Decomposition of Hydrogen Peroxide on Supported Metals
1307
Formation of Superoxide Ion during the Decomposition of Hydrogen Peroxide on Supported Metals Yoshio Ono, Tsuyoshi Matsumura, Nobumasa Kitajima, and Shun-ichi Fukurumi Department of Chemical Englneerlng, Tokyo Institute of Technology, Ookayama, Meguro-ku, Tokyo, Japan (Received December 14, 1976) Publication costs assisted by the Tokyo Institute of Technology
Presence of the superoxide ion (02)in the aqueous phase during the decomposkion of hydrogen peroxide over supported metals has been revealed by quenching the reacting solution and measuring the ESR of the quenched ice. Observation of the superoxide ion in an aqueous phase gives direct evidence for the involvement of free-radical intermediates during peroxide decomposition by solid catalysts. The kinetics of hydrogen peroxide decomposition and of superoxide formation were studied. The rate of superoxide formation is expressed by d[OJ/d[W/I;1 = hz[H2Ozlt- h3[0J, where [H202],is the total analytical concentration of hydrogen peroxide, W is the amount of catalyst, and F is the flow rate of the solution. A reaction mechanism for the formation of the superoxide ion has been proposed based on the Haber-Weiss mechanism. It was also found that the superoxide ion trapped in ice decayed with second-order kinetics above -70 "C with an activation energy of 50 kJ/mol.
Introduction Decomposition of hydrogen peroxide has been the subject of many investigations. The decomposition mechanism in aqueous solution with metal cations is well established based on the classical Haber-Weiss mechan i ~ m . l - In ~ some cases, free-radical intermediates have been evidenced by the ESR However, the mechanism of decomposition over heterogeneous catalysts is by no means e~tablished.~-~' As for the decomposition over metals, Weiss proposed the following radical mechanism:" S
+ H,O,
H,O,
---+
S'
+ OH- + *OH
(1)
k
+ *OH3 H,O + .HO,
s + + o ,k3s --+s+o,
(2) (3)
k
S + .HO, -.% S + + HO; S'
(4)
k
+ H O L Z S + .HO,
(5)
Here, S and S+ represent the uncharged and charged parts of the metal surface, respectively. The principal role of the metal surfaces is in the electron transfer processes, (I), (31, (4), and (5). This mechanism is based mainly on kinetic results and no direct evidence for free-radical formation has been obtained. It is not specified whether the free-radical species exist on the surface or in the aqueous phase. Thus, it is of interest to obtain direct evidence for the formation of free-radical intermediates during peroxide decomposition by solid catalysts. For this purpose, we have applied a flow technique to radical detection.13-15 H ydrogen peroxide in an alkaline solution was passed through a column packed with the catalyst (metals supported on alumina) and was frozen at the outlet with liquid nitrogen. From the ice thus formed, a strong ESR signal due to the superoxide ion (0'3 was detected. This is the first direct evidence that the decomposition of hydrogen peroxide over heterogeneous catalysts involves free-radical processes and has the character of a heterogeneous-homogeneous reaction. Formation of the 02 ion in the aqueous phase has been studied mainly from the viewpoint of the biological importance of the radical.16-19 The way we adopted may
provide a simple method to prepare this radical in aqueous solution. In this work, the kinetics of superoxide ion formation by the decomposition of hydrogen peroxide over supported metals will be examined. The thermal stability of the radical trapped in ice will be also given.
Experimental Section Catalysts. Platinum, palladium, and silver supported on alumina were obtained from Nippon Engelhard. The diameter of the pellets is 1.0-1.5 mm and the metal content is 0.5 wt 5%. Gold supported on alumina was prepared as follows: Alumina from Surnitomo Chemicals was sieved into 15-24 mesh and heated in air a t 550 "C for 5 h. The alumina was then immersed in an aqueous solution of HAuC14 (obtained from Wako Pure Chemical Industry) overnight, dried, and calcined in air a t 400 "C for 1 h. Reagent. An aqueous solution of hydrogen peroxide (30 wt 5%) was obtained from Mitsubishi Gas Chemical Industry. The concentration was determined by titration with potassium permanganate (0.1 N). Procedure. The catalyst (0.15-1.5 g) was packed in 7-mm Pyrex tubing which was temperature controlled a t 2 "C. The HzOzsolution was fed from a reservoir at a rate of 8 cm3 min-' with a roller pump. The concentration of hydrogen peroxide was determined by titration of the solution at the outlet with potassium permanganate under acidic condition. For determination of the superoxide ion concentration, the outflow was directly led into liquid nitrogen through a nozzle. The ice thus formed was picked up and the amount of the radical was determined by ESR at 77 K. The ESR measurements were carried out with a JEOL X-band spectrometer (PE-1X). The radical content was determined by comparison of the area under the absorption curve for the anion and that for 1,l-diphenyl2-picrylhydrazyl (DPPH) in benzene. The change in Q factor for the 02-and DPPH measurements was calibrated by use of an ESR marker included in the ESR cavity. Results and Discussion ESR Spectrum of 0;. Figure 1 shows a typical ESR spectrum of the solution frozen at the exit of the catalyst The Journal of Physical Chemistry, Vol. 81, No. 13, 1977
1308
Ono et 81.
TABLE I: R a t e Constants for H y d r o g e n Peroxide Decomposition with Various Catalysts a t 275 K
106k,,
1O2k,:
(g of catalyst)-' L min"
(g of catalyst)-' L min-'
103k,, (g of catalyst)-' L min-'
4.0 2.6 2.2
0.2 0.3 0.2 1.8 f. 0.2 0.26 f. 0.02 4.0 k 0.3
AgIALO, Pt/A1,3, Pd/A' 0,
AuIA *03 A403 Charcoal a At a
50i 3 1.3 i k , / k , = (1.9t 0.2) x 4.5 t 0.3 5.0 i 2.7 i: 0.2 4.9 f. 2.1 i: 0.2 4.1 i 8.4 i 0.2 5.8 i
i: i i
0.1 0.3 0.4 0.3 0.3
KOH concentration of 1.0 N.
-
0.5
-
0 K1.0 OH
0
-
Pt / A I 2 0 3
0
El,=
t
3.0 N
( H 2 0 ~ )0.1 ~ 00 M
L
5.0 mT
-e%\-
2.0 N
I r Y
___.,
N
1.0 ( W / F ) x l d 2( g - c a t n m i n l l )
2.0
(b) 1 . 0 h
1
2.001
Flgure 1. ESR spectrum of the 02'radical produced during the decomposition of H202over Ag/AI2O3in 1.0 N KOH aqueous solution.
( a ) 1.0
-I
i
'
v
B .& '
0
3.0 N AgIA1203
(H202)o 0.100
1
(H202)n
A
* 0
I
0
0.5
1.0 1.5 ( W/F)x10-2 ( g - c a t , m i n / l )
o A 0
0
0.141 M 0.310 M 0.203 M 0.103 M
0.100 M 0.194 M 0.296 M
0.40C M
The Journal of Physical Chemistry, Vol. 81, No. 73, 1977
(
2.0
1
1
wi
M
1.0 2.0 F ) x 1 o-2 ( g - c a t m i n i I )
Flgure 3. In (1 - x ) plotted against the contact time ( W I F ) for the H202decomposition over (a) R/AI2O3and (b) Ag/AI2O3at various KOH concentrations at 275 K; the initial concentration of hydrogen peroxide was 0.100 M.
Kinetics of HzOz Decomposition. The conversion (x) of HzOzwas measured as a function of catalyst amount (W) in the column. The feed rate (F)of the solution was kept constant at 8.0 cm3 min-l. Figure 2a,b shows a plot of In (1- x) against WIF for decomposition over Ag/Al2O3and Pt/AlZO3,respectively. In each case, a straight line was obtained and the slope of the line was independent of the initial concentration of HzOz. Thus, the decomposition is the first order with respect to peroxide concentration: -d[HzO*l,/d[W/Fl = kl[H'OZlt (6) where [HzOZ],is the total analytical concentration of hydrogen peroxide. First-order kinetics hold for HzOzdecomposition by all the catalyst used in this study. Figure 3a,b shows that the rate of decomposition both on silver and on platinum is almost independent of the concentration of potassium hydroxide. The values of the rate constant, kl,for the various catalysts are summarized in Table I. 02-Formation on Ag/A1203.The concentration of the superoxide ion at the column exit was calculated from the ESR intensity of the frozen solution. The concentration of the superoxide ion formed by decomposition on Ag/ AlZO3was plotted as a function of the contact time ( W / F ) in Figure 4. The radical concentration increases with the contact time and then, through a maximum, it decreases
1309
Decomposition of Hydrogen Peroxide on Supported Metals
9
( KOH)
1.5
m 0 r
X h
I N
0 u
"
0
0
1.0 2.0 3.0 (WIF)x162(~-catmln/l) AglA1203
1.0 ( WIF)
K O H 1.0 N
Feed Rate 8.0 cm3/mln Figure 4. Concentrationof 02plotted against the contact time ( W / F ) during the H202decomposition for different initial concentrations of H202 at 275 K; the catalyst was Ag/A1,03; the KOH concentration was 1.0 N.
3.0
2,o
xld2 ( 9 - c a t
minil)
Figure 5. Concentration of 0; plotted against the contact time ( W / F ) during the H202decomposition for different KOH concentrationsat 275 K; the catalyst was Ag/AI2O3;the initial concentration of H202was 0.300
M.
with further contact with the catalyst. At the maximum, an 0; yield of as high as 0.62 mM is obtained. The contact time which gives the maximum yield of 0; is independent of the initial concentration of hydrogen peroxide and is 140 (g of catalyst) min L-'. The rate of superoxide formation can be expressed by the following equation:
t
Rate eq 7 was confirmed in the following manner. The initial rate of Oz- formation is given by
O:
d[O;lld[WIFl = ~z[H2021ot (8) where [H20zlotis the initial analytical concentration of hydrogen peroxide. The initial rate of 02- formation determined from the slope of Figure 4 in the low W / F region is proportional to the initial concentration of hydrogen peroxide and the value of k z is determined as 5.0 f 0.3 X L min-' (g of catalyst)-'. Integration of eq 7 using eq 6 gives
- exP(-k 1 KIF)1 (9) The contact time which gives the maximum yield of superoxide is obtained by differentiating eq 9 and is expressed by (W/F)Inax =
In ( k l / k d I ( k l -
123)
(10)
This agrees with the experimental fact that the contact time which gives the maximum yield of superoxide ion is independent of the initial concentration of- hydrogen peroxide. By setting (W/F),,, = 130 (g of catalyst) min L-' and kl = 4.0 f 0.2 X W3(g of catalyst)-' L min-', one obtains the k3 value of 1.3 f 0.1 X L min-' (g of catalyst)-'. Using the determined values of kl,kp, and k3, the solid curves in Figure 4 were drawn. The agreement of the calculated lines with the experimental results is fairly good. Effect of KOH Concentration. The effect of the concentration of potassium hydroxide on 02-formation is illustrated in Figure 5, which shows that the initial rate formation is independent of KOH concentration; of 02the rate constant kz is independent of KOH concentration,
0.100 M
I I
1.0 ( WIF ) x 1 PtIA1203
KOH
2.0 6( g~- c a t m i n i l )
3.0
1.0 N
Flgure 6. Concentration of 02'plotted against the contact time ( W / F ) during the H202 decomposition over R/A1203 at 275 K; the KOH concentration was 1.0 N.
but the maximum 02- concentration increases with increasing KOH concentration. As described earlier, the first-order rate constant kl for HzOz decomposition does not depend on the KOH concentration, then the value of k3, the decay rate constant of O;, should decrease with increasing KOH concentration. From Figure 5, the values of k3 are determined as 13 f 1 X ( of catalyst)-' L min-' ([KOH] = 1.0 N), 6.8 f 0.5 X 10- (g of catalyst)-' L min-' ([KOH] = 2.0 N), 4.5 f 0.2 X (g of catalyst)-' L m i d ([KOH] = 3.0 N). 0; Formation on Pt/Al,O,. As stated earlier, the rate of decomposition of hydrogen peroxide over Pt/Alz03is first order. Figure 6 shows the dependence of the superoxide concentration on the contact time (W/F)in this system. The superoxide concentration is an order of magnitude smaller than that observed in the case of Ag/A1203. The superoxide concentration is proportional to the initial concentration of hydrogen peroxide. Under our experimental conditions, it decreases with increasing contact time (W/F). It was also found that the superoxide concentration increased with increasing KOH concentration. By assuming the same kinetics as in the case of Ag/A120a, the ratio of the rate constants k z and k3 is estimated from the results in Figure 6 and is listed in Table I. Using the determined ratio, kz/k3, and the rate constant for decomposition, kl,the solid curves in Figure 6 were drawn. The agreement between the calculated lines and the experimental points is good.
f
The Journal of Physical Chemistry, Vol. 81, No. 13, 1977
1310
Ono et ai.
I "0
1.0
20
3.0
( w / F )xic2(g-catmin/l ) K O H 1.0 N ( H 2 0 2 ) . 0.100 M Flgure 7. Cornparison of 02concentration plotted against the contact time ( W / F )during the H202decomposition over various catalysts at 275 K; the initial concen?ration of H202was 0.100 M; the KOH concentration was 1.0 N.
Decomposition over Other Catalysts. The decomposition of hydrogen peroxide and the formation of the superoxide ion were studied also on Pd/A1203,Au/A1203, alumina, and charcoal. The decomposition is first order in every case and the rate constants are listed in Table I. The concentrations of superoxide ion were compared together with the cases of Ag/AlZO3 and Pt/A1,03 in Figure 7. Applying the same kinetic equation (eq 7 ) as in the case of Ag/A1203,the kinetic parameters k2 and k3 are estimated and listed in Table I. Mechanism of 02- Formation. Observation of the superoxide ion in the aqueous phase clearly indicates that the decomposition of hydrogen peroxide on heterogeneous catalysts involves free-radical processes at least partly and that the radical formed on the surfaces can desorb into the homogeneous phase as a superoxide ion. Thus, we assume that some of the superoxide ion formed on the surface desorbs without transferring an electron into the metal. €€ one assumes the mechanism proposed by Weiss," eq 1-6, and uses the steady-state approximation for .OH, .HOz, and S+,then the rate of Hz02decomposition, r, is given bylo
x LH2021t (11) where 0 = [SI+ [s+1, rs = ( h 3 s h j s ~ H O ~ ( ~ ) ~ H ~ 0 2 ( a ) / k l s h 4 s ) 1 ' 2 ,
and and KH202(a) are the dissociation constants of the perhydroxyl radical and hydrogen peroxide in the adsorbed states. The dissociation constant of H202in aqueous solution, KH202,is 1.7 X 10-l' M at 20 "C." Under our experimental conditions, KH2O2is much greater than [H+1; >> [Htl. IfKH202(a)>> [H'l and rs
