1894
ROBERTM. HOFFMAN [CONTRIBUTION FROM THE
VOl. 56
LABORATORY OF PHYSICAL CHEMISTRY, UNIVERSITY OF WISCONSIN]
The Photolysis of p-Caryophyllene Nitrositel BY ROBERTM. HOFFMAN
It was observed in 1898 by Kremers2 that ,Bcaryophyllene nitrosite is decomposed in solution by red light with the liberation of nitrogen. It is also decomposed by heating to 80-110°.3
@z:6sHP4~,~a
gen liberated, and = yield of nitrogen molecules per quantum, using the notation of Holmes and Daniels* in which the experimentally measured product or reactant is designated by a superscript and the material which absorbs the quantum by a subscript.
Experimental Procedure The nitrosite was kindly given to us by ProTABLEI fessor Edward Kremers of the Pharmacy DeQUANTUMYIELDS FOR NITROGEN partment of the University of Wisconsin. Its x, A. Lamp t , min. p , CC. o ~ ~ ~ ~ , ~ ~ , absorption spectrum has been studied by M i t ~ h e l l . ~ 6439 Cd arc 57.5 0.124 0.40 He found that absorption in the visible region was 6439 58 .133 .39 confined to a narrow band with its head at about 6439 58.5 .lo3 .33 6800 A. 'This was confirmed in our laboratory. Zn arc 65 ,083 .36 6362 6362 43.5 .053 .44 Monochromatic light was obtained with a small Tungsten 360 .I90 .35 6800 Hilger monochromator. The light sources were quartz capillary lamps6of cadmium or zinc, and in A v . 0.38 some cases tungsten lamps. The thermopilePrecautions were taken in these experiments to galvanometer system was calibrated with a standexclude oxygen, as its presence affected the quanard lamp (C-132) furnished by the Bureau of tum yield. Standards. A micro gas buret was attached diQuantum Yields by the Optical Method.rectly to the reaction cell and thermostated a t 28'. In measuring quantum yields by the bleaching The volume of nitrogen liberated was measured to out of the color, two difficulties must be over0.001 cc. 'by weighing displaced mercury. Nitrocome. First, Beer's law may not be strictly benzene was used as the solvent in these experiapplicable to the absorbing system. Second, from ments on the evolution of gas, on account of its low the nature of the measurement, the light transvapor pressure. mission must undergo a large change and this fact In the optical method, a ten-volt tungsten lamp makes it difficult to determine the exact amount of was used with a storage battery floated across the energy absorbed. Neither log I,,/I, which is prodynamo circuit to decrease variations in intensity. portional to the concentration, or the absorption If there was any loss in the monochromatic nature (l-I)/Io will yield a straight line when plotted of the light by the use of the tungsten lamp, it did against time; therefore the value at the midpoint not affect the quantum yields within the range of will not give the true average absorption. Furexperimental error. Two identical cells were used, thermore, a value for the average absorption is one filled with pure solvent as a reference. Toluene useless if the intensity of the light source has was used as the solvent in these experiments. changed during the experiment. The dark reaction at room temperature was These difficulties were overcome in the followf0un.d to be negligible. ing general method. The initial concentration is Experimental Results determined either by starting with a known amount, or more conveniently by reference to an Quantum Yields for Nitrogen.-These results experimentally determined curve relating concenare summarized in Table I where X = wave tration to light absorption as in Fig. 1for p-caryolength, t = time in minutes, v = volume of nitrophyllene nitrosite. The photolysis is continued (1) Further details concerning this work may be obtained from a thesis prepared under the direction of Farrington Daniels and subuntil the final concentration is zero, as shown by mitted by the author to the University of Wisconsin in 1933 in partial the light transmission becoming 100%. The abfulfilment of the requirements for the degree of Doctor of Philosophy. (2) E. Kremers, Pharm. Arch., 2, 273 (1899). sorbed energy is calculated by averaging both the (3) Valenzuela and Daniels, PhiliQpine J . Sci., 34, 187 (1927). absorption and the light intensity over short in(4) Mitchell, J . Chem. Soc., 3238 (1928). (5) Hoffman end Daniels, THISJOURNAL, 54, 4225 (1932).
(6) Holmes and Daniels, ibid., 66, 633 (1934).
THEPHOTOLYSIS OF /3-CARYOPHYLLENE NITROSITE
Sept., 1934
crements and summing these up. This is, in effect, a graphical integration of the total absorbed energy and is accurate whether the function is linear or not. For these data no dependence on Beer’s law is necessary. Lastly, log 1ojI (or -log 1/10)is plotted not against time but against a function which will give a straight line when Beer’slaw holds. IO = incident light I
= transmitted light
A
=
1895
molecules decomposed” since Kl is proportional to the quantum yield and contains the factor to convert the energy summation to total quanta. The quantum yields obtained by this method are shown in Table 11.
average absorption
t = time elapsed since start At = length of increment in minutes
CO = initial concentration Ct = concentration at time t C d = amount decomposed = CO- Ct then ( AtAlo) = energy absorbed during increment 2; ( A ~ A I = ~ )total energy absorbed during time t
The amount decomposed a t time t is proportional to the total energy absorbed, therefore we‘ may write Co - Ct = Ki 2: ( A t d l o ) also Ct = K2 log (I/lo)t then CO- K2 log (I/lo)t = Ki 2: (Atdlo) -log (I/lo) = (Ri/K*)2; (Atz410)
and
- Co/K2
This equation yields a straight line by plotting -log (Ij10) against Zf,( AtAlo). When -log (Illo)becomes zero, we have K I 2; ( AtA10) = CO and the Beer’s law coefficient Kz has disappeared. This equation is equivalent t o the statement that “quantum yield times quanta absorbed equals TABLEI1 QUANTUM YIELDSBY THE OPTICALMETHOD Gas present in cell
The first t h e e experiments in Table I1 show, as before, that oxygen has a marked effect on the reaction. A limited number of data for the fourth experiment in Table I1 are given in Table I11 to illustrate the calculations and all the calculated points are graphed in Fig. 2. In Table I11 IOand 1 are recorded simply as galvanometer deflections in centimeters. It will be
*ClsHaN?Oa ClrH~rN~Os
2.8 2.9 2.9 1.4 1.4 1.3
Air Oxygen Oxygen Nitrogen Nitrogen Nitrous oxide
4 8 12 Grams of pcaryophyllene nitrosite per liter. Fig. 1.-Concentration-light adsorption curve for 8-caryophyllene nitrosite in toluene, cell depth 13.5 mm., X = 6800 A.
TABLE I11
QUANTUM YIELDBY THE OPTICALMETHOD min.
cm.
10,
I, cm.
0 2
95.9 95.2 95.0
35.3 36.8 37.1
0.638 ,618 .609
206 4 7
99.8 99.2 99.3
58.0 57.8 57.8
130 60 195
80.1 80.2 67.7
75.2 75.7 64.0
A‘,
5
av.
10.
A, av.
DJAIQ
95.6 95.1
0.628 .614
300 117
300 417
0.4425 .4179 .4078
Eighteen 0.419 ,417 .417
increments omitted 0.453 102.6 .418 99.5 99.2 .417
9560 166 290
30,251 30,417 30,707
0.2350 .2343 ,2343
Eighteen 0.061 .056 .055
iricrements omitted 80.0 0.065 50.1 .059 ‘73.7 .056
675 284 804
73,927 74,211 75,015
0.0273 ,0250 .0246
1
- Ill0
2: ( A t A I d
-Log IlIo
ROBERTM. HOFFMAN
1896
noted in Fig. 2 that the reaction is much faster a t the beginning due to traces of oxygen which remain even after sweeping out the system with nitrogen. To exclude this effect, a later point is taken as the start of the reaction, namely, where -log 1,’10 equals 0.359. From Fig. 1 this value gives an initial concentration of 0.0213 g. in about five cubic centimeters of solution. Figure 2 shows that, at completion, energy corresponding to 80,000 galvanometer deflection-minutes has been absorbed. Subtracting the energy absorbed during
Vol. 56
rial was completely decomposed by direct light without using the monochromator. The nitrogen trioxide was removed either by freezing or by absorption in potassium hydroxide. Theory Reaction Mechanism.-The knowledge of /3caryophyllene nitrosite is a t present so incomplete as t o make difficult the choice of a reaction mechanism. It seems probable that the main, over-all reaction is an autooxidation as follows Reaction I 3B-CisHzaN20r
+ 2hv
--f
N2
+ CisHz4 + CisHzaNzOc $- CisHzrNzOs
A termolecular collision involving two activated molecules and one normal molecule would be very improbable, therefore the following steps are postulated to explain Reaction I.l0
20,000 40,000 60,000 80,000 ISnergy absorbed, 2’0 (Ai AI,,). Fig. 2.--Quantum yield by the optical method. 0
Step I Step IIa Step IIb Step I11 Step IV
+
B-ClSH24N20~ 6 v +P-CUHZ~NZO~* ,9-Cla&4NzOs* ---f Ct6HZ4 f NZOS B - C ~ ~ H Z ~ N+ ~ O S A-CI~HZ~NZOS * 2 A-C16Hz~NzO~ +(CI~HUNZOS)~ (CEHZ~NZOS)~ CIJHZ~NZO~ -+ Nn CisHz4 CisHz4Nz04 f CuHzrN2Oa
+
+
+
The agreement of this mechanism with our data the rapid initial oxygen reaction, one obtains is shown in Table I V where the experimental re80,000- 6250 or 73,750 galvanometer deflection- sults are compared with the theoretical results deminutes. The galvanometer factor (including a rived from Reaction I. correction factor of 1.035 for reflection losses) is TABLEIV 4.57 X loL4quanta/cm. deflection-minute. The THEORETICAL AND EXPERIMENTAL QUANTUMYIELDS molecular weight of the nitrosite is 280. The Theoretical Theoretical from (corrected quantum yield is Reaction I for NnOs) Experimental (0.0213/280) X 6.06 X 1 0 2 8 = 1.37 73,750 X 4.57 X loi4
The Gaseous Products.-Kremers has shown that nitrogen is the main gaseous product of the photolysis of this nitrosite, although later Schreiner and Kremers?state that nitrogen oxides appear to be liberated also. Using a micro gas buret similar to the one described by Blacet and Leighton,8 the gas was found to contain 13.9% (15.2 and 12.6) nitrogen trioxide and 86.1% (84.8 and 87.4) nitrogen. The nitrogen trioxide was absorbed with potassium hydroxide. As the Nz03in the buret is largely dissociated into NOz, Nz04, and NO, its undissociated volume was calculated from the equilibrium constants given by Verhoek and D a n i e l ~ . The ~ volume of nitrogen gas per mole of nitrosite decomposed is 6478 cc. This value is the average of five determinations in which the mate(7) 0. Schreiner, “The Sesquiterpenes,” Pharm. Rev. Pub. Co.. Milwaukee, 1904, p. 74. (8) Blacet and Leighton, Ind. Eng. Chcm., Anal. Ed., 8, 266 (1931). (9) Verhoek and Daniels, THISJOURNAL, 03, 1260 (1931).
@%HrrNzQs CiaHnNnOt @CSH~N~O~
Nz/CisHz4”10s
0.50 1.50 0.33
0.46 1.46 0.32
0.38 1.37 0.29
However, the theoretical values must be corrected for the occasional escape of Nz03 as in Step IIa. One hundred molecules of liberated gas would consist of 14 molecules of NzO3 and 86 molecules of nitrogen. The N& molecules would require 14 quanta and the nitrogen molecules would require 172 quanta, a total of 186 quanta. The is 86/186. Thenumcorrected value for @i!&2,N20a ber of nitrosite molecules decomposed would be 272. The corrected value for @E:$:; then is 272/186. The ratio of Nz to CN,H~~NZO~ decomposed would be 86/272. Discussion Step I consists in the absorption of a quantum of light. Since most nitrosites and even liquid (IO) The symbols used below have the following meanings: 9 , designates one of several isomers of caryophyllene [see Deussen, J. prakf. Chcm.. 114, 75 (1926)],* signifies a molecule in an activated state, and A refers to an isomer of the nitrosite as discussed later.
Sept., 1934
TRANSITION POINTS OF HYDRATED HYDRAZONIUM SALTS
Nz03 show a characteristic blue color with absorption over a narrow range at 6800 A., the N20a group must absorb the energy. The Nz03 is now completely removed as in Step IIa or shifted in position as in Step IIb, forming an isomer designated here for convenience as the delta modification. Step IIb may possibly involve the removal of N203as in Step IIa and then subsequent. rearrangement and recombination. This isomerization allows the formation of the bimolecular c13mpound as in Step 111. The final Step I V yields nitrogen, some isomeric form of caryophyllene and the two nitrosates containing the Nz04 and N 2 0 6 groups. The direct formation of nitrosates rnay account for the photochemical effect of oxygen in accelerating the photolysis as discussed previously. The intensity of the light was changed four-fold without affecting the quantum yield as shown in Table I. The straight line obtained in Fig. 2 shows that the quantum yield was not affected by a change in concentration from a 0.0152 to 0.0009 molar. The mechanism presented, while not definit ely proved, agrees well with the physical data based on quantum yields for a product as well as for the disappearance of the original molecule. Any la.ter mechanism founded on new chemical facts should be subjected to this test. It is suggested that such
1897
a procedure may be helpful in many problems involving complicated molecules. It was hoped that a study of the simpler amylene nitrosite would yield further information on this problem. Preliminary experiments showed that the quantum yield is greater and that the products contain more nitrogen oxides. The author wishes to express his gratitude for the helpful advice of Professor Farrington Daniels, throughout the course of this work. Summary 1. The photolysis of /3-caryophyllene nitrosite in red light has been measured quantitatively with a monochromator. 2. The gaseous products have been analyzed. 3. The quantum yield for nitrogen liberated is 0.38 molecule per quantum. 4. An improved method is presented for determining quantum yields by the decrease in the absorption of light. 5. The quantum yield for the disappearance of the nitrosite is 1.37 molecules per quantum. 6. Oxygen has been shown to affect the reaction. 7. A reaction mechanism has been proposed. The value of photochemical data in studying the reaction of complicated molecules is illustrated. MADISON, WISCONSIN
RECEIVED JUNE 15, 1934
[CONTRIBUTION FROM THE DEPARTMENr O F CHEMISTRY OF THE OREGON STATE COLLEGE ]
Studies on Hydrazine: Transition Points and Dissociation Pressures of Hydrated Hydrazonium Salts BY
B. E. CHRISTENSEN AND E. c. GILBERT
Only a small number of hydrazonium salts
