R.P.BORKOWSKI AND P. LkUSLOOS
4044
y x for biacetyl excited near the 0-0 transition is roughly unity.16 At shorter wave lengths such as used here it decreases to the extent that photodecomposition from the II’& state (which must be initially formed) and from the first singlet excited state becomes important. -Assuming no internal conversion to the ground state and noting that a t short wave lengths biacetyl does not fluoresce, then yx 1 - 4, where 4 is the quantum yield of the primary process. As an estimate for 4 for the conditions of our experiment the value of 0.34 reported by Bell and Blacetl’ at X 2630 and 100” was used. With yx computed in this fashion, Fig. 5 shows a test of eq. 111. The results are ED/€* = i 7 0 SO and KlrU = 10,000 1000. We did not measure Euje-4 for benzophenone and for biacetyl in the vapor phase a t A2750 -4. and t = log”, but this result appears reasonable. In solutions, E D / E A‘v 600 a t that wave length.
*
(16) G. B. Porter, J . Chem. P h y s . , 32, 1587 (1960). (17) W. E. Bell and F. E. Blacet, J . A m . Chem. Soc., 76, 5332
(1954).
Vol. s4
If the transfer of energy takes place on every collision Ks
= m(L~/1000)!’2 [i?r~?T(~bfo ~a f
L 1 f ~ ) / A 1 f ~ l‘12 f.~]
where is the sensitizing cross section for biacetyl sensitization and MDand MAare the molecular weights of the donor and acceptor, respectively. If YDI is taken to be the sum of the collision radii for such molecules (- 10 A.) then the mean life of the benzophenone appears to be TD s 1.4 X 10+ sec. or longer. A great deal more work remains to be done on this system, but in principle the qualitative behavior is understood. It is apparent that direct lifetime measurements are becoming necessary and should be undertaken in a systematic way A. complementary study of the quenching of benzophenone phosphorescence by biacetyl was not undertaken because of the high temperatures necessary to obtain sufficient benzophenone in the vapor phase The author wishes to thank Dr. Mary Cox for preparing the quinine bisulfate calibration curve used in this work.
e.] The Vapor Phase Fluorescence and its Relationship to the Photolysis of Propionaldehyde and the Butyraldehydes1 [CONTRIBUTIOS FROM
THE hTATIOX.4L
BUREAUOF
STASDARDS, \ v A S H I N G T O N
2 5 , D.
BY R. P. BORKOWSKI~ AND P. XUSLOOS RECEIVEDMAY31, 1962 The effects of pressure, teriiperature and wave length on the fluorescence of n-butyraldehyde and isobutyraldehydc have been investigated. Xt 3310 k. the fluorescence yield decreases slightly with increase in pressure, while at shorter wave lengths this trend is reversed. Similar observations were made in the case of C2H5CH0and C2F5CH0. 9 n increase in temperature reduces the relative fluorescence yields a t all wave lengths. -4lthough no phosphorescence was observed, addition of biacetyl leads to a strong emission which could be ascribed to triplet excited biacetyl formed in the Energy transfer reaction: B +A B3 (I). The importance of this process is strongly dependent on wave length and pressure. The analysis ,Of the products formed in the photolysis of n-butyraldehyde and n-butyraldehyde-biacetyl mixtures indicated hu + C2H4 CHaCHO (II),is strongly inhibited t h a t at 3310 A. the intra~nolecularrearrangement process, n-C3H;CH0 by biacetyl. The efficieJicy of this quenching process parallels t h a t of process I. On the basis of these observations it is concluded that a t 3310 A. process I1 occurs from a triplet excited state.
+
+
+
+
used without further purification. .-lnhvdrous pentafluoroIntroduction propionaldeh>-de was obtained from Merck and Co. and Although the photochemistry of the simpler ali- used without a n y further purification. .Ill of these comphatic aldehydes is now fairly well e ~ t a b l i s h e d , ~pounds were thoroughly degassed and stored a t -80”. very little information is available in the literature The oxygen was assayed reagent grade and was obtained the Air Reduction Co. of the emission characteristics of CcHnCHO, from rl T-shaped high-quality quartz fluorescence cell with C2F5CH0,n-C3H+2H0 and i-C3H7CH0. The pres- plane windows, 56 mm. long and 28 mm. in diameter, atent work was undertaken in order to obtain in- tached to a vacuum line, was used for the fluorescence experiformation about the fluorescence of these com- ments. T h e light source was a Hanovia SH-type mediummercury lamp, used in conjunction with a Bausch pounds as a function of temperature, pressure and pressure and Loinb grating monochromator of 250 nirn. focal length wave length, and to correlate these data with the with entrance and esit slits 1.O m m . wide. One light from photochemical observations. the monochromator passed through the absorption length of tlie cell to a 1P28 plintomultiplier tube (calibrated against Experimenta a thermopile a t the Sational Bureau of Standards) for The propionaldehyde, It-butyraldehyde and isobutyraldcIiyde were obtained from Eastinan Kodak Co. and were purified 011 a spinning band distillation colunin. Only those fractions were used which were shown to be pure on a gasliquid chromatograph equipped with a flame ionization detector. Acetone (Spectrograde), acetaldehyde and biacetyl were also obtained from Eastman Kodak Co. and (1) This research was supported by a grant from the U. S. Public Health Service, Department of Health, Education and Welfare. ( 2 ) Sational Academy of Sciences-Xational Research Council Pust-doctoral Research Associate 1962-19133. ( 3 ) For a review see: E. \%’. I
: r 7 1
Fig. 1.-Phosphorescence emitted by biacetyl-acetaldehyde and biacetyl-propionaldehyde mixtures.
Effect of Biacetyl on the Emission.-As is shown in Figs. 1 and 2, a rather strong phosphorescence emission was observed when relatively small amounts of biacetyl were added to C2H,$2HO, nC3H7CH0and i-CaH7CHO. Addition of oxygen to any of these mixtures always reduced the emission to the same yield as that observed for the pure aldehydes, indicating that the relative fluorescence yield was not affected by the addition of biacetyl. From the initial slopes 9f Figs. 1 and 2 i t may be concluded that a t 3340 A. and 33' biacetyl is most effective in the case of CH3CHOjand least effective in the case of i-C3H7CH0. .4t relatively high biacetyl concentrations the emission from CHsCHO CIHBCHOand n-C3H7CH0, respectively, levels off and approaches approximately the same relative value. The plots in Fig. 1 clearly indicate that a decrease in wave length or iricrease in temperature ( 5 ) T h e d a t a on CHaCHO have been included fox comparison purposes. An extensive study on tbe fluorescence and phosphorescence of C H C H O has been carried o u t by C. S. Farmenter and W. A. Xoyes, Jr., and will be published shortly.
I
1
2
E 3 b;iTvL
*2ESSdiE
m r ,
Big. 3.-Effect of biacetyl on the eth)-lene formation frviii the photolysis of %-butyraldehyde.
Discussion Fluorescence.-The results of Tables TI to V indicate that although the observed variations of the relative fluorescence are rather small, there are two distinct trends which have to be considered: (a) the yueiiching of the fluorescence with increasing concentration a t the longer wave lengths and (b) a t the shorter wave lengths the enhancement of the emission with increasing concentration.
VAPORPHASE FLUORESCENCE OF ALDEHYDES
Nov. 5, 1962
4047
TABLE VI RELATIVEBIACETYLPHOSPHORESCENCE AS A FCXCTIOX OF TOTAL PRESSURE FOR VARIOUS BIACETYL-ALDEHYDE MIXTURES X,, = 3310 a , T = 33' CzHsCHO-n-CzH?CHO--z-CaH,CHOC2HsCH0, hex = 3130 %, , T = 33" r--CHaCHO--
-
-
--O 112-----. f i , m r n . 0 x 102
108 78 56 40 29 14
7 - 0 011L-y p , mm. 0 X 103
7-0 099--p , mm. Q X 30' r-
I /
56 40 29 21
72 72
99 71 52 36 26 19
93 95 92 88 86 81
in8
105 107 108 109 104 106
Biacetyl/aldehyde -0 167--7 p , mm Q X 1 - 0 2
68 49 35 25 18 13
i1
67 65 59
121 124 121 108 99 95
7
r--0 175------. p , mm Q X 102
26
33 28 20 14 10
19
5
$14 GY 49 35
7 - 0 -088p , mm. 0 X 102
76
184 133 95 69 49 25
72 03 53 40 20
-0
0 0 b -
p , mm. Q X 102
108 99 7% 52 37 27
39
37 30 22 15 10
TABLE VI1 order process such as PHOTOLYSIS OF n-CaH7CH0 IN THE PRESESCE OF BIACETYL Ao' A ---+ 2A (6) .4t 3304 A,, T = 30", p(n-CsH?CHO) = 51.5 mm. The fact that a t the higher temperatures the fluores-
+
9
RCO
x
t,
Rc~r
x
(CHaCO)z, mm.
min.
cc./rnin.
0 0.43 1.4 4.7 9.6
150 150 150 150 150
1.72 1.64 2.74 4.86 8.23
104,
104,
Rc~H~
x
1-04,
cc.irnin.
cc./min.
..
0.51 .34 .14 ,084 ,058
0.0; .32
.45
.95
Rca11,
x
104, cc./min.
1.2 1.96 1.65 2.42
..
Similar effects of concentration on the fluorescence emission have been reported for CH3CH0.6 I t may also be noted that in the case of acetone, self-quenching of the fluorescence and phosphorescence has been reported in ai1 earlier More recently,8 however, no apparent effect of pressure was observed on either the phosphorescgnce or the fluorescence emitted by acetone a t 3130 A., while a t the shorter wave lengths the phosphorescence emitted by acetone shows a pronounced increase m ith increasing pressure. A similar observation was made in a study carried out in this Laboratoryg on the phosphorescence emitted by CH8COCF3. In the latter case i t was also shown that a t the longer wave lengths the phosphorescence undergoes self-quenching. In this respect the fluorescence emitted by the aldehydes follows similar trends with pressure as the phosphorescence emitted by CH,COCF, and CHaCOCH:,. The effect of pressure on the fluorescence emission of the aldehydes c a i i be accountcd for by the reaction scheinc
+
-4
iLV
--+
.I','
.4,' --+ decompositioii A,' --+ A,,3
An'
+ A +Ao' + A +-4 + hv -40'
(1) (2) (3) (4) (5)
where A,' is an aldehyde molecule in a vibrational level of the excited singlet state, Ao' an aldehyde molecule in the lowest vibrational level of the upper singlet state, Am3,an aldehyde molecule in an upper vibrational level of the triplet excited state. Step 3 has been included here to account for the presence of triplet excited aldehyde molecules. The pressure effect observed a t longer wave lengths for C2H5CH0 and C2FbCHO necessitates the inclusion of a second(6) E. Murad, J. Phys. Chem., 6 4 , 942 (1960). (7) H. J. Groh, Jr., G. W. Luckey and W. A. Iioyes, Jr., J. Chem. Phys., 21, 115 (1953). ( 8 ) J. Heicklen, J. A m . Chem. SOL., 81, 3863 (1959). (9) P. Ausloos and E. Murad, J. Phys. Chcm., 6 6 , 1519 (1961).
cence becomes independent of pressure a t 3340 A. is consistent with the occurrence of reaction ( 6 ) a t low temperatures. Effect of Biacety1.-It has been demonstrated beforelo that the addition of biacetyl to acetone enhances preferentially the phosphorescence emitted by biacetyl, while the acetone singlet fluorescence efficiency remains unaffected.ll These observations have been explained by a transfer of excitation energy from acetone to biacetyl. -403
+B
=
B3
+A
(7)
The validity of this interpretation was corroborated by the fact that addition of biacetyl to acetone also causes a reduction of the primary quantum yield of decomposition of acetone. At relatively high biacetyl concentration, the remaining product yield was ascribed to the decomposition of the upper singlet state of acetone. hiore recently, a study carried out in the same laboratory12 on the emission and decomposition yield of diethyl ketone in the presence of varying amounts of biacetyl was interpreted by essentially the same mechanism. It may be suggested that the results obtained in the present investigation can also be explained by the energy transfer reaction 7. Figure 1 illustrates that by addition of only 1% of biacctyl to acetaldehyde, a yield of phosphorescence is obtained which does not undergo any further increase with increase in pressure of biacetyl. Also the results of Table VI show that the emissioii yield in the plateau region is independent of total pressure. A plot of the phosphorescence emitted by C*H&HO-(CH,CO)? mixtures versus !he pressure of biacetyl (Fig. 1) shows that a t 3340 A. the initial slope is less than the one observed for CH3CH0, while the efficiency of biacetyl is still smaller for the mixtures containing n-C3H,CH0 and i-C3H7CH0 (Fig. 2 ) . It may, however, be noticed that for the latter two compounds the effect of total pressure on the yield of phosphorescence is also more pronounced. These observations can be accounted for by a competition between the reactions An'-+-Am3
+D
Ama
(3) (8)
(10) H. Okabe and W. A. Noyes, Jr., J . A m . Chem. Soc., '79, 801 (1957). (11) J. Heicklen and W. A. Koyes, Jr., ibid., 81, 3858 (1959). (12) D. S. Weir, {bid., 83, 2629 (1901).
K.P. BORKOWSKI A N D P. XUSLOOS
+ +
+ +
.Im. il +A " 3 x Ao3 B +A B3 B 3 +B hv
+
(9) (7) (10)
- i o 3 represents a low vibrational level of the triplet
state. It should be noted that instead of step 3, the following reaction sequence may eventually have to be considered .in' + A +.\"' + x (4) XO'
+
Am3
(11)
Vol.
+
T Z - C ~ H ~ C H Ohv
+C ~ H I+ CHBCHO
s-l
(I)
The yield of ethylene should a t reasonably low conversions not be affected by secondary reactions and may consequently be considered as a reliable measure of the relativeoquantumyield of process I . The fact that a t 3340 A. biacetyl reduces the relative rate of formation of CaH4by nearly a factor of ten indicates that process I can be readily quenched by biacetyl. I t may consequently be concluded that an excited triplet aldehyde molecule is involved in process I. The correlation between the quenching of the primary decomposition and the formation of the triplet excited biacetyl is demonstrated by the plot ( K c - Xi) given in Fig. 3 which levels off a t about the same pressure of biacetyl as the phosphorescence which is presented in Fig. 2 . The fact that addition of biacetyl does not appreciably affect the yield of propane may be considered as an indication14 the dissociative process
,kcording to the mechanism given above the difference in magnitude of the effect of total pressure for the various mixtures can be accounted for by a variation in dissociative lifetime of the triplet or singlet excited molecule according to the sequence i-CaH7CHO > n-CzH7CHO > CsHsCHO > CH3CHO. If one accepts reaction 3 it may also be expected that the shorter the incident wave length, the higher CaH;CHO + hv +CaH; + CHO the degree of vibrational excitation of Am8. It (11) thus follows that a higher pressure will be required occurs from a singlet electronic state and/or from a to deactivate the triplet aldehyde molecule to the triplet electronic state excited to a high vibrational lower vibrational levels from which energy transfer level. to biacetyl can take place. In agreement with this Recent observations'j seem to contradict our interpretation is the observation that the phos- conclysion that the olefin elimination process a t phorescence emitted ky C2H5CHO-(CH3C0)2mix- 3340 A. occurs by way of a triplet excited molecule. tures is lower a t 3130 A. than a t 3340 A.but is more It has indeed been suggested by Borrell and Norrish strongly snhanced by an @crease in total pressure that in view of the very low yield of ethylenelfioba t 3130 A . than a t 3340 A. I n this connection it served in the mercury-sensitized decomposition of may be noted that recent measurements of the n-C3H7CH0, as compared to the direct photolysis, phosphorescence emitted by CH3COCH?v9 and only aldehyde molecules excited to the upper singlet CH3COCF39did show that a t the lower incident state can undergo process I . The latter conclusion wave length the emission is enhanced rather was based on the Wigner spin conservation rule. strongly by increasing the pressure. These observa- If we accept that this selection rule does hold for tions, which have been accounted for by a very heavy atoms, the triplet aldehyde molecule formed similar mechanism,x indicate that the absence of in the process any phosphorescence or e f i c t of biacetyl is not neces.I Hg ("1) +&n3 Hg sarily proof that only molecules excited to an upper singlet state undergo decomposition, but can just can be expected to be in a much higher energy level as well be explained by the formation of molecules than the one produced a t 3340 A. This may lead excited to high vibrational levels of the triplet state to a considerably higher relative probability of the in process 3. I t is obvious, however, that the effects dissociative process 11, which is energetically less of pressure a t the different wave lengths could favorable than the intramolecular rearrangement equally well be explained by reactions 4 and 11. process I. On this basis all one can deduce from a From the data presented in this paper i t cannot be comparison of the products formed in sensitized and deduced what fraction of the aldehyde molecules non-sensitized decomposition is that decomposition excited to the upper singlet state undergo a transi- of a molecule excited to a high vibrational level of tion to the triplet state. The rather small effect of the triplet state may not contribute appreciably wave length and structure on the relative fluores- to the formation of ethylene. Considering the cence yield for the various aldehydes would seem information which is available a t present it may be to indicate that the importance of reaction 3 or 11 tentatively suggested that the occurrence of process does not undergo any drastic variations. This is I by way of an excited triplet state prevails a t the corroborated by the fact that a t 3340 A.the maxi- longer wave lengths, while an excited singlet state mum phosphorescence yield emitted by biacetyl is may contribute to the decomposition a t the shorter of the same order of magnitude in mixtures contain- wave lengths. ing CHaCHO, C2HsCH0 and n-C3HiCH0, re(14) As evidenced by the formation of methane a t high biacetyl concentrations, direct decomposition of biacetyl becomes important spectively. The Effect of Biacetyl on the Primary Decomposi- and complicates t h e reaction mechanism. (15) P. Borrell and R. G. W. Sorrish, Pioc. R o y . SOC.(London), tion.-It is well known13 that by absorption of a A262, 19 (1961). photon n-CaH7CHO undergoes an intramolecular (le) A few sensitized experiments which were carried o u t in this rearrangement resulting in the elimination of Laboratory did show t h a t t h e yield of ethylene is smaller than in t h e direct photolysis. b u t b y no means negligible. At 80' a value of 0.08 ethylene. was obtained for the ratio C ~ H I / C O . From the pressure dependence
+
(13) F. E. Blacet and J. G. Calvrrt, J . Afu. C h m . Soc., 73, U i l ( 19.51 ) .
+
on the rate of formation of CgHI i t could he concluded t h a t the ethylene wa6 mainly formed by a sensitized decomposition.
