10
The Journal of Physical Chemistry, Vol. 82, No. 1, 1978
N. Washida, H. Akimoto, and M. Okuda
S. R. Fahrenholz, F. H. Doieiden, A. M. Trozzolo, and A. A. Lamola, Photochem, Photobiol., 20, 505 (1974). L. A. Paquette, D. C. Liotta, and A. D. Baker, TetrahedronLett., 3681 (1976). (24) 6. C.'Frost and J. S. Sandhu, Indian J , Chem., 9, 1105 (1971). (25) J. W. Turner and W. Herz, J . Org. Chem., 42, 1657 (1977).
(26) C. Dufraisse, J. Rlguady, J.J. Basselier, and N. K. Cuong, C . R. Acad. Sci., 260, 5031 (1965). ( 2 7 ) D. F. Evans and J. N. Tucker, J . Chem. Soc., Faraday Trans. 2 , 72. 1661 11976). (28) C. 'S.Footk, E. R. Peterson, and K.-W. Lee, J. Am. Chem. Soc., 94, 1032 (1972).
Formation of Singlet State Molecular Oxygen in the Reaction of H N. Washlda,"
H. Akimoto,
+ O2
and M. Okuda
National Institute for Environmental Studies, P. 0. Yatabe, Tsukuba, Ibaraki, 300-21 Japan (Received June 8, 1977) Publication costs assisted by the National Institute for Environmental Studies
Singlet state molecular oxygen was detected in the reaction of hydrogen atoms with O2 using a photoionization mass spectrometer. Singlet oxygen could be photoionized by argon reeonance lines. Signals of singlet molecular oxygen were proportional to [H][02][M]t,as reported most recently by Hislop and Wayne.g The following mechanism is proposed for the formation of singlet oxygen in this system: H + 02 + M H02 + M (1); H02 + H Hz 02*(singlet) (2d). Reaction 2d was estimated to be about 0.015 f 0.003 of the total reaction of HO2 H. Measurement of the emission of 02(b1Zgt)showed that at least part of the singlet molecular oxygen initially formed in reaction 2d is 02(b1&+). Adding the results by Westenberg and de Haas and Hislop and Wayne to our results, the fractions for the formation of molecular oxygen in the reaction of H O2 are obtained as follows: HOz H H2+ 02(X32,-),f = 0.60; Hop + H H2 + 02(a1A,),f = 0.015; HOz + H H2 + 02(b1X:), f = 0.000 28.
-+
-
+
+
-
-
Introduction It is well known that the reaction of H atoms with O2 forms €302 radicals in the third-order process H+O, t M - H 0 , t M
(1)
HO2 formed in reaction 1 has been detected directly by IR absorption,l laser magnetic resonance,' and mass ~pectrometry.~ There are numerous direct measurements of the rate constants for reaction 1 near 300 K, which are summarized in recent report^.^^^ An excess of H atoms destroys the H02 radicals via reactions 2. There are several reports6 on the fraction ratio /* H, t Q, A H = - 57.1 k c a l / m o l (2a) HO, t H-,c H,O t 0 ~ H = - 5 5 . 3k c a l / m o l (2b) OH t OH A H = - 38.5 k c a l / m o l (2c) of reaction 2. The most recent report7 showed the fol-
lowing ratios using ESR measurements of the absolute H atom and steady state OH and 0 atom concentrations: ilt2a:k2b:k2c= 0.62:0.11:0.27. The rate constant for reaction 2 seems to be very fast as expected for a radical-atom reaction. Recently Giachardi et ale8found the emission spectra of H02and 02(b'Z1 +) in the H O2 reaction system. Very recently Hislop and Wayne9 measured the emission spectra of 02(b124+)for various experimental conditions and reported that the oxygen formed by reaction 2a includes singlet state molecular oxygen 02(b1Zg+)in a fraction of This report concerns the measurement of singlet state molecular oxygen formed in the reaction of H + O2 using a photoionization mass spectrometer, which can be used to detect singlet state molecular oxygen, Oz(alAg) and 02(b1Zg+).lo
+
Experimental Section Experiments were carried out using a cylindrical Pyrex fast-flow reactor coupled to a photoionization mass spectrometer, as described previous1y.l' The hydrogen atoms were generated by microwave discharge in a He-H' mixture. The oxygen was added
+
-
downstream through a multiholed movable inlet which was coaxial with the hydrogen atom flow. The concentration of hydrogen atoms was determined by titrating with an excess of NO2 and measuring the resulting NO with the photoionization mass spectrometer. The sensitivity of the instrument was calibrated with known partial pressures of NO. The total and partial pressures of gases were measured by an MKS Baratron gauge. The linear flow velocity in the reaction was 13.2 m s-l at 3 Torr and 14.0 m s-l at 4 Torr. The gases were Hz (Nippon Sanso, 99.99999%), He (Nippon Sanso, 99.9999%), O2 (Nippon Sanso, 99.99%), NO (Matheson, 99.0%), and NOz (Matheson, 99.5%), all used without further purification. The reacting gases were sampled through a thin Pyrex pinhole (0.2 mm diameter) into the ionization region of a quadrupole mass spectrometer. Ions were counted by a Daly type detector,12 a thin aluminum-coated plastic scintillator, and pulse counting electronics. An argon lamp with a LiF window was used to photoionize singlet molecular oxygen and NO. The argon resonance lines (11.83 and 11.62 eV) will ionize 02(a1Ag)and Oz(blZg+),but not ground state 0'(X32,-).10 The concentration of singlet state molecular oxygen was calculated from the relative ionization cross section for NO and OZ(a1Ag),l3 assuming the ionization cross sections for 02(a1Ag)and 02(b1Z.,f)are equal. The measurement of the emission spectrum of 02(b1Xgt) was done using another flow tube, a monochromator (Nikon P 250), a photomultiplier (Hamamatsu T.V. R649-S), and a pulse counting system.
Results and Discussion Mass spectra of the H atom plus O2 reaction are shown in Figure 1. A large signal at mass 32 (singlet state 0,) and a small signal at mass 34 (H202and 1601sO)could be detected. Other signals could not be found over a wide mass range. The signals at mass 32 increased linearly with the addition of O2 for constant H atom concentration, constant
0 1978 American Chemical Society 0022-3654/78/2082-001~8$01.00/0
The Journal of Physical Chemistry, Vol. 82, No. 1, 1978
Formation of Singlet State Molecular Oxygen
b) H
c ) 02
19
1 1
I
JL OO2
10 [H
I
i
c
n
I H 1t mTorr
Flgure 2. Singlet molecular oxygen signals obtained by the addition of 0,: [He] = 3.35 Torr, [HI = 26.15 mTorr, and reaction time = 5.86 ms.
total pressure, and constant reaction time (Figure 2). On the other hand, for constant O2 concentration, the mass 32 signals were proportional to the H atom concentration (Figure 3). Since the absolute concentrations of singlet oxygen are about 50 times higher than the estimated value for the concentration of Oz(b12++)by HWY9it is probable that most of the singlet state molecular oxygen formed in this reaction is in the ala, state, because it is difficult to imagine that the cross section for photoionization by the argon resonance lines for 02(b1Z,+)is a factor of 50 larger than that for 02(a1A,). The reaction time dependence of 02*signals were measured by changing the movable inlet of the fast flow reactor. The results are plotted in Figure 4. Plots against the product of H atom concentration and reaction time for two different H atom concentrations yielded a single straight line. The hydrogen atom loss by O2 in the flow tube could be neglected. The hydrogen atoms are consumed by reaction 1,followed by fast reaction 2, and therefore two hydrogen atoms should be consumed by a reaction with 02. The
0
Flgure 3. Signals of singlet molecular oxygen as a function of H atom concentration: [He] = 3.45 Torr, [O,] = 0.110 Torr, and reaction time = 5.86 ms.
50
I O2 I
20 1 mTorr
10'~ molecule ~ e ccm-3
Figure 4. Singlet molecular oxygen signals vs. [HI tfor two different conditions: (0)[He] = 3.44 Torr, [HI = 26.15 mTorr, [O,] = 0.114 Torr; ( 0 )[He] = 3.44 Torr, [HI = 6.71 mTorr, [O,] 0.113 Torr.
calculated hydrogen atom losses were less than 10% of the initial concentrations for most of the points in Figures 2-4. Only a few points gave losses near 15%. The quenching of singlet oxygen by hydrogen atoms seem to be very slow, otherwise Figure 4 would not give a straight line. The following mechanisms are possible for the formation of 02(a1A,) during the reaction of H + 02: HO, t H-+ H, t O,*(singlet)
(a)
(2a)
H 0 2 formed in reaction 1reacts with an excess of H atoms and forms Oz*(alA and/or blZ,+). This mechanism was proposed by HWggas the formation mechanism of 02(blZg+). (b) HO,
+ H-
H2*+ 0 ,
H,*
+ 0,
(3)
H,t O,* (4) In this case, vibrationally excited hydrogen molecules are formed and then energy is transferred to the oxygeq molecules. (c) H
-+
+ H t 0,
(d) 2H0,
-+
4
H,O,
H,
+ O,*
+ O,*
(5)
(6)
20
The Journal of Physical Chemistry, Vol. 82, No. 1, 1978
N. Washida, H. Akimoto, and
M. Okuda
TABLE I : Estimated Values of k z d / k , Initial conditions Figure
[He], Torr
2 3 4
3.35 3.45 3.44 3.44
0
0.88
2.61
3.56
0
5 5
[N,], Torr 0 0
[ H I , mTorr 26.15
Varied
0
Reaction time, ms
[O,], Torr Varied 0.110 0.114 0.113 0.110 0.112
26.15 6.71 11.35 12.58
kzdlk,
5.86 5.86
0.012 0.014
Varied Varied Varied Varied
0.013 0.019 0.019
Av 0.015
AH = -42.5 kcal/mol for this reaction makes it possible to excite molecular oxygen.
-
( e ) HO, t H -+ OH* OH* t 0, OH
-+
H,O
i
k
0.003
I
I
I
1
I 2
I 4
I 6
a
+ OH + 0,"
This formation mechanism is similar to mechanism b. ( f ) OH t HO,
__
+ O,*
(9)
OH formed in reaction 2c reacts with H 0 2 and AH = -72.1 kcal/mol is enough energy to excite molecular oxygen. (9) H + 0, t (M) HO,* + ( M ) HO,* + 0 , HO, ~ + O,(a'Ag)
(10)
+
(11)
This mechanism has been proposed by Giachardi et al.s The results in Figures 2-4 show that the concentration of 02* formed in the reaction of H O2 is proportional to [H][02]t. The reaction of H atoms with O2 forms HOz radicals in the first step by reaction 1. H02radicals react with an excess of H atoms in reaction 2. Reaction 2 is thought to be very fast reaction, near cm3 molecule-1 s-l, since it is an atom-radical reaction. Therefore H 0 2 radicals should reach their steady state concentration for our experimental conditions, hydrogen atom concentrations (5 -30 mTorr), and reaction times (0.7-13 ms). The steady state concentration of H02radicals is shown by eq I. If singlet state oxygen is formed by mechanism
100-
E
c -s' eol
- 50-
t0 N
I
+
[H02I 8s = ( h i / k 2 ) [ 0 2 ][MI
(1)
a, the concentration of 02*(singlet)is represented by eq I1 where hzd refers to the rate constant of reaction 2d. = ( h 2 d / h 2 ) h l [HI Lo21 [MI t (11) The results in Figures 2-4 satisfy eq 11. The three body effect was examined using nitrogen as a carrier gas, as shown in Figure 5. The signals for singlet molecular oxygen increased in the N2 system. The increased slope in Figure 5 agreed well with the reported rate constants for reaction 1, kIN2/hlHe = 3.4.4 These results show that the concentration of singlet oxygen is proportional to [H][02][M]t and mechanism a agrees well with these results. If singlet oxygen is formed by mechanisms b and e, [02*] should be proportional to [HI [O2I2[M]t,which disagrees with our results in Figure 2. In the case of mechanism c, [02*] should increase with [HI2. This also disagrees with the results in Figure 3. For mechanisms d and f, 02*should be proportional to [O2I2[Ml2tassuming a steady state for HOz and OH. The result in Figure 3 shows that [02*] is not independent of H atom concentration. According to the mechanism of Giachardi et a1.,8 mechanism g, 02*should be independent of a third body, M, which does not satisfy the present results. The results in Figures 2-5 suggest strongly that singlet oxygen is formed in reaction 2d. Since the rate constants for reaction 1 have been r e p ~ r t e d the ,~,~ value of kzd/k2 in eq I1 can be estimated from each slope of Figures 2-5. The
lo,*]
0.
0 0
i H Ii 10'2molecule 5ec cm-3
Flgure 5. Singlet molecular oxygen signals vs. the product of reaction time and hydrogen atom concentration for two different conditions: (0) [He] 0.88 Torr, [N2] 2.61 Torr, [HI = 11.35 mTorr, [O,] = 0.110 Torr; (0)[He] = 3.56 Torr, [N2] = 0 Torr, [HI = 12.58 mTorr, [O,]= 0.1 12 Torr.
value of 1.57 X cm6 molecule-2 s-l was used for hlHe and 5.33 X cm6 molecule-2 for hlNz, which were reported by Kury10,~because the values of Kurylo4 and of Wong and Davis5 are very close and we had no reason to select one. For the hlo2, the value of klNz was used assuming that third-body efficiencies of Nzand O2 are quite similar. The most difficult problem is the absolute concentration of 02*.The values in Figures 2-5 were calculated from the relative ionization cross section for NO and 02(a1A,).13If all of the singlet oxygen is 02(a1A,), the values are correct. As mentioned before, the large difference between the estimated concentration of singlet oxygen in the present results and that by HW for 02(b11;p+) means that most of the singlet state molecular oxygen observed in these experiments is in the alAgstate, although O2(b1Zg+)is also formed initially in reaction 2a and the ionization cross sections of 02(b1Zg+)for the argon resonance lines are unknown. Calculations for k2d/k2 were done using the slopes in the figures and eq 11. The results, as listed in Table I, are in good agreement. The average fraction for reaction 2d is 0.015 f 0.003 of the total reaction of HOz H. The reported fraction kz,/k2 is 0.62 by WH' and 0.33 by Clyne and Thrush.14 If we use the fraction for reaction 2a, 0.62,' and the fraction for the formation of the O,(blZ,+), 2.8 x the following summary can be made: H, t 0 , ( X 3 X g - ) f = 0.60 P f = 0.015 HO, + H - H, + O,(a'Ag)
+
\
H,
+ Oz(b'Zg+)
f = 0.00028
The result of the N2 system in Figure 5 supports our concentration estimate of singlet oxygen. Both 02(a1A,)
The Journal of Physical Chemistry, Vol. 82, No. 1, 1978 21
Formation of Singlet State Molecular Oxygen
reaction 13, 5 X cm3 molecule-' 8,typical concentration of the Oz(alAg),5 X 10'l molecule ~ m -and ~ , reaction time, 10 ms, give the concentration of the 02(b1Z,+)formed ~ , is too low by reaction 13, 1.3 X lo4 molecule ~ m - which compared with the directly formed 02(a'Ag) and 02(b1Z+) by reaction 2d. HW also could not find evidence of t i e energy pooling reaction 13 in the measurement of the 02(b'Zg+)emission. The concentration of the 02(a1Ag)in the HW experiments can be estimated from the fraction of k z d / k z obtained in our study. The estimated concen~ , their retration of 02(a1A,), lo1' molecule ~ m - and action times, 50 and 130 ms, also give too low a concentration of Oz(blZg+)compared with the directly formed ~. it is reasonable Oz(b'Z,+), lo9 molecule ~ m - Therefore that we and HW could not find any effect of the energy pooling reaction in the measurement of the singlet oxygen. HW measured the IR chemiluminescence of H 0 2 in the H O2 reaction system. They reported that the H02* bands were proportional to [H][02][M], in accord with reaction 10. They reported also that they could not find evidence of the emission of Oz(a'Ag) at 1.26 pm, where the emission of HOz ('A', 001 2A", 000) overlapped. Our results show that the concentration of 02(a1Ag)is fairly high. It is difficult to estimate whether the emission from the 1011-1012 molecule cm-3 of Oz(alAg)is strong enough to compete with the emission from H 0 2formed by reaction 10 or not, since the rate constant of reaction 10 and the radiative lifetime of HOz at 1.26 pm are known. N
N
+
I
1w
U
LO21
I
I
200
300
-
rnTorr
Figure 6. Emission intensity of O2(biZg+) at 762 nm vs. [O,].
and 02(b1Z:B+)have long radiative lifetimes compared to our experimental reaction times, and the quenching of 02(b1Zg+)by O2 and He is negligibly small according to the reported rate constants.'6 However the deactivation rate by N2,152.2 X cm3 molecule-' s-l, is fast enough to quench 02(b1Zg+).Using the above rate constant one can calculate that 85% of the Oz(b1Zgt) should be quenched by the 2.6 Torr of Nz in the 10-ms reaction time. Since no decrease in the singlet oxygen signals was observed (see Figure 5), this means that either the 02(b1Z +) is quenched by Nz and forms 02(a1Ag),as suggested by Braithwaite et a1.16 O z ( b l z g +t) N, o z ( a l A g ) + N, (12) -+
and that the difference of the ionization cross section between 02(b1Zg+)and 02(a'A,) is small, or that most of the singlet oxygen produced by reaction 2d is 02(a'Ag). Considering our results together with the HW results, the latter is more reasonable. Emission measurement of the 02(b1Z+ -+ X3ZP+)system gave results similar to those reported %yHW. The (0,O) band at 762 nm increased linearly with the addition of O2 (Figure 6). This result means that the 02(b1Z,+)is formed initially by reaction 2d rather than by the energy pooling reaction 20,(a'A,)
-+
O Z ( b 1 x g ++) 0 , ( X 3 x e - )
(13)
The rate constant for the energy pooling reaction 13 has been reported by Derwent and Thrush.l' The rate of
Acknowledgment. We thank Professor Kyle D. Bayes for his kind suggestions and T. Nishijima of ULVAC for his technical assistance in using the photoionization mass spectrometer.
References and Notes (1) H. E. Hunziker and H. R. Wendt, J. Chem. Phys., 60, 4622 (1974). (2) H. E. Radford, K. M. Evenson, and C. J. Howard, J . Chem. Pbys., 60, 3178 (1974). (3) S. N. Foner and R. L. Hudson, J . Chem. Phys., 21, 1608 (1953); 36, 2681 (1962). (4) M. J. Kurylo, J. Pbys. Chem., 76, 3518 (1972). (5) W. Wong and D. D. Davis, Int. J. Chem. Kinet., 6, 401 (1974). (6) A. C. Lloyd, Int. J . Chem. Kinet., 6, 169 (1974). (7) A. A. Westenberg and N. de Haas, J. Phys. Chem., 76, 1586 (1972). (8) D. J. Giachardi, G. W. Harris, and R. P. Wayne, Chem. Phys. Lett., 32, 586 (1975). (9) J. R. Hislop and R. P. Wayne, J . Chem. Soc., Faraday Trans. 2 , 73, 506 (1977). (10) I. T. N. Jones and K. D. Bayes, J. Chem. Phys., 59, 3119 (1973). (11) N. Washida and K. D. Bayes, Int. J . Chem. Kinet., 8, 777 (1976). (12) N. R. Daly, Rev. Sci. Instrum., 31, 264 (1960). (13) I. D. Clark and R. P. Wayne, Mol. Phys., 18, 523 (1970). (14) M. A. A. Clyne and B. A. Thrush, Proc. R. Soc. London, Ser. A , 275, 559 (1963). (15) L. R. Martin, R. B. Cohen, and J. F. Schatz, Chem. Phys. Left., 41, 394 (1976). (16) M. Braithwaite, E. A. Ogryzlo, J. A. Davidson, and H. I. Schiff, J. Faraday Trans. 2 , 72, 2075 (1976). Chem. SOC., (17) R. G. Derwent and B. A. Thrush, Trans. Faraday Soc., 67, 2036 (1971).