Langmuir 1995,11, 519-526
519
Iron Oxide-Mediated Degradation, Photodegradation, and Biodegradation of Aminophenols C. Pulgarin Institute de Genie de I'Environement, Laboratoire de Genie Biologique, Ecole Polytechnique, Federale de Lausanne, CH-1015 Lausanne, Switzerland
J. Kiwi* Institute de Chimie Physique, Ecole Polytechnique Federale de Lausanne, CH-1015 Lausanne, Switzerland Received July 5, 1994. I n Final Form: November 9, 1994@ The results of the photocatalytic degradation and the influence of photocatalytic pretreatment on the biodegradability of 2-aminophenol using iron oxides are reported. Various reaction parameters like substrate concentration,a-FezO3 content, amount of oxygen present, type of iron oxide, pH of the solution, temperature, and contribution of the dark reaction were investigated. By the use of several experimental techniques like 'H-NMR studies, C02 evolution,total organiccarbon spectrophotometry,and high-pressure liquid chromatography,the more relevant features ofthe degradationmechanism observed in aminophenols were established. a-Fe203 appears to react only with aminophenols having electron-donatingcharacter through the formation of a surface complex. No similar degradationwas observedwhen phenol,nitrophenol, or chlorophenolwas used as a substrate. The reaction proceeds in the dark and under light and is favored by the amount of oxygen present. The rates of photooxidation were found to vary by about 2 orders of magnitude with different iron oxides. Photocatalysis mediated by a-Fez03 used as pretreatment step enhanced the biodegradability of 2-aminophenol. A favorable BODdCOD ratio of 0.40 is achieved when photocatalytic pretreatment is applied as compared to a value of 0.23 found for non-pretreated solutions.
Introduction Phenolic wastes result from many industrial processes. They are residuals from phenolic resin manufacturing and petroleum refining. Aminophenols are specific intermediates in the manufacturing of dyes and in the dying of furs.l The level of phenolic compounds in municipal wastes has become increasingly serious in recent years. It has been observed that the treatment of phenols with HClO in treatment plants results in the formation of persistent and toxic chlorophenols.2 More recently, phenolic compounds have been photodegraded by new methods via Ti021r3using high-energy photons. The objective of this work was to explore the possibility of using iron oxides in the dark and under light for the degradation of phenols and to ensure that total mineralization takes place. I t was also our interest to examine the biodegradability for some phenolic compounds in view of the conflicting literature values4 reported for the biological degradation of this compound. a-Fe203, the most common iron oxide, has been used extensively during this work. It has a narrow band gap (2.2 eV) and shows better photoelectrochemical response a t longer wavelengths in the visible region than TiOz. It has a flat band potential ofO.1V a t pH 0 (vs NHE).5 It is thermodynamically stable toward photoanodic decomposition. a-FezO3 shows very Abstract published inAdvance ACSAbstracts, January 1,1995. (1)(a) Ollis, D. F.; Al-Ekabi, H. (Eds) Photocatalytic Purification and Treatment of Water and Air; Elsevier: Amsterdam, 1993. (b) Matthews, W.R. in Photochemical Conversion and Storage ofEnergy; Pelizzetti, E.; Schavello, M. (Eds.) Kluwer: Dordrecht, 1991;pp 427449. (2)Pitter, P.; Chudoba, V. Biodegradability oforganic Substances in the Aquatic Enuironment; CRC Press: Boca, Raton, FL 1990. (3)(a)Legrini, 0.;Oliveiros, E.; Braun, M. A. Chem. Reu. 1993,93, 671. (b) Augugliaro, V.;Palmisano, M.; Scbiavello, M.; Selafani, A.; Marchese, L.; Martra, G.; Miano, F.App1. Catal. 1991,69,323.(c) Fox, A. M. Acc. Chem. Res. 1983,16,314.(d) Helz, G.; Zepp, R.; Crosby, D. Aquatic and Surface Chemistry; Lewis: Boca Raton, FL 1994. (4)Howard, P. Handbook of Environmental Degradation Rates; Lewis: Washington, DC, 1989.
low electron mobility cm2Ns) although it is unstable in acidic solution and its efficiencies are low. Iron oxides have been investigated in photocatalytic processes. a-Fe203 has been found to be stable in the transformation of sulfites, oxalates, and peroxosulfates.6 Iron oxides are widespread in nature and are of significance in the processes taking place in ecosystems. They regulate the concentration and distribution of plant nutrients and heavy metals in the natural cycle.'Z2 a-FezO3 is the most abundant oxide, but FeOOH (geothite) and y-FezO3 are the end products of the corrosion of iron. In this paper we plan to show that a-FezO3 reacts with phenols containing N-donor groups and leads to the degradation of this compound. A weak interaction between the lone electron pair of the NHZ group and the higher valence states of Fe(+3,+4) will be shown to lead to complex formation that seems to be active in the degradation of the substrate. Oxygen has been used as oxidant although it is less thermodynamically favored than H202,5,8 but in this way the oxide corrosion due to the addition of H2Oz-induced radicals is avoided.
Experimental Section Photolysis experiments were carried out with dispersions placed in 60 mL Pyrex glass reaction vessels. Illumination with A > 290 nm was carried out by means of a Hanau Suntest lamp having 2.1kV A electricalpower and 76.5mW/cm2(AM1).Dark control experimentswere also carried out in thermostatedbaths
@
(5)(a) Kennedy, H.J.;Frese, W. K. Electrochem. Soc. 1978,25,723. (b) Hardee, L. K.; Bard, J. A. J. Electrochem. SOC.1977,124,215.(c) Gratzel, M.; Kiwi, J.; Morrison, C.; Davidson, S.; Tseung, C. J . Chem. Soc., Faraday Trans. 1 1985,81,1883. (6)(a)Leland, K. J.; Bard, J. A. J.Phys. Chem. 1987,91,5076. (b) Herrmann, J.;Mozzanega, N. M.; Pichat, P. J.Photochem. 1984,102, 1211. (c) Maruthamuthu, P.; Gurunathan, K.; Subramaniam, E.; Ashokkumar, M. Bull. Chem. Soc. Jpn. 1991,64,1933. (7)(a) Balzani, V. Tetrahedron 1992,48, 10443. (b) Balzani, V.; Campagna, S.; Denti, G.; Serroni, S. Supramolecular Photochemistry. Antenna effect inpolynuclear metal complexes. NATOASISer. C . 1992, 376,233. (c) Lehn, J. M. Angew. Chem. Int. Ed. Engl. 1990,29,1304. (8) Kormann, C.; Bahnemann,W. D.; Hoffmann, R. M. J.Photochem. Photobiol. A 1989,48, 161.
0743-7463/95/2411-0519$09.00/0 0 1995 American Chemical Society
520 Langmuir, Vol. 11, No. 2, 1995 at the temperatures indicated. Prior to the start of the experiments the suspension was stirred for a V2 h to normalize the absorption taking place before illuminati~n.~ Experiments were carried out under aerobic conditions. Spectrophotometric analysis on the supernatant of centrifugated samples was performed via a Hewlett-Packard 386/20 N diode array. The progress of the photoinduced degradation of the substrate was monitored by measuring the total organic carbon (TOC; Shimadzu-500,equippedwith anAS1502automatic sample injector). CO2 was followed by gas chromatography (GC) by means of a Poropak QS column. Chemical oxygen demand (COD) was measured via a Hach DlU2000opticalunit. Proton NMR analysis (BrukerACP-200)was meant to identifythe intermediatesformed during the photocatalyzed degradation. High-pressure liquid chroamtography (HPLC)was carried out via a Varian 5500unit. The gradient was regulated with ammonia acetate buffer (100 mM/L) and methanol was used with an inverse phase column Spherisorb 5 ODS-2. The signals were observed at A = 281 nm. Samples were decanted at regular intervals while being purged with argon and immediately centrifuged at 5000 rpm for 5 min, after which the supernatant was removed and tested for FelI1by iodometric analysis and complexation with thiocyanate. lo The optical densities at 351 and 476 nm were determined (with substraction of blank) using a Hewlett-Packard diodearray. The oxidation potentials of the phenols were determined in a cyclic voltameter Autolab-20. The peroxide in solution was determined quantitatively by the Is- spectral absorption method.5c The peak for the iodineiodide complexwas assessed in each case showing an absorption at 354 nm of E = 26000 M-l cm-l. Using a Tecator titrator the nitrate found during the degradation was determined via the cadmium reduction method. By addition of sulphanilamide, a diazo dye is formed and the color intensity was measured spectrophotometrically at 540 nm. Ammonia was determined in water by flow injection analysis when an aqueous sample was injected into a carrier stream and mixed with a sodium hydroxide-boric acid mixture. The resulting color change was measured via a Tecator instrument at 590 nm. The geothite used was a product of Bayer AG with a BET surface area of 14.7 m2/gconsisting of needle shape units 0.06 pm long. a-FezO3with 150m2/gsurface area was obtained from BASF, Ludwigshaven, Germany. It was prepared by oxidation of Fe(C0)5 at T > 500 "C. It did not form big agglomerates as seen by TEM and was isometric in shape. Electron microscopy (TEM)was carried out with a Phillips 300s instrument. The XPS measurements were performed using a Leybold surface analyzer equipped with a separate gas reaction cell attached to the spectrometer chamber. Thebase pressure of the spectrometer mbar and during the measurements was 8 x was 2 x mbar. The XPS spectra were recorded using a twin anode X-ray gun (Mg KdAl Ka) at a power of 200 and 250 W, respectively, and a hemispherical energy analyzer(LeyboldEA11/100). X-ray powder diffraction patterns were recorded from 0" to 90" (20) using a Hijaku diffractometer. Biochemical oxygen demand (BOD)determination was carried out via a BSB controller(WTW). This unit was thermostated at 20 "C and urban waste water after primary decantation was used as inoculum. (9) Kiwi, J.; Pulgarin, C.; Peringer, P.; Gratzel, M. Appl. Catal. Environmental 1993,3,85. (b) Kiwi, J. J . Environ. Toxic. Chem. 1994, 13,1569. (c) Okamoto, K.; Tanaka, H.; Itaya, A. Bull. Chem. SOC.Jpn. 1985,58,2023. (10) Kolthoff, I . M.; Sandell, E. B. Quantitative Chemical Analysis; Macmillan Co.: Toronto, 1994. (11) Eisenhauer, R. H. J . Water Proc. Chem. Fed. 1964,36, 1116. (12) (a) Ollis, D.; Hsiao, C.; Budiman, L.; Lee, C. J . Catal.'1984,88, 89. (b) Turchi, S.; Ollis, D. J . Catal. 1989,119, 480. (13) (a)Knakmuss, H-J. InMicrobial Degradation ofXenobiotics and Recalcitrant Compounds; Lessinger,M. A., Nuegch, J., Eds.; Academic Press: New York, 1981,p 189. (b) Okamoto,K.;Yamamoto,Y.;Tanaka, H.; Itaya, A. Bull. Chem. Soc. Jpn. 1985,58, 2015. (14) Kiwi, J.; Pulgarin, C.; Peringer, P.; Gratzel, M. New. J . Chem. 1993, 17, 487 (and references therein). (15) (a) Lee, K. S.; Fox, B. G.; Munck, E. J . Am. Chem. SOC.1993, 115,6450. (b) Orgaz, F.; Rawson, H. J . Non. Cryst. Sol. 1986,82,378. (16) (a)Meites, L. Handbook ofAnalytica1 Chemistry; McGraw Hill, New York, 1963. (b)Bard,J. A. Encyclopedia OfElectrochemistry;Marcel Dekker: New York, 1973. (17) Hansch, C.; Leo, A.; Taft, W. R. Chem. Rev. 1991,91,165.
Pulgarin and Kiwi
0
10
20
30
time (hours)
Figure 1. Photodegradation measured by total organic carbon as a function of time under different conditions for 2-aminoM)in the presence of 2.5 g/L a-FeeO3. phenol (5.6 x
Results and Discussion A. Kinetics of 2-hinophenol Degradation. Blank experiments using 2-aminophenolin the dark and under light in air showed 4 and 7% decrease in TOC (mg of C/L) over 24 h, respectively. The results for the characteristic a-Fe203-catalyzed decomposition of 2-aminophenol are shown in Figure 1. The concentration of 2-aminophenol used was 5.6 x M having an initial TOC of 400 mg of CL. This value was found after a long series of preliminary runs indicating good reproducibility a t this concentration. The amount of a-Fe203 used was 2.5 g/L with an area of 375 m2/L. Gases were bubbled a t the rate of 20 m u m i n for the time of the experiment. With Ar purging in the dark a small decrease ( k g during the degradation in the mentioned scheme.
02
- - -
B. Mechanismof PhenolsDegradationby a-FezOa. t
IYNH',
no
\
P enamine
lb COz
+
H 2 0 + NH,' +NO;
(9)
The proposed scheme is supported by TOC or dissolved organic carbon measurements (DOC) in Figure 1and by the HPLC reporting the disappearance of the initial substrate in Figure 2. COZ evolution was observed from the beginning and went on during the whole process. XPS studies as well as titration will report later ammonia and nitrates as degradation products. The reaction steps 5-9 are not balanced stoichiometrically. They are intended to show that the degradation of the aromatic ring and cyclic intermediates (as shown by NMR; see Figure 5) is a much faster process than the mineralization of the
Amino compounds form complexes with Fe species when higher oxidation states (+3, +4)15 are involved. a-Fez03 can easily accept electrons from amino compounds as shown by the degradation of 2-aminophenol in Figures 1-5. Light or dark experiments for a-Fe203carried out on 2-chlorophenol, 2-nitrophenol, and phenol did not show any meaningful degradation ( ~ 5 %for ) these compounds. They lacked the electron-donor capacity of 2-aminophenol. That the electron-donating capacity is an important parameter controlling the reaction is shown in Figure 6. The oxidation potential of 2-aminophenol is measured vs Ag/AgCl in saturated KC1 solution (The Ag/AgCl couple lies a t 0.199 VNHE). Phosphate buffer has been used a t pH 7. In the voltammogram shown in Figure 6, curve A indicates the formation and reduction of platinum oxide. The values for the oxidation potentials obtained for 2-aminophenol (trace D), phenol (trace C), 2-nitrophenol (trace B), and 2-chlorophenol were 0.296, 0.669, 0.949, and '1.2 V vs NHE. Figure 6 does not show the peak for 2-chlorophenol since this value is above 1.2 V. These measurements were necessary since the half-wave oxidation potentials found in the literature for these compounds were measured under different experimental conditions and could not be compared.16 The formation of a supramolecular species would be possible due to the low oxidation potential of 2-aminophenol in addition to the ortho effect of the NH2 as a substituent group in the ring. The Hammett constant reported for 2-chlorophenol, 2-nitrophenol, phenol, and 2-aminophenol17are 0.23,0.78, -0.37, and -0.66. The progressive change in the values
a-Fe203-MediatedDegradation of Aminophenols
Langmuir, Vol. 11, No. 2, 1995 523
a
..
-,
.... . ..
..
.-
Figure 6. Cyclic voltametry showing (A) R,(B) nitrophenol, (C) phenol, and (D)aminophenol voltammograms. For other details, see the text.
absorbs light up to 570 nm. Oxygen has also been observed to play a major role in the observed degradation of the substrate. This effect is readily understood in terms of the reaction
:NH,-R b
+ 0,
-
(:NH,- -0,)light .NH,+
+ 0,-
(io)
Under light the excited intermediate in eq 10 would be a stronger oxidant than the corresponding ground state, leading to a more effective degradation of 2-aminophenol.'"
D. Overall Reaction Observedin the Degradation of 2-Aminophenol. The a-FezOs used (see Experimental Section)has a surface area of 150 m2/g. This large surface area allows for the adsorption of meaningful amounts of substrates a s in our case. For particles like the one used (0.06 pm) there is enough potential gradient to separate the h++, from the e-cb produced5
The e-cb cannot react with 0 2 bound to surface since the value E -0.15V (pH 2) has been founds for the reduction of Oz/HOz*Or E -0.N (pH 12) for the reduction in alkaline media.18 Since the does not react with oxygen it may lead to dissolution of a-FezOs:
but this reaction is known to be short circuited by the catalyst via reactions of the holes, hfvb:
Figure 5. 'H-NMR spectra of 2-aminophenol at irradiation times of (a) 0 h, (b) 1h, and (c) 4 h. The resolution has been changed in part c to show in more detail the methylenicprotons.
for electron-donating capacity in the list lends further support to our claim that the electron transfer role from 2-aminophenol to a-FezO3is important in the degradation of the pollutant. In separate experiments 3-aminophenol degraded a t -60% of the rate of 2-aminophenol, indicating the importance of the position of the substituent in relation to the electron density in the ring during the degradation.
C. Possible Supramolecular Chemistry Taking Place during Degradation. a-FezO3 and 2-aminopheno1 form a complex involving some supramolecular character, that is, weak interaction between the electrons of both components. This complex would involve the donation of the electron pair of :NHz to Fe(II1). Light in the visible region is able to excite this species since a-FezO3
Our experimental results show that Fe3+is produced. The above mentioned mechanism is substantiated by the fact that we did not observe Fez+via o-phenanthroline. The Fe3+ produced would recycle in a Fenton-like way:
-.
+
D h+vb D+ D = 2-aminophenol Fe3+
+ H,O,
light
Fe"
+ HO, + H+
(14)
(15)
subsequently being the recycled iron used in eq 15. If the photoproduced holes in eq 13would effectively react with surface OH- and produce OH radicals, we would also observe degradation of phenol, nitrophenol, and chlorophenol, which is not the case. The photodegradation of (19)(a)Sawyer, T. D.; Valentine, S. J.ACC.Chem. Res. 1981,14,393.
(b) Airey, P. L.; Sutton, H. C. J. Chem. Soc., Faraday Tram. 1 1976, 2 , 2542.
524 Langmuir, Vol. 11, No. 2, 1995
Pulgarin and Kiwi
degradation time of 2-aminophenol under light. Adding lop2M H202 in the dark or under light increases the Fe(II1) ion concentration in solution in both cases to 3 and 5 M, respectively. Peroxides present in solutions of iron oxides have been reported to produce Fe3+ions.lg
3
-
n
EE
E. Effect of Catalyst Type and Surface Area on the Degradation. Figure 7 shows the effect of catalyst -0 concentration on the disappearance of 2-aminophenol. Trace a decreases =lo% when no catalyst and purging c with oxygen under light are used. Traces b, c, d, and e .s n show the effect of 10, 60, 100, and 200 mg/80 mL, 0 .-c respectively, of Fe2O3 on the degradation of 2-aminoE ' phenol under light and oxygen purging. This figure shows the favorable effect of the increase of catalyst concentration on the rate of 2-aminophenol degradation. This rate follows an apparent first-order kinetics because the value will depend on the type of iron oxide used, the nature of the gas atmosphere present, etc.14 The penetration of 0 0 10 20 30 360 nm light into 01-Fe203~O is attenuated by 98%in 1 pm. We used higher concentrations of a-Fe203 during this work time (hours) (2.5 g/L) since the average particle size20 of Fe2O3 was Figure 7. Effect of a-Fe2Os concentration on the photodeg0.06 pm, being made up of 0.01 pm particle size crystals. M): (a)no a-Fe203, and radation of 2-aminophenol(2.8 x The incident photons traverse many particles before they (b) 0.125 g/L, (e) 0.75 g/L, (d) 1.25 g/L, and (e)2.5 g/L a-Fe203. are fully attenuated due to the transpartent character of The degradation is followed by HPLC. the small particles used. Therefore a high particle concentration was necessary. We have worked here at 2-aminophenol therefore involves photochemistry of a relatively low 2-aminophenol TOC (mg of C L ) values of surface complex and not a semiconductor-assisted pho400 mg of C/L since 0 2 is rapidly depleted in the region todegradation. Thereforethe present study does not claim of light penetration. that a-Fe203 is useful as a photocatalyst. The degradation observed on 2-aminophenol(2.8x loW3 If 2-aminophenol (D) is adsorbed strongly on the iron M) after 4 h was studied in a stream of oxygen in the light oxide surface, then reaction 14 will proceed favorably and when (a) Fluka (1.9 m2/g);(b) Geothite Bayer (14 m2/g), lead to the observed degradation in spite of this reaction and (c) BASF a-FezO3 (150 m2/g) are used. In all cases being in competition with reaction 13. The generated hole 2.5 g/L of catalyst has been used. The amount of in reaction 11 has a thermodynamic potential of 2.5 V. degradation of 2-aminophenol in relation to their BET But the photooxidation could also go through the 3.3 areas after 4 h was 10, 20, and 95% respectively. V bandgap corresponding to a LMCT (Fe(II1)-O(I1)) F. Biodegradation of 2-Aminophenolbefore and transition.lsC In both cases there is sufficient themoafter Photochemical Treatment. The fate of organic dynamic potential to oxidize 2-aminophenol and more pollutants in the environment depends on their suscepinterestingly also phenol, nitrophenol, and chlorophenol. tibility to biodegradation. No biodegradation rates were Unpublished resultslsdfrom work under way via an FTIR found for this compound in ground and waste water technique show strong adsorption and electronic polareffluent^.^ The biochemical oxygen demand (BOD) of ization between 2-aminophenol and a-Fe203. These two 2-aminophenol(5.6 x M) is shown in Figure 8. The last effects seem to be very weak, as reflected by the FTIR aim was to quantify the oxygen utilized during the results for iron oxide interaction with phenol, nitrophenol, degradation of p-NTS by urban waste water bacteria. and chlorophenol. The intervention of iron oxides on Three bottles were taken for each sample containing degradation processes involving phenols will then depend standard N,P nutrients required during the biological on the possible formation of complexes on the iron oxide process of the Vidy (VD, Switzerland) urban biological surface in which the adsorption plays a determinant role. waste water treatment plant. This solution was further The oxidation with 0 2 has been shown to be important 24 h, filtered through cotton, and used as in the mineralization of many environmetalp o l l ~ t a n t s , ~ ~ ~ Jdecanted ~ inoculum. The oxygen consumed was measured via the and in our case: instrument indicated in the Experimental Section. Blank BOD reference runs are carried out with bacteria only. 4C,H,NO 5g/,0, -24C0, 12H,O The forrnula used to calculate the BOD reported in Figure NH,' NO,N, (16) 8 is 2
W
Q)
(P
(v
+
+
+
+
+
The photodegradation yield observed when irradiation is carried out in the presence of oxygen and a-Fe203 was about 0.2 kW h/ppm C per L. Ammonia and nitrate have been determined according to the Experimental Section. In the presence of the catalyst, light, and oxygen, the substrate produced 30% of the stoichiometric amount of ammonia,4% of the nitrate, and about 10%of the nitrogen. The stoichiometric amount refers to the total amount of product if this is the only N-containing product observed. In the dark about M of Fe3+is present up to 12 h, the time needed for total degradation of 2-aminophenol. Under light irradiation the concentration of Fe3+increases slightly or about 50% up to 4 h. This is the total
BOD = [BOD measured - 0.01 {percentage inoculum x
BOD inoculum}]/[{100 - percentage inoculum} x 0.011 The percentage of inoculum used was 10%. The solutions used before photochemical treatment contained 2-aminophenol (540mg/L) or the equivalent of 400 mg of C L . As seen in Figure 8a the degradation decrease in BOD was not significant. Figure 8b presents the evolution for the BOD values for the same solution but pretreated 2.5 (20) (a) Ostertag, W.; Defazet 1979,434.(b) Smith, R.A. Semiconductors, Cambridge University Press: Cambridge, 1993. ( c )Matijevic, E.Langmuir 1994,10,8.
a-Fe203-MediatedDegradation of Aminophenols ""
1
Langmuir, Vol. 11, No. 2, 1995 525
Before photochemical treatment
' /-
0
1
3
2
time
4
before treatment
6
5
0
(days)
Figure 8. Biochemical oxygen demand before and after 2.5 h pretreatment applied to a solution containing 2-aminophenol vs time. h with a Suntest lamp (see Experimental Section). Due to the dearomatization of 2-aminophenol, bacterial attack would take place on the intermediate degradation products (of 2-aminophenol). It is noted that after 3 days the bioprocess alone provides better conversion after photochemical treatment. It is readily seen from Figure 8b that after photochemical treatment no adaptation time is needed and 0 2 consumption starts from the beginning. Before photochemical treatment a BODEJCODratio of 90/ 380 = 0.23 is observed in Figure 8a. A pretreated solution has initially a COD of 204 mg O h and shows after 5 days a BOD value of 82 mg O h . This indicates a BODEJCOD ratio of 0.40. This value being bigger than 0.23 means that photochemical pretreatment facilitates the biodegradation process. The NH2 group has been shown to be susceptible t o bacterial attack, leading to generation of a n OH by oxidative deamination.2,21 The measured uptake of 0 2 falls short of the value for total combustion to C02 H2O (Figure 8a). Therefore only a portion of the substrate is oxidized to completion. Phenol shows a BODEJCODvalue of 0.29. This would confirm that amino aromatics present nonclassical behavior during b i o d e g r a d a t i ~ n . ~2-Ami,~,~ nophenol was found to degrade about 30% faster than 3-aminophenol, showing the determining influence of the substituent on the observed degradation rates. Figure 9 shows that photochemical pretreatment accelerates the attack on the substrate by bacteria found in biological waste water treatment. We report data for DOC (dissolved organic carbon or TOC, total organic carbon) for a solution photochemically pretreated for 2.5 h. It is seen from Figure 9 that an adaptation period of 1day was necessary when no photochemical pretreatment was applied. After 2.5 h ofpretreatment, the aromatic moiety no longer seems to exist (see Figures 1and 4). Thus after the ring cleavage takes place the bacterial action proceeds more readily on smaller aliphatic compoundswith a higher degree of oxidation. Bacterial attack proceeds more
+
(21) Pitter, P.Acta Hydrochim. Hydrobiol. 1986, 13, 453.
/
1
3
2
4
5
6
time (days)
Figure 9. Percentage of carbon degradation as a function of time before and after 2.5 h photochemical pretreatment. Refer to the text for other details. readily on pretreated solutions, as shown in Figure 9b. The attack proceeds without an adaptation p e r i ~ d , ~ J ~ , ~ ~ as reported previously by our laboratory for other compounds.
G. Physical Characterization of the a-FezOsIntervening the Catalysis. The crystallographic phase of the hematic a-Fe203did not change during the catalysis. At time 0 (Figure 2, light, 0 2 ) hematite with -3% FeO was found by X-ray diffraction. At the end of the process (t > 4 h) the FeO had disappeared since a n overall oxidative process had taken place. Finely crystalline a-FezO3 was observed by this technique. Electron microscopy (TEM) revealed regular isometric articles of hematite with a basic particle unit of -100 and a n average aggregate unit diameter of 0.06 pm. No variation of the aggregate size was observed before and after the catalysis. X-ray photoelectron spectroscopy (XPS)measurements were carried out on unreacted and reacted samples of a-Fez03. On both samples the Fe 2p 3/2 signal doublet was observed a t 711.1 eV, indicating no valency changes for Fe in a-FezO3. The binding energy (BE) taken as reference for Fe in a-Fe2Os was 706.7 eV. Nitrogen was found as ammonium on a-FezO3 after use. The C found by XPS on the a-Fe203 surface was predominantly aliphatic in nature with traces of aromatic compounds after a 4 h run (Figure 2, light, 02).
x
Conclusions This study shows for the first time the use of a-FezOs in the mineralization of 2-aminophenol. The photooxidation of 2-aminophenol in a n aerated iron oxide suspension could be due to (1)a ligand-to-metal charge transfer process within the iron(II1) surface complex with concomitant oxidation of the complex and reduction of the iron surface, which is subsequently reoxidized by oxygen in the presence of air, (2)photo Kolbe type reactions in which the holes are scavenged by 2-aminophenol and the electrons by the surface iron that is subsequently reoxi-
Pulgarin and Kiwi
526 Langmuir, Vol. 11, No. 2, 1995 dized by oxygen, and (3) the OH radicals resulting from the oxidation of surface OH groups. The photoproduced holes do not seem to effectively react with surface OHgroups to produce OH radicals since the degradation of phenol, chlorophenol, and nitrophenol is not observed in a-Fe203-mediated processes. The photocatalyzed degradation of 2-aminophenol is more related to a surface complex and not t o a semiconductor-assisted photodegradation. Anew aspect presented in this study is the integration of photocatalysis via a-FezO3 coupled with biological treatment. This is important since a cost-effective technology could be applied to other persistent pollutants with electron-donating character. In our case, photocatalysis
initially induces rapid ring dearomatization followed by slower mineralization of the aliphatic intermediates. A knowledge of the fundamental mechanism of photocatalytical decay for our model compound may help to determine the most efficient oxidation of pollutants under visible light in waste treatment plants.
Acknowledgment. We appreciate the help of R. Cloux with the lH-NMR work, V. Vlachopoulus with the cyclic voltametry measurements, J-Paul Schwitzguebelwith the ammonidnitrate determination, and P. Albers of Degussa A.G. with the XPS experiments. LA9405206