohol and benzene. A small amount (0.1 8.) of a red, amorphous solid remained. The solution was evaporated and the residue recrystallized from alcohol; yield of 2-propylphenanthroxazole 0.9 g. (35%); m. p. 85-86'. A mixed melting point determination with a sample prepared by Stein and D a p showed no depression.
Summary 1. It has been shown that primary amines which have two hydrogen atoms on the alphacarbon atom react with retenequinone and phenanthraquinone (or their isosters, the quinonimines) to form the corresponding oxazoles. 2 . The course of the reaction has been shown to consist of the following probable steps: (1) an aldol-type of condensation between the quinone and the amine; ( 2 ) a shift of hydrogen in the two adjacent triad systems from carbon to oxygen; (3) addition of OH across a -N=CHlinkage; and (4) oxidation of a dihydrooxazole (by a quinone) to an oxazole. PHILADELPHIA, PA.
RECEIVED APRIL28, 1943
DEPARTMENT OF CHEMISTRY, LABORATORIES OF THE MOUNT SINAIHOSPITAL,AND FROM SCIENTIFIC BUREAUOF BAUSCH AND LOMEOPTICAL Co.]
THE
The Fluorescence of Vitamin A BY HARRYSOBOTKA,' SUSANKANNAND ERICH LOEWENSTEIN The fading green fluorescence of vitamin A under ultraviolet irradiation has been used for its histochemical demonstration in slices of liver and other animal organs2 An attempt by one of us (E. L.) to utilize this fluorescence for the quantitative analysis of vitamin A solutions by a photoelectric fluorometer led to the observation that vitamin A preparations in alcoholic solution display upon irradiation an initial steep increase in fluorescence, followed by complete destruction of fluorescence during prolonged irradiation. In (1) Supported by a Grant from Nutrition Foundation Inc. A report of the work was given a t the Conference on vitamin research held at Gibson Island, Maryland, under the auspices of the A.A.A.S., July ZCth, 1943. (2) H. Popper, Proc. SOC. Exptl. B i d . M e d . , 43, 133 and 234 (1940); Arch. Path., 31, 766 (1941); H. Popper and R . Greenberg, i b i d . , 31, 11 (1041). R . Greenberg and H. Popper, A m . J . Phyriol., 134, 114 (1941).
contrast to this, the fluorescence of vitamin A solutions in ether, chloroform, or benzene shows sometimes a small initial drop, but always assumes quickly a level of steady intensity which decreases but slowly. The fluorescence of vitamin A in non-polar as well as in polar solvents is fairly proportional to its concentration over a range from 0.1-5.0 I.U./ ml. under our experimental conditions. Although not quite as sensitive as the Carr-Price reaction, fluorescence may serve as a satisfactory basis for an analytical method. In concentrations below 0.1 I.U,/ml. erratic results are often obtained. In polar solvents such as methanol, ethanol, or isobutanol fluorescence follows a curve such as "1" in Fig. 1. Symbatic curves are obtained for varying concentrations, the ordinates (intensity
