1;ig. 4 for different fibers in the dry state and in Fig. 5 ftn the same fibers in the swollen state. When the curves are straight lines, this would highly wpport (although not fully prove) the conclusion of independent scattering by the particless; but this is seldom the case, as Warren and Biscoe4 for example pointed out 111 their experiments on carbon black. The curves of Fig. 4 show that in the case of dry fiber5 no straight lines are ever obtained. The curves obtained with swollen fibers (Fig. 5), on the other hand, differ from those of dry fibers by being perfectly straight as woiild hr the case when the above assumptions are fulfilled. The conclusion, therefore, is obvious that due t o the swelling process, the particles are so far moved apart that inter-particle interference no longer plays a part, and the calculation of the particle size from the slope of the curves appears permissible. -4riother important conclusion froiti t h e straight shape of the curve is that the size of the particles will not vary t o i I'tigi. txtent drouiid an 'tverage \.a1ue. I n Table I the slopes of the curves obtained with different fibers following this new swelling technique art' giveii, these values are proportional to the sizes of the particlei iii t h e different fiher-, indicating their ratio4 1 ' 4B1,F 1 li.idiiis of gyrntion,
DiTnirtir
bilirr
Jute I Iydrolyzetl jute I%Y Rayon (viscosei Rayon (fiber G) Rayon (fortism) Ramie
centrated alkali gives 9-hydroxy-9-carboxyfluorene3 and with more dilute alkali fluoren-9-one. The intermediate formation of 9-hydroxy-9c.arboxy-4,5-diazofluorene followed by removal of formate ions thus provides an explanation. 1-10, ,CO,-
Iiiglett and Smith further state that this is "the first reaction i n which carbon monoxide has been observed to be lost from a 1,2-diketone grouping." However, several 1,%diketones have long been known to lose carbon monoxide by the action of heat : diethyl 3,4-diketoadipate (ketipink ester), 2,4,&7-tetraketooctane, 1,B-diphenylI ,3,4,6-tetraketohexane and 1,4-diphenyl-1,4-di-
cyano-2,3-diketobutane.j It appears from this, and from work carried out in this Laboratory, that for loss of carbon monoxide to occur d t temperatures below about L(i0" (for %keto esters) or 230" (for 1,2--diketonesj, the following grouping should be present, both IT] ?--ketoesters anti it1 1 ,L'-tliketone\ R
s c'o
C'O
CII
Y
W l l ere
Values for radii of gyration are gven 111 two column5 111 Table I, one using formula (d) arid the othei formula ilii The values of the particle diameter given also i n Table I were calculated from the two sets of radii of gyration hut under the assumption that the particles in both c ha\e the shape of an elongated cylinder I _ _ -
(3) If the various assumptions are fulfilled on which Guinier 5 theory IS based, includtng tndependent icattering the theorv re quires these curves t o be straight lines (4) T Riscoe and I3 I Wirreii J A p p l i r i P h I1 942)
-_
.
X = CHCOR' or OR' K = 11, alkyl or aryl T = electrt)ri-,ittractiiig group
____
13) A Barver Rei , 10, 125 '1877) 1' Friedlander, r b d 11877) (4) K Anachutz and F K Japp, rbid , 11, 211 (1878) 15) \Y Wi4icenu5, rbid 18, 811 (1893)
10, 331
SCHOOL OF CHEMISTRY rIYIVERSITY OF SYDYEY
, X S K' , A U ~ T R A L I ARECEIVRI) J U V E q , 1950
4,4'-Difluorobenzhydrol 1'
4
(;1
YIfIFK A Y D
K c
h I N N
'Tlie authors recently reported1 the synthesis of The Loss of Carbon Monoxide from 1,Z-Di-4,4'-difluorobenzhydrol, a compound previously ketones unrecorded in the literature, along with its simple BY P. 1-1. GORE A S D G h. I~IJT.HBS properties and some of its reactions. I n FebruIn a recent communication, C . E. Inglett and ary, 1950, Picard and Kearns reported2the syntheG. IT. Smith' have recorded the formation in sis of the same compound by a different series of dmost quantitative yield of ~,3-diazafluoren-n- reactions. Whereas our benzhydrol melted a t one by alkaline treatment of 1,lO-phenatithroline- 47.347.5', theirs melted a t 92-93". Rigid comparisons of the two compounds indicate different 5,ti-quinone. 'They regarded this removal of the elements of structures. In our hands reduction of 4,4'-difluorobenzocarbon niorioxide as a "very surprising development," yet the reaction is perfectly analogous to phenone with zinc and alkali gave the desired benzthe benzilic acid-type rearrangement' of phen- hydro1 (11), whereas attempted reductions by anthrene-5,6-quinone, which with fairly con- aluminum isopropoxide yielded an unidentified substance (V) in the hands of Picard and Kearns 11) G. E. Inglett and G P Smith, THISJ O U R N A L , 72, 842 (1950) The reaction schemes are as follows: (2) J Liebig, Ann., 25, 27 (1838), N Ztniu, ibid , 31, 329 (1839), M Bdqler, Ber (1908); etc
,
14, 323 (18x1)
1
(:attrrniann,
Ann
347, 304
(1) I?. A Gunther and R C BIinn THIS JOURNAL. 72,4282 (1950). ( 2 ) Picard and Kcarni, Can J Rcsca7ch, WB, 36 (1960).
Noms
0
I1 (11) m. p. 47.3-47.5'
(1) m. p. 107-108" oxime rn.p. 134-135"
t
Picard and Kearns AI(0Pr)z _ (91%1 __f
sideration by the fact that both conlpounds I1 and V retain essentially the same nielting points after repeated meltings, after reCrOs 0 crystallizations from a variety of solvents, 0 -+----H and after prolonged drying a t reduced pressures over desiccants. (IV) (84%) (V) It has not been possible to apply the exm. p. 102-103" rn.p. 92-93' isting data for compound V to any reasonoxime m. p. 135-136' able structures of possible products from Compounds I and IV are identical, as shown by the treatment of 4,4'-difluorobenzophenone with mixed melting points, and by comparisons of their aluminum isopropoxide oximes and 2,4-dinitrophenylhydrazones. It would ordinarily be anticipated that the The analytical comparisons carried out in these aluminum isopropoxide reduction of the ketone laboratories are summarized in Table I. would proceed normally, with greater chances for stray reactions with a zinc and alkali reduction. TABLEI Hence, the results obtained are puzzling. The AXALYTICAI, COMPARISONS OF COMPOUNDS I N SYNTHETICclaim of Picard and Kearns is, nevertheless, in SCHEMES error, and the substance supposed by them to be Analyses, % the substituted benzhydrol was not completely Mol. wt. Carbon Hydrogen Compound Sourcea Calcd. Found Calcd. Found Calcd. Found characterized and in fact does not even show the I G&B 218 220 7 1 . 9 6 ... 3 . 9 3 required analysis for this substance. The data I 2,4-dinitrophenylsummarized here do support the assigned struchydrazone G&B ... ... .. .. 397 407 ture of our compound I1 as 4,4'-difluorobenzhyI1 G&B 220 218 7 0 . 9 0 7 0 . 5 8 4 . 5 2 4 . 5 8 drol. I1 $-nitrobenzoate G&B 369 368 65.03 65.16 3 . 5 5 4 . 0 0 Experimental.-The reactions diagrammed 204 206 7 6 . 4 6 7 6 . 9 0 4 . 9 4 5 . 1 9 I11 G&B and sources of parent ketones are presented and I11 3,3'-dinitroderivative G &B . , . ... 53.07 5 3 . 2 6 2 . 7 4 2 . 9 4 discussed in the literature cited, with two exceptions. The conversions of I1 and I11 t o I and IV P&K ,., ... 71.96 ., . 3.93 .. G&B of V to IV were accomplished in the yields shown V P & Kb 220 361 7 0 . 9 0 73.57 4 . 5 2 4 . 7 2 with the aid of chromic anhydride in acetic acid, " A s prepared by Gunther and Blinn (G & B) or by in the usual manner. All other comparison Picard and Kearns ( P & K ) . As prepared by P & K , data were obtained with accepted procedures. but recrystallized and analyzed by G & B from sample Melting points are uncorrected. obtained from Dr. Picard.
Fc)-g-c>F
Compounds I1 (b. p. 150-154" (5 mm.)) and V (b. p. 143" (3 mm.)) exhibit depressed melting points in admixture. They are probably not dimorphic since supercooling the melts of each and cross seeding did not effect conversion of one to the other. Their ultraviolet absorption characteristics are shown in Table 11. TABLEI1 ABSORPTION CHARACTERISTICS EtOH Compound Amin.
e
239 202 v 241 367 I1 V" 240 367 a I1 = 45.3%, v =
I1
+
EtOH hmax.
e
EtOH hmin.
c
EtOH Xmax.
265 1431 269 868 271 265 1352 269 867 271 265 1463 269 900 271 54.7%.
(3) See Wilds, "Organic Reactions," Vol 11, John Wiley & Sons, Inc., New York, N . Y., pp. 178-223.
DIVISIONOF ENTOMOLOGY UNIVERSITY OF CALIFORNIA CITRUSEXPERIMENT STATION RIVERSIDE, CALIFORNIA RECEIVED JUNE 5, 1950
The Preparation and Properties of 5-Ethyl-5(1-methyl-1-nitroethyl)-barbituric Acid BY E. B. HODGE
e
1211 1234 1266
Solvation phenomena are eliminated from con-
The preparation of ethyl ethyl-(1-methyl-1-nitroethy1)-malonate has recently been reported. Since i t seems that there is no record of the hyp(1) E. E. Van Tamelen and G . Van Zyl, THISJOURNAL, 71, 835 (1949). Since this note was written, the preparation of 5-methyl-5(1-methyl-l-nitroeteyl)-barbituric acid has been reported; E. E . Van Tamelen and G. Van Zyl, ibrd., 72, 2979 (1950).
