Environ. Sci. Technol. 1982, 16, 396-402
Maenpaa, R.; Hynninen, P.; Tikka, J. Pap. j a p u n . 1968, 50, 143-150.
Rogers, I. H.; Manville, J. F.; Sahota, T. Can. J. Chem. 1974, 52, 1192-1199. Bowers, W. S.; Fales, H. M.; Thompson, M. J.; Uebel, E. C. Science (Washington, D.C.) 1966, 154, 1020-1021.
(31) Leach, J. M.; Thakore, A. N.; Manville,J. F. J . Fish. Res. Board Can. 1975, 32, 2556-2559. (32) Ekundayo, 0. J . Chromatogr. Sci. 1980, 18, 368-369. Received for review February 23, 1981. Revised manuscript received November 16, 1981. Accepted March 22, 1982.
Oxidation of Phenol and Hydroquinone by Chlorine Dioxide Johannes Edmund WaJon,+David H. Rosenblatt,* and Elizabeth P. Burrows US Army Medical Bioengineering Research and Development Laboratory, Fort Detrick, Frederick, Maryland 2 170 1
Rates of reaction of chlorine dioxide with phenol and with hydroquinone were determined with a stopped-flow spectrophotometer in the pH range 4-8. Second-order rate constants increase with increasing pH, consistent with a mechanism in which both the free phenol and the more reactive phenoxide anion react with C102. Removal of an electron from the substrate by CIOz to form a phenoxyl radical and C102- ion is the rate-determining step. Subsequently, in the case of hydroquinone, C102removes another electron from the radical, forming p-benzoquinone and another C102- ion. In the case of phenol, CIOz adds to the phenoxyl radical para to the oxygen, and p-benzoquinone is formed with concomitant release of HOC1. The mechanism for phenol reaction accounts for (i) the immediate formation of p-benzoquinone without apparent intermediacy of hydroquinone, (ii) the chlorination observed in solutions containing excess phenol, and (iii) the production of only 0.5 mol of ClO;/mol of C102consumed. When aqueous chlorine is used as a disinfectant in phenol-bearing waters, malodorous chlorophenols are formed which persist for long periods unless large excesses of chlorine are applied (1). However, tastes and odors are completely avoided when chlorine dioxide in much smaller doses is used (2, 3), and it is generally believed that chlorinated products will not form when chlorine dioxide is used as a disinfectant. At significantly higher concentrations of both reactants (millimolar or greater) at nearly neutral pH, increasing ratios of C102to monohydric phenols give larger amounts of oxidized products (e.g., p benzoquinones and oxalic acid) relative to chlorinated products (e.g., chlorophenols and chloro-p-benzoquinones) ( 4 , 5). Chloro-p-benzoquinones, but not chlorophenols, have been found even when CIOz was in excess ( 4 ) . With dihydric phenols such as hydroquinone only oxidation occurs when C102 is in great excess (6),but chlorination may occur when C102is not in excess (5). It was supposed that in these concentrated solutions chlorination occurred before any oxidation, because p-benzoquinone and chlorinated p-benzoquinones do not react with chlorine ( 4 ) . In contrast, it was proposed, from studies in more dilute solutions at pH 2, that phenol is oxidized in the rate-determining step, and any excess phenol is then chlorinated (7). It was also postulated that HOC1, formed slowly from the disproportionation of HC102,reacts rapidly with more HC102 to form C120z,which then chlorinates the phenol. However, this mechanism did not satisfactorily explain the yield of chlorite or chlorate from the reaction. Alternatively, HOCl may form directly from reaction of C102with a variety of phenols, especially guaiacols (8). Chlorination would then be a result of direct reaction of the phenol with HOCl or C12rather than with C1202,which also forms from t National Research Council Postdoctoral Associate
396
Environ. Sci. Technol., Vol. 16, No. 7, 1982
1980-1981.
HOCl and C102- but quickly decomposes to CIOz, C103-, and C1-. Rates of reaction between CIOz and phenol have been measured between pH 0 and 2 (7) and between pH 2.5 and 4.5 (8). The rate increases with increase in pH, and from the dependence of the second-order rate constant on pH, it has been proposed that both the free phenol and the phenoxide anion react with C102, though the phenoxide anion is more reactive by several orders of magnitude (7). Extrapolation of the rate constant suggested that the reaction should have a half-life of several milliseconds at millimolar concentrations and pH 7. It was the aim of the present investigation to clarify the stoichiometry and mechanism of the oxidation of phenol by C102and especially to establish the extent of formation of chlorinated products at neutral pH under conditions approximating those of water treatment. Rates of reaction between 5-250 X M phenol and 2-65 X loT5M CIOz (13-440 mg/L of C102)were determined with a stoppedflow spectrophotometer in order to confirm the extrapolation of Grimley's data to pH 7 and to provide support for the mechanism proposed here. The reaction between 13-100 X M hydroquinone and 2-35 X M C102 was also investigated. Experimental Methods Preparation of Solutions. All solutions were made with glass-distilled, deionized water which was irradiated with a mercury vapor lamp, then boiled, and regassed with Nz before use. Solutions of CIOz (0.017 M) were prepared from reagent grade potassium persulfate and sodium chlorite (9) and stored in a low actinic glass bottle at 2 "C for no longer than 6 months. Phenol (MC/B reagent) was purified by distillation under nitrogen and stored at 2 "C. p-Benzoquinone (Baker reagent) and 2-chlorohydroquinone (Pfaltz and Bauer) were purified by steam distillation. The following chemicals were purified by recrystallization: NaCIOz (MC/B), 4-chlorophenol (Chemical Service Co.), 2,6-dichlorophenol (Aldrich),hydroquinone (Aldrich), 2,5-dichlorohydroquinone(Pfaltz and Bauer), and 2,5-dichlorobenzoquinone (Pfaltz and Bauer). 2Chlorophenol (Chemical Service Co.), 2,4-dichlorophenol (Eastman), and sodium hypochlorite (570, Baker) were used as received. 2,6-Dichlorobenzoquinonewas prepared from 2,4,6-trichlorophenol (IO), and 2,3-dichlorobenzoquinone from 2,3-dichlorophenol (11)by Cr03/acetic acid oxidation. 2,6- and 2,3-dichlorohydroquinoneswere prepared by reduction of the respective dichlorobenzoquinones with NaBH, in ethanol. Acetic acid, sodium acetate, NaH2P04,Na2HP04,and NaC104, used in preparing buffer solutions (129,were all reagent grade chemicals. The concentration of NaC104 in these solutions was 0.1 M, while those of phosphate and acetate were usually 0.01 and 0.02 M, respectively. A Markson Science Inc.
Not subject to U.S. Copyright. Published 1982 by the American Chemical Society
Model 884 micro pH electrode and a Corning Scientific Instruments Model 12 pH meter were used to measure pH. Solutions of oxychlorine species were analyzed iodometrically with a Fischer-Porter amperometric titrator. The concentration of C102+ HOCl was determined at pH 7 by adding a sample to 200 mL of water containing phosphate buffer (2 mL) and 5% KI (1mL) and titrating immediately with 0.005 64 N phenylarsine oxide (Fisher Scientific Co.). Chlorite was determined subsequently in this solution at pH 2 by adding 2 mL of 6 N H2S04and titrating with phenylarsine oxide after 6 min. Kinetic Analysis. Kinetic studies were performed in a Durrum (Dionex Corp., Sunnyvale, CA) Model 110 stopped-flow spectrophotometer in which 0.075-mL solutions of C102 and the phenol were mixed in a 2-cm path length, 0.066-mL cuvette with a deadtime of 1.6 me. The reaction temperature was 25.0 f 0.1 “C. Reactions were followed by measurement of light absorbance, done by monitoring the disappearance of C102 at 359 nm with a Biomation (Gould, Inc., Santa Clara, CA) 805 waveform recorder and a Tektronix 7613 oscilliscope. The data from the least noisy of similar curves of absorbance vs. time, spanning 95-99% of the reaction, were stored in a Data Graphics Datos 305 interface and then transferred to a digital computer. The pH of the reacted solution was determined. As will be shown, the stoichiometry of the reaction can be represented by eq 1. The reaction was found to be first phenol
+ 2C102
k2
products
(1) order with respect to each reactant on the basis of the initial rates of reaction, which were determined by fitting a straight line through the first 1-2% of the absorbancetime data by linear regression. The applicable rate expression is -d[ClOz] /dt = 2kz[PT] [ClO2] (2)
where [PT] = total concentration of the phenol and k2 = observed second-orderrate constant. Kinetic studies were carried out both with a molar excess of C102and a molar excess of the phenol. When the phenol was in excess, it was present in at least a 5-fold and up to a 120-fold stoichiometric excess. Equation 2 thus becomes -d[ClOz] = 2kob,d[C102] (3) where (4) IZobad = k2[PT1 and kobsd is the pseudo-first-order rate constant. Upon integration and use of Guggenheim’s method (13),this can be expressed as In ( A - A? = -2kobsdt In ( A , - A,) + In e-kobedA- 1 (5) where A = absorbance at time t , A ’ = absorbance at time t + A, A, = absorbance at time t = 0, A , = absorbance at time t = m, and A is a time interval greater than the half-life of the reaction. Pseudo-first-order rate constants (k,,,,) were determined by plotting the function In ( A A 9 vs. time for the initial 25-50% of the absorbance-time data by using the linear regression model RLONE (International Mathematical and Statistical Libraries Inc., Houston, TX). If phenol is in stoichiometric excess, eq 2 can also be expressed as (14)
+
which can be written 1 t=-ln X
B(A,-A) D-A/2
(7)
where a, = total initial concentration of phenol, bo = initial concentration of CIOz,X = (bo- 2a0)k,B = ao/bo,C = ( A , - Ao),and D = BC 1/2Ao. Rearranged and expressed exponentially, eq 7 becomes
+
A=
DeXt- BA, ext/2 - B
If phenol is not in stoichiometric excess, eq 2 becomes (bo/ao)(A,- A ) kt = In (9) 2ao - bo (bo/ao)(A,- A,) - 2(A - A,) Making the substitutions X = (2ao- bo)k,B = bo/ao,C = ( A , - Ao),and D = BC 2Ao and rearranging as before, we get
+
DeXt- BA, (10) 2ext - B Second-orderrate constants, k,were determined directly by fitting either eq 8 or eq 10 to the initial 2550% of the absorbance-time data by using the curve fitting simulation, analysis, and modeling (SAM) program (15). The parameters X , C, and A , were adjustable. Product Analysis. A solution of C102, NaC102, or HOCl was mixed rapidly and uniformly with a phenol solution by forcing the two together in the junction of two l-mm diameter glass tubes. Mixing of 20-mL solutions was achieved within 10 s. The mixtures were kept at 25 f 0.3 OC and were sampled for analysis of residual oxidants and organic compounds. The first samples were taken 45 s after mixing. Residual oxidants were measured at pH 7 and 2 as described above. In solutions where it had been shown that all the C102was consumed within several seconds, any titer of a pH 7 sample was presumed to be due to HOCl alone. However, no satisfactory method was found to determine the concentrations of C102and HOCl when they were both present. All published titrimetric methods (16-23) were found to be inconsistent. The concentration of C10, determined after either all the C102 or all the HOCl had dissipated was presumed to be the same as that present initially. In cases where it was likely that neither HOCl nor C102had completely disappeared, we assumed that the pH 7 titration formed C102-, which appeared as “excess” CIOz- on titration at pH 2. Any portion of the pH 7 titer not accounted for as CIOz via the pH 2 titration was presumed to be HOC1. Organic compounds were analyzed by HPLC. Acetylation followed by GC analysis (24) was found unsatisfactory because the acetylation procedure produced chlorinated hydroquinones and benzoquinones. The liquid chromatographic system consisted of the following: (1)two Model 6000A pumps and Model M660 solvent programmer (Waters Assoc.); (2) SF-770 variable wavelength detector (Schoeffel Instrument Corp.) set at 220 nm; (3) Sigma 10 Data Station (Perkin-Elmer Corp.); (4) 300 X 3.9 mm 10 pm pBondapak C18reversed-phase column (Waters Assoc.). A linear gradient elution program was used in which the eluent changed from 100% 0.02 M KH2P04pH 2.8 to 50% acetonitrile-water (80/20) in 30 min at 1.5 mL/min and 1200 psi. Under these conditions, phenol, mono- and dichlorophenols, hydroquinone, mono- and dichlorohydroquinones, p-benzoquinone, and mono- and dichloro-pA=
Environ. Sci. Technol., Vol. 16, No. 7, 1982
397
Table I. Stoichiometry of the Reaction of C10, with Phenols [C102 I/ 10-4,M (I) Phenol + C10,
AC10,
[reductant ] I 10-4,M
PH
12.62 5.00 5.00 5.00 2.32 2.32 0.91
6.89 6.93 6.83 6.90 6.84 6.88 6.87
Aphenol
A benzoquinone
AC10,-
Aphenol
AC10,
(a) Phenol in Excess
5.25 6.40 3.15 0.94 2.90 0.94 0.94
0.82 1.00 1.00 1.09 0.65 1.06 0.72 __
1.96 1.72 3.33 2.94 2.00 2.86 1.85
av 2.4
f
0.6
0.9
0.62 0.52 0.50 0.62 0.44 0.52 0.51 __
+ 0.2
0.5 + 0 . 1
( b ) C10, in Excess 3.09 2.78 3.09
0.91 0.91 0.43
6.84 6.90 6.87
1.52 1.47 2.08 __
0.61 0.63 0.67 __
av 1.7 f 0.3 ( 2 ) Hydroquinone
0.6
+ C102
2.70 3.09
3.80 0.48
0
6.98 6.92
1.6 1.2
0.30 0.46 0.38 __
* 0.1
0.4 + 0.1
1.00 0.68
0.84 0.61
I
0
0.80 x 1U4M HOC1
o 0.80x 1u4uHOCI-
- 2.32 x
lO-'M PHENOL 8.88
8.8 x 1 0 . ' ~ N~CIO,- 2.32 x 1 6 ' PHENOL DH 8.01
0.4
x
i
- 2.32 i I O - ~ MPHENOL pH
benzoquinones at concentrations 1 5 X lo4 M were eluted within 35 min. Results
Products and Stoichiometry. When 1-7 X M solutions of C102 (5-45 mg/L) were mixed with phenol in a phenol to chlorine dioxide mole ratio 10.75 at pH 6.85-6.95, C102 disappeared within seconds, although an oxidizing titer at pH 7, presumed to be due to HOC1, remained for periods up to about 10 min. The products detected within 1 min were p-benzoquinone, 2-chlorophenol, 4-chlorophenol, 2,4-dichlorophenol, and 2,6-dichlorophenol. After 10 min, no oxidant titer was present, the concentration of the four chlorophenols had increased substantially, hydroquinone had also formed, and the concentration of p-benzoquinone had decreased slightly. Environ. Sci. Technol., Vol. 16, No. 7, 1982
8.84
_--- -
TIME ( MIN 1
Figure 1. Products of the reaction of 2.90 X M CIO, with 2.32 X M phenol at pH 6.84 after 4.0 min: (1) hydroquinone; (2) benzoquinone; (3) phenol; (4) 2-chlorophenol; (5) khlorophenol; (6) 2,6-dichlorophenol; (7) 2,4dichlorophenol.
398
Z . B J X ~ O - ~ CIO, M
p240CP HZ4DCP 0
0
zoo
100
TIME (
SEC
:;3
+ -
)
Figure 2. Rates of formation of chlorophenols from reaction of C102 or HOC1 with phenol at pH 6.9.
Figure 1 presents a typical chromatogram, and Figure 2 shows the rates of formation of the chlorophenols. Over the next 2 h, the concentration of hydroquinone increased rapidly while the concentration of benzoquinone decreased in almost a 1:l correspondence (see Figure 3). Chlorinated hydroquinones or chlorinated benzoquinones were not detected at any time. From the results of seven experiments (see Table I), the ratio of C102consumed to phenol oxidized was found to be 2.4 f 0.6, and of the phenol oxidized, 90 f 20% was oxidized to p-benzoquinone. The ratio of C102- formed to C102consumed was found to be 0.5 f 0.1. When 3 X lo4 M C102solutions (20 mg/L of C102)were mixed with phenol in a C102/phenol molar ratio >3, all the phenol was consumed within seconds. p-Benzoquinone
1.0 T
b '0 BQ
v
1.26
1 0 - 4 ~CIOS 0.62
D
0.526
i
10%
x
1 0 ' ' ~ PHENOLPH 8.8s
BENZOQUINONE pH 8.92
o
2.90 x 10-414
+
2 . 7 8 I. IrT4M C l o p i 0 . W I 1O"M
ci6? + 2 . w
x
io-'H
PHENOL PH 8.84 PHENOL pH 8.BO
Flgure 4. Absorbance-time plots for the reaction of C102 with phenol at pH 7.0. b
, I
I
i
lb TIME
i 1
0
3
~
~
li?
12
~
Figure 3. Rates of change of concentratlons of hydroquinone and benzoquinone during the reaction of C102 with phenol at pH 6.9.
was the only organic product detected, but the yield was only 64 f 3% (Table I). The ratio of CIOz consumed to phenol oxidized was 1.7 f 0.3, and the ratio of CIOzformed to C102 consumed was 0.4 k 0.1. Over the next 2 h, the concentration of benzoquinone decreased steadily while that of hydroquinone increased (Figure 3). No chlorinated products were detected even though the presence of HOCl was suspected (on the basis of oxidant titer) for periods up to 10 min. When 3 X lo4 M CIOz solutions, either in molar excess or molar deficiency, were mixed with hydroquinone, the sole products detected were p-benzoquinone and CIOz-, whose concentrations remained unchanged for many hours. There was no evidence for the presence of HOCl at any time. The ratio of C102 consumed to hydroquinone oxidized was 1.6 when hydroquinone was in excess and 1.2 when CIOz was in excess. The ratio of CIOz- formed to CIOz consumed was 1.0 when hydroquinone was in excess and 0.7 when CIOz was in excess. While the accuracy of the oxidant analyses may be suspect, especially in solutions containing excess CIOz,due to the simultaneous presence of several oxychlorine species that could not be determined separately, it nevertheless appears that 1 mol of hydroquinone consumes 2 mol of CIOz,forming 2 mol of C102-, while 1 mol of phenol also consumes 2 mol of C102 but forms only 1 mol of CIOz-. (Note added in proof: This stoichiometry was later confirmed in each case through ion chromatographic measurement of CIOz- formed.) Kinetics. Reaction kinetics of CIOz and phenol and of CIOz and hydroquinone were observed in the pH range 4.5-8.0 at 25.0 "C. Absorbance-time data for three reactions at pH 7 are shown in Figure 4. In solutions with stoichiometric excesses of phenol or hydroquinone, plots of In ( A - A3 as a function of time, in accord with eq 5, were linear for at least 5 half-lives of the reaction, con-
t B 10 0'0
\ 0'08
;
O'ld
0'2,
TIME
'
0'11
'
0
Go
I BEC I
Figure 5. Pseudo-first-order plots for the reaction of C102 with phenol at pH 7.0.
Table 11. Second-Order Rate Constants at 25.0 "C for the Disappearance of C10,
7.05 2.28 5.17 1.14 2.22 7.05 2.28 5.17 2.46 7.01 12.5 5.17 2.56 7.01 12.5 5.17 2.65 6.92 2.14 50.0 13.4 2.69 6.92 2.14 50.0 13.0 2.61 6.99 12.5 50.0 13.2 2.65 6.99 12.5 50.0 2.84 6.89 2.14 126.0 34.2 2.70 6.89 2.14 126.0 34.6 2.74 6.92 12.8 126.0 28.6 2.26 6.92 12.8 126.0 2.34 6.91 23.0 126.0 32.0 2.54 6.91 23.0 126.0 2.81 When no value is listed for hob+, k , was obtained directly by fitting a second-order kinetic expression t o the data.
sistent with the assumption of pseudo-first-order behavior. A typical plot, from which the value of k o mwas calculated, is shown in Figure 5. Rate constants (kz),calculated from kobsd by means of eq 4, were constant with varying initial phenol or hydroquinone concentrations, as shown in Table Environ. Sci. Technol., Vol. 16, No. 7, 1982
399
Table 111. Second-Order Rate Constants a t 25.0 "C for t h e Oxidation of Phenol by C10,
PH
ionic strength, M
Y
7.99 7.05 7.00 6.90 6.90 5.81 5.78 5.76 4.60 4.58 4.52 2.18 1.75 1.45 1.06 0.17
0.129 0.200 0.200 0.200 0.120 0.111 0.111 0.111 0.110 0.110 0.110 1.00 1.00 1.00 1.00 1.00
0.763 0.736 0.736 0.736 0.767 0.772 0.772 0.772 0.773 0.773 0.773 0.683 0.683 0.683 0.683 0.683
k , , M-l
~
s-l
0.2) x l o 5 0.1) X l o 4 0.1) x lo4 0.1) x lo4 0.1) x l o 4 0.05) X l o 3 (1.8 i 0.1) X l o 3 (1.8 t 0.05) X l o 3 (1.2 t 0.05) x 10' ( 1 . 3 t 0.05) X 10' (1.3 t 0.05) X 10' 0.80b 0.44b 0.34b 0.2gb 0.24b
(3.5 i (2.3 i (2.6 t (2.6 t (2.6 t (1.8 i
Table IV. Second-Order Rate Constants a t 25.0 "C for the Oxidation of Hydroquinone by C10,
-
s - 1i
3.2 X 3.8 X 3.4 X 2.7 X 2.6 X 2.1 X 2.0 X 1.9 X 1.3 x 1.2 X 1.1x 0.80 0.44 0.34 0.29 0.24
a
10'
lo4 lo4 10' 10'
lo3 lo3 lo3 10' 10'
loz
11, thus confirming overall second-order behavior. For several of-these reactions, the absorbance-time data were also fitted to eq 8, and values of kz so obtained directly weie equal to the values obtained with eq 4 and 5 (see Table 11). Values of kz for solutions with CIOz in excess, obtained by means of eq 10, were equal to those for solutions with phenol or hydroquinone in excess (see Table 11). Second-order rate constants (kz)decreased with decrease in pH. In the case of phenol, k2 decreased 10-fold for each unit pH decrease above pH 4 (see Table 111),and began to level off at pH 6 and began to level off at pH
