Environ. Sci. Technol. 1989, 23, 533-540
Catalytic Autoxidation of Chemical.Contaminants by Hybrid Complexes of Cobalt( I I ) Phthalocyanine Andrew P. Hong, Scott D. Boyce,t and Michael R. Hoffmann”
Environmental Engineering Science, W. M. Keck Laboratories, California Institute of Technology, Pasadena, California 9 1 125 Cobalt(I1) tetraaminophthalocyanine (Co”TAP) and cobalt(I1)tetrasulfophthalocyanine (ConTSP) were linked to the surfaces of a chemically modified silica gel, a functionalized Ti02, and an amino-substituted macroporous copolymer. The hybrid catalysts were found to have catalytic activity for the aqueous-phase oxidation of hydrazine, hydrogen sulfide, sulfur dioxide, and thiols by molecular oxygen. In general, the hybrid catalysts exhibited longer catalytic lifetimes than their homogeneous analogues. In most cases, attachment of the catalytic complexes to the solid supports resulted in reduced catalytic activity; however, in selected cases, the hybrid catalysts were more active than their homogeneous analogues. The effective diffusivity for S O:- within macroporous polystyrene/divinylbenzene beads was found experimentally to be 2.6 X lo4 cm2 s-’. The observed catalytic activity has been interpreted within the framework of a kinetic model with mass-transport limitations.
Introduction Metal-catalyzed autoxidation may provide an efficient and economical method for pollution control. Autoxidation in this context is defined as the oxidation of a reductant by molecular oxygen (I). In general, reactions of the triplet ground state of O2with singlet spin-state reductants proceed very slowly because they involve changes in spin multiplicity and a considerable degree of bond alteration of deformation in the formation of products. The oxidation of many inorganic and organic substrates by oxygen is often accelerated in the presence of multivalent first-row transition-metal cations, Mn+,such as Mn2+,Fe2+/Fe3+, Co2+/Co3+,Ni2+,and Cu2+(1,2). Metal-phthalocyanine complexes have been shown to be effective homogeneous catalysts for the autoxidation of aldehydes (3-5))aromatic hydrocarbons (6),ascorbic acid (7),hydrazine (81,hydroxylamine (9),hydrogen sulfide (lo),mercaptans (11,12), phenols (13),and sulfur dioxide (14)in organic solvents and in water. The catalytic activity of the metal-phthalocyanine complexes depends on the nature of the central metal. For the first-row transition metals, catalytic activity with respect to autoxidation has the following general trend: Cor’ > Fen > Mnn > Cu” N Ni”. This trend reflects the relative capacity of the corresponding metal-phthalocyanine complex to reversibly bind molecular oxygen (15).The reactivity of cobalt(I1) 4,4~,4”,4”’-tetrasulfophthalocyanine (Co”TSP) in aqueous solution has been compared with that of oxidases and dioxygenases (3,8,12,14,16,17). A major disadvantage associated with the application of homogeneous catalysts for pollution control involves the problem of separating the catalyst from the reactants and products. This drawback may be overcome by attachment of the homogeneous catalyst to the surface of an inert solid, which can be readily removed from the reaction mixture or can be used in a fixed-bed reador. Maas and co-workers (18)and Schutten et al. (19)have previously studied the Present address: The PQ Corp., 280 Cedar Grove Rd., Philadelphia, PA 19444. 0013-936X/89/0923-0533$01.50/0
kinetics of thiol oxidation to disulfides in the presence of cobalt(I1) phthalocyanines attached to polyacrylamide, polyvinylamine, and a polystyrene/ divinylbenzene copolymer. They reported that the rate of thiol autoxidation was accelerated in the presence of the hybrid catalysts as compared to the homogeneous control. With this in mind, we synthesized several new cobalt(I1) phthalocyanine hybrid catalysts using silica gel, Ti02,and polystyrene/divinylbenzene copolymers as solid supports. In addition, we determined their catalytic activity with respect to the autoxidation of several reduced sulfur compounds. The observed catalytic activity has been interpreted within the framework of a kinetic model with mass-transport limitations.
Experimental Section Synthesis of Cobalt(11)-PhthalocyanineComplexes. The tetrasodium salt of cobalt(I1) 4,4’,4’’,4‘‘’-tetrasulfophthalocyanine (CoIITSP) was synthesized from sodium 4-sulfophthalate according to the method of Weber and Busch (21). Cobalt(I1) 4,4’,4’’,4”’-tetraaminophthalocyanine (CoIITAP) was prepared as follows: a mixture of sodium 4-aminophthalate (19 mmol) (ICN Pharmaceuticals), ammonium chloride (12 mmol), ammonium molybdate (6.6 mmol), urea (0.13 mol), and cobalt(I1) sulfate 7-hydrate (6.6 mmol) was ground until it appeared to be homogeneous, and then the mixture was heated under refluxing conditions at 170 “C in nitrobenzene (20 mL) for 6 h. The crude product was washed with methanol and then heated to boiling in 0.5 N HC1 (400 mL). After filtration, the solid was dissolved in hot (70 “C) dimethyl sulfoxide (DMSO; 250 mL) to remove insoluble impurities. The product precipitated upon addition of H20 (500 mL) and was collected by centrifugation. Further purification involved repetitive washings with boiling H20 and centrifugation. Following initial treatment with absolute ethanol, pure Co”TAP was obtained by heating the solid in ethanol under reflux for 5 h. The structure of each complex was verified by elemental analysis (Galbraith Laboratories) and by comparison of UV/vis (HP 8450A) and ‘H NMR (Varian Model EM390) spectra with published reference spectra (21,22). Preparation of the Silica Gel Support. The synthesis of the catalyst support involved treatment of silica gel with a silylating reagent (23,24). In the preparation of 1, a suspension of silica gel (22 g; Fisher; specific area, 330 m2 g-’) and (3-chloropropyl)trimethoxysilane(7.5 g; Aldrich, redistilled) in xylene (150 mL) was heated under reflux for 8 h. Addition of imidazole (3.9 g) to 1 resulted in the formation of 2. After filtration, the functionalized silica gel was washed thoroughly with acetone and dried under vacuum; 2 was determined to have 1.7 X mol of N g-l as a result of the synthesis. Attachment of ConTAP and ConTSP to Silica Gel. Two different methods were utilized to anchor the Co(11)-phthalocyanine complexes to the modified silica gels as shown in Figure 1. In the first method, attachment of the catalyst to the solid support was achieved linking the surface ligand to the modified silica gel to the peripheral
0 1989 American Chemical Society
Environ. Sci. Technol., Vol. 23, No. 5, 1989
533
SILICA GEL SCHEME 1:
1 r=1
I
I
TAP
(I)
DMSO
SCHEME 2 :
SILICA GEL
I
0 -Si-(CH2)3-N
A
I
N
Co'TAP
DMSO
OCH
I
-0-
6CH3
POLYSTYRENE/DIVI NYLBENZENE
2HN+ J P (
S0,Na
Figure 1. Schematic diagram Wustrathg the syntheses of hybrid cobatt(II)-phthalocyanine complexes with silica gel supports 1 and 2 and cross-linked polystyrene/divlnylbenzene copolymers.
amino group of the Co"TAP. A mixture of the modified silica gel 1 (10 g) and Co"TAP (6.7 g) in DMSO (30 mL) was heated at 80 OC for 4 h under constant stirring. The hybrid cobalt-phthalocyanine complex 3 was isolated by filtration and washed successively with warm DMSO and 0.1 N NaOH in order to remove excess ConTAP and HC1. 534
Environ. Sci. Technol., Vol. 23, No. 5, 1989
Purification of the reaction product 3 was accomplished through the extraction of impurities with H20 by using a Soxhlet apparatus. The solid was dried in an oven at 100 "C. The purified hybrid catalyst 3 was determined to have 2.7 X mol of Co/g of material and thus results in a surface coverage of approximately 150 Co"TAP mole-
cules/pm2 (Le., -0.06% of the silica gel surface is covered with the catal ic complex, assuming a per molecule coverage of 400 ). The second synthetic method involved the direct coordination of the surface imidazole functionality to the metal center of the phthalocyanine complex. Synthesis was achieved as follows: an aqueous solution of Co"TSP (7.8 X lod M) was added to a column packed with the modified silica gel 2 and eluted dropwise. A similar procedure was followed for Co"TAP with DMSO as the solvent. The reaction products, 4 and 5, were washed several times with DMSO and H20, respectively. The initial washing was followed by an acetone wash. The products were collected by filtration. Analysis of the eluents by UV/vis spectrophotometry yielded a cobalt concentration of 1.1X lo+ mol of Co g-' for both hybrid catalysts 4 and 5. In these cases, each hybrid catalyst had approximately 6 molecules of active catalyst/pm2 or a surface coverage of 0.0024%. Attachment of Co"TSP to Polystyrene/Divinylbenzene. CoIITSP was coupled to a macroporous, amino-substituted polystyrene/divinylbenzene matrix, 6, by direct complexation of the NH2 functionality of the modified surface to the metal center of the CoIITSP. The preparation of 6 involved the addition of Amberlite IRA-93 resin (10 g; Rohm and Haas; specific area, 35 m2 g-l) to 500 mL of an aqueous Co"TSP solution (2.5 X M). The reaction mixture was agitated continuously for 1h at room temperature. Stirring was accomplished with a rotary evaporator in order to prevent mechanical damage to the particle structure. The product was isolated by filtration, and the solid was washed several times with dionized water. Spectrophotometric analysis of the filtrate showed an effective cobalt concentration of 4.8 X lod mol of Co/g of solid support. Based on this analysis approximately 9% of the porous particle surface is covered by the active catalyst. Attachment of Co"TSP to Ti02. The synthesis of Ti02-Co"TSP (7) has been described in detail by Hong et al. (25). Green cobalt(11)tetrasulfonylphthalocyanine (Co"(S02C1),P) was obtained by treatment of CoIITSP with PC&, while Ti02 (Degussa, P25 anatase) was functionalized with (3-aminopropy1)triethoxysilanein xylene for 1 2 h. A sulfonamide linkage between Co"(S02C1),P and the silane groups on the Ti02 surface resulted in the stable attachment of the catalyst to Ti02. Ti02-Co"(S02Cl),P was hydrolyzed in 2.5 M H2S04to yield Ti02CoIITSP. Reagents. All reagents were of analytical grade. Buffers were prepared with sodium phosphate (Mallinckrodt), sodium borate (Mallinckrodt), and sodium hydroxide (J. T. Baker). Constant ionic strength ( p ) was maintained with sodium perchlorate (G. F. Smith). The water used to prepare the buffer and reagent solutions was obtained from a Milli-Q water purification system (Millipore) and had a resistivity of 18 MQ cm. Solutions of reagents and buffers were prepared with deoxygenated water obtained either by purging with N2 or by Schlenk line vacuum techniques as described by Conklin and Hoffmann (26). Kinetic Procedures. Kinetic measurements were made by following the time rate of change of either the substrate concentration, with a continuous-flow spectrophotometric cell (HP 8540 UV-vis spectrophotometer), or the dissolved oxygen concentration, with a DO probe (Orion Model 9708-00) linked to an ion analyzer (Orion Model 901). Data were analyzed on-line with an IBM-XT computer. The pH of the reaction mixture was monitored simultaneously with a combination electrode (Orion 91-62, Ag/AgCl reference).
$
The reactions were conducted in a water-jacketed, glass and Teflon reactor with a total volume of 2.0 L. The design and operation of the batch reactor system have been described previously (10, 24). A constant temperature of 25 f 0.1 "C was maintained during the course of the reaction by a Haake Model FK-2 water circulation system and temperature controller. Reactions were initiated by adding a known volume of substrate stock solution to a buffer/catalyst mixture. Dissolved oxygen levels were established by dispersing N2 and O2 gas mixtures into the buffered suspension before the addition of the substrate. EDTA was added as a sequestering reagent to reduce the catalytic effect of homogeneous trace-metal contaminants (20). Mannitol and sodium cyanide were utilized as free-radical scavengers and complexation inhibitors, respectively. Sorption/desorption experiments were run to evaluate the diffusivity of S032-in the porous catalyst supports. Liquid-phase S(1V) concentrations were followed by periodically withdrawing aliquots and removing the solids by filtration; the resulting solution was measured for S032spectrophotometrically.
Results The hybrid catalysts, 3-6, were tested for catalytic activity with respect to the autoxidation of hydrogen sulfide, sulfur dioxide, 2-mercaptoethanol (R = CH2CH20H), cysteine (R = CH2CH(NH2)C02H),and hydrazine by measuring either the rate of O2 consumption or the rate of substrate loss according to the following stoichiometries: HS- + 2 0 2 H++ S042(1)
+ + + + - +
2S0322RSH
N2H4
0 2
0 2
O2
2S042-
RSSR N2
H202
2H20
(2)
(3) (4)
where the R groups in mercaptans have been defined above. In order to compare the performance of the solid-supported catalysts with their homogeneous counterparts, the corresponding homogeneous reactions were also carried out. The results for these experiments are summarized in Table I. The most efficient hybrid catalyst appeared to be the CoIITAP derivative 4, which was formed upon the complexation of the imidazole group on the functionalized silica gel directly to the Co(I1) center on Co"TAP. The next most efficient catalyst appeared to be the ConTAP derivative on silica gel that was bound through the peripheral amino substituent of the phthalocyanine ligand. The Co"TSP derivatives 5 and 6, bound to their respective surfaces through the central metal, appeared to have lower catalytic activity. The reactivity of the latter hybrid species did not seem to depend strongly on the nature of the supporting matrix, although the slightly higher activity of 6 compared to 5 may be due to the fact that the IRA-93 support of 6 has a typical pore dimension of 1000 A whereas the silica gel support of 5 has a substantially smaller pore size. With the exception of S(IV), the reactivity of the metalphthalocyanine complex was lower when supported on a solid substrate. Since many of the active catalytic centers are located within the pores of the support matrix, diffusion of the substrates and O2 through pores may be rate limiting. In terms of substrate specificity, HS03- appears to be most efficiently catalyzed in the hybrid catalytic systems followed by N2H2,HOC2H4SH,HS-, and H02CC(NH2)HCH2SH,in that order. In the case of the catalytic autoxidation of S(IV), [02] vs time profiles for reaction mixtures containing 1 mM Envlron. Scl. Technol., Vol. 23, No. 5, 1989 535
Table I. Summary of Kinetic Results for the Autoxidation of Selected Substrates in the Presence of Homogeneous and Heterogeneous Cobalt(11)-Phthalocyanine Complexesa uo,
substrate
pH
HS03HSO
