4. Small amounts o f biacetyl greatly incrcase the fluoresceiice of acetone *>. I t is concluded that the fluorescent emitters are the same for both acetone arid biacetyl, a t least as regards the bands at .WOO -4. and longer wave lengths. ii Mechanisms to account for these facts are
advaneed. An adequate explanation is based on c ery probable deactivation of excited acetone rtiolecules by biacetyl, the excited biacetyl being respoiisible for the fluorescence. However, fluorewenre of thi, CEIICO radical cannot be e.:cluded, cl
P ~ ( , ~ ~ ~ ,
[CC)NTRIBIr170N FRO\I T H E M E T C A I T CFIEMICAI, LABORATORY OF
BROWNUNIVERSITI']
The Absorption Spectrum of Biacetyl between 1500 and 2000 i.
In connection with work on the spectroscopy and photochemistry of ketones it was thought advisable to study the absorption spectrum of biacetyl. This substance is formed during the photochemical decomposition of acetone' arid is a possible product, along with other diketones, during the irradiation of ethyl methyl ketone. Since absorption coefficients of the ketones below 2000 A. are far higher than in the lied: ultraviolet, it was hoped that the absorptioii spectrum might be used to detect even small amounts of biacetyl. As would be expected the juxtaposition of two chromophoric groups causes a displacement of the absorption spectrum toward longer wave lengths. Thus 111 the near ultraviolet biacetyl vapor absorbs to 4670 A&.,2 whereas acetoneJ arid ethyl methyl ketonei show absorption to only betweeii 3200 and 3300 Below these absorption regions both biacetyl arid the other ketones show a broad zone of transmission extending to below 2000 A.,where an inteiise absorption takes place. Biacetyl shows bands with fine structure from 1670 to -1.100 A. at ivhich point the bands become diffuse. Continuous absorption is observed below %SO0A.z and exteridiiig to about 2500 A. I,ardy5 arid L i ~ t h yhave ~ iiivestigated the spectra of biacetyl and of glyoxal, respectively, i i i hexane solution. In biacetyl four bands occur between 4513 and 4033 A,,a wide band at 2WO ( 1 ) Barak and Style L V u / , ~e 136, 7 O i i l O J i ) Si-ence and XViId, ibid 138, 206 (1936) ( 2 ) Cf bponer, Molrkulqxktren L'erlag voit T u l i i i , springer, Berhr1,Vol I 1'J3>,p 117 ( 3 ) Koyes, Duncan and hlanning J C h e m P h y s 2, 717 (l'J34) 1 his paper contains reference, to enriier work See d13o N o i e i I o v i i Faraday S o c , 83, 1 4 9 5 t1037) 4 ) Duncan 1311s arid Woycs THI>J o r ' H \ r A L G8, 14.54 1 lO3hl
a i d another a t 1950 A. being observed also. Glyoxal between -l(jl,3 and 31 19 shows seventeen barids and in addition there are bands a t LhOO and 1950 A The substitution of a CH? group for a hydrogen atom widens the narrow bauds, reduces their number and increases the absorptioii coefficierit
A.
Experimental The biacetyl (Eastman) was fractionated several times lxfuie being introduced into the storage bulb, after which I I wa5 again fractionated three time5 a t low pressures in the vacuuin line Calciuin chloride was used a5 a drying agent and the boiling point of the fraction taken was 87XY". The freezmg point was approximately -4", although n o value is recorded in the literature No vapor pressure data could be found, but the observed vapor pressures were about 0 01 min a t -?(I", 1 mm a t -30"and 13 mm at 0" A quartz cell with relatively thin windows was interposed between the hydiogen discharge tube and the slit of thc vacuum Spectrograph to avoid introducing the biacetyl directly into thr spectrograph Thus an absorbing column about 15 cm in length and vapor pressure, from 0 01 to 1 5 mm were used. The spectrograph ha5 been dc scribed llford I plates were used ?he di\per\ion of the spectrograph with the setting used wLii found- t o be 8 48 -4 per mm. ni the first order T h L X LrUl 0 2 i A line of C I11 and a number of the fout Lh posi live bands of carbon monoxide were used7 as wavc length .t.tndardq Sinc-e a yuartr cell w a s ~ e m p l o ~ e dthe , lower l l n ~ UT ~ i oI)zcr\at~onwa5 about 1560 A ,
Results 'fwenty-six bands of biacetyl were found in all These are listed in Table I, together with rough estiniates of the intensities based on the pressurc a t which the bands appeared and their visually estimated intensities 011 the plates. The red edges o f the bands appeared to be most sharp and are reported. Most o f the bands were sufficivntlj;I
I
