Photolysis of aqueous free chlorine species (NOCI and OCI-) with 254 nm ultraviolet light

Article (PDF Available)inJournal of Environmental Engineering and Science 6(3):277-284 · May 2007with1,287 Reads
DOI: 10.1139/S06-052
Abstract
The quantum yields of the UV photolysis of free chlorine (OCl− and HOCl) at 254 nm were measured in a series of batch reactor experiments from pH 5 to 10 and at various concentrations. When the concentration of free chlorine is low (3.5 mg Cl/L) to moderate (70 mg Cl/L), the quantum yields of HOCl and OCl− are 1.0 ± 0.1 and 0.9 ± 0.1, respectively. When the concentration increases to higher levels (>70 mg Cl/L), the quantum yield of HOCl photolysis increases significantly, whereas the quantum yield of OCl − photolysis is essentially independent of concentration. In addition, based on the experimental results obtained in this research, a mathematic model was developed that can be used for the prediction of the quantum yield for the UV photolysis of free chlorine at 254 nm. The quantum yields predicted by this model agree very well with the measured data. Also, the dependence of free chlorine decomposition on the fluence (UV dose) and the effect of water quality on the quantum yield of free chlorine species were investigated in this research.
277
Photolysis of aqueous free chlorine species
(HOCl and OCl
) with 254 nm ultraviolet light
1
Yangang Feng, Daniel W. Smith, and James R. Bolton
Abstract: The quantum yields of the UV photolysis of free chlorine (OCl
and HOCl) at 254 nm were measured in a
series of batch reactor experiments from pH 5 to 10 and at various concentrations. When the concentration of free chlorine
is low (3.5 mg Cl/L) to moderate (70 mg Cl/L), the quantum yields of HOCl and OCl
are 1.0 ± 0.1 and 0.9 ± 0.1,
respectively. When the concentration increases to higher levels (>70 mg Cl/L), the quantum yield of HOCl photolysis
increases significantly, whereas the quantum yield of OCl
photolysis is essentially independent of concentration. In
addition, based on the experimental results obtained in this research, a mathematic model was developed that can be used
for the prediction of the quantum yield for the UV photolysis of free chlorine at 254 nm. The quantum yields predicted by
this model agree very well with the measured data. Also, the dependence of free chlorine decomposition on the fluence
(UV dose) and the effect of water quality on the quantum yield of free chlorine species were investigated in this research.
Key words: ultraviolet irradiation, free chlorine, fluence, UV dose, quantum yield, disinfection.
Résumé : Les rendements quantiques de la photolyse UV à 254 nm du chlore libre (OCl
et HOCl) ont été mesurés dans
une série d’expériences en réacteur discontinu à des pH de5à10etàdiverses concentrations. Lorsque la concentration
en chlore libre est de faible (3,5 mg Cl/L) à modérée (70 mg Cl/L), les rendements quantiques du HOCl et du OCl
sont
de 1,0 ± 0,1 et 0,9 ± 0,1 respectivement. Lorsque la concentration augmente à des niveaux plus élevés (>70 mg Cl/L),
le rendement quantique de la photolyse du HOCl augmentait de manière significative, alors que le rendement quantique
de la photolyse du OCl
est essentiellement indépendante de la concentration. De plus, en se basant sur les résultats
expérimentaux obtenus au cours de cette recherche, un modèle mathématique a été mis au point qui peut être utilisé
pour prédire le rendement quantique de la photolyse UV du chlore libre à 254 nm. Les rendements quantiques prédits
par ce modèle concordent très bien avec les données mesurées. La dépendance de la décomposition du chlore libre sur la
fluence (dose UV) et l’effet de la qualité de l’eau sur le rendement quantique des espèces de chlore libre ont également été
examinés au cours de cette recherche.
Mots-clés : rayonnement ultraviolet, chlore libre, fluence, dose UV, rendement quantique, désinfection.
[Traduit par la Rédaction]
Introduction
In many ultraviolet (UV) disinfection installations, free chlo-
rine is present in the water as it passes through the UV reactors.
A good example is the E.L. Smith Water Treatment Plant of
EPCOR Water Services in Edmonton, Canada, which has one
of the largest UV disinfection systems in the world. Here the
UV equipment was installed after the chlorine microorganism
reduction (MOR) unit (EPCOR 2004). A major advantage of
this combined process is that the UV light can inactivate proto-
Received 22August 2005. Revision accepted 17August 2006. Pub-
lished on the NRC Research Press Web site at http://jees.nrc.ca/ on
23 May 2007.
Y. Feng, D.W. Smith, and J.R. Bolton.
2
Department of Civil
and Environmental Engineering, 3-133 Markin/CNRL Natural Re-
sources Engineering Facility, University of Alberta, Edmonton, AB
T6G 2W2, Canada.
Written discussion of this article is welcomed and will be received
by the Editor until 30 September 2007.
1
This article is one of a selection of papers published in this
special issue on application of ultraviolet light to air, water, and
wastewater treatment.
2
Corresponding author (e-mail: jb3@ualberta.ca).
zoa (e.g., Cryptosporidium spp., and Giardia spp.), which are
difficult to treat by chlorination (Bolton et al. 1998; Bukhari
et al. 1999; Clancy et al. 2000). From another point of view,
the interactions between free chlorine and UV light could have
some negative effects on the disinfection ability of this kind
of process. Örmeci et al. (2005) found that the UV fluence is
influenced slightly by free or combined chlorine, which can in-
crease the UV absorbance of the drinking water. However, the
free chlorine itself can be affected significantly by UV light.
For example, in the daily operation of outdoor swimming pools,
one usually finds that the free chlorine level shows a marked
decrease on sunny days, arising from photolysis by the UV
portion of sunlight. In an experiment in which UV disinfection
was employed in water chlorinated upstream of the UV dis-
infection unit, Zheng et al. (1999) reported that the higher the
fluence (UV dose) applied the greater the chlorine dose that was
required to keep a fixed chlorine concentration in the effluent
water. At the highest fluence (UV dose) of 4825 mJ/cm
2
em-
ployed, the chlorine demand was about 5 times that obtained
without UV applied. In addition, according to the authors’ pri-
vate communication with staff of EPCOR in Edmonton, Can-
ada, the concentration of free chlorine residual shows a distinct
decrease as the water passes through the UV reactor in the E.L.
Smith Water Treatment Plant cited above.
J. Environ. Eng. Sci. 6: 277–284 (2007) doi: 10.1139/S06-052 © 2007 NRC Canada
278 J. Environ. Eng. Sci. Vol. 6, 2007
Fig. 1. Dependence of the ratio HOCl/OCl
on pH (pK
a
= 7.5).
OCl
-
60
HOCl
Percentage of HOCl or OCl
-
in aqueous chlorine solution %
pH
567
8
9
10
0
20
40
100
80
When Cl
2
is added to water, it reacts according to
[1] Cl
2
+ H
2
O HOCl + HCl
HOCl is a weak acid such that the relative concentrations
of the acid (HOCl) and its conjugate base (OCl
) are strongly
dependent on pH and follow the equilibrium
[2] HOCl OCl
+ H
+
(pK
a
= 7.5at 25
C)
The lower the pH, the higher the percentage of HOCl; at
higher pH, OCl
will be the predominant species. According
to the equilibrium of reaction 2,
3
more than 99% of the free
chlorine is HOCl at pH 5 and similarly more than 99% is OCl
at pH 10. Figure 1 shows this relation graphically.
There havebeen some earlier studies concerning the reactions
occurring in the UV photolysis of aqueous free chlorine. Pri-
marily hydroxyl radicals (·OH) and chlorine radicals (·Cl) are
generated in the photodecomposition of free chlorine in water
(Nowell and Hoigné 1992a). Because of the presence of some
organic matter in drinking water, two chain-reactions have been
proposed for the decomposition of HOCl by Oliver and Carey
(1977):
·OH radical chain
·OH + RH →·R + H
2
O[3a]
·R + HOCl RCl OH[3b]
·Cl radical chain
·Cl + RH →·R + HCl[4a]
·R + HOCl ROH Cl[4b]
Buxton and Subhani (1972a) carried out a systematic inves-
tigation of the photodecomposition mechanism of hypochlorite
ions (OCl
) in aqueous solution. They found that the quan-
tum yield for OCl
decomposition is about 0.85 at 254 nm.
3
Morris (1966), Snoeyinkand Jenkins (1980), and Sawyeret al. (1994).
Nowell and Hoigné (1992a) studied the free chlorine exposed
to sunlight or UV light at 254 nm and found that the produc-
tion ratio of ·OH generated by OCl
(pH > 8) is only about
0.1 and HOCl (pH 5) yielded 0.7 ·OH in sunlight and 0.9 ·OH
with 254 nm UV light. However, the molar absorption coeffi-
cients of HOCl and OCl
at 254 nm reported by Nowell and
Hoigné (1992b) are much higher than those obtained by other
researchers. Table 1 summarizes published values of the molar
absorption coefficients at about 25
C.
Since in practice the pH is such that both HOCl and OCl
are present in most water or wastewater treatment, the overall
quantum yield of free chlorine decomposed is more important
than the individual quantum yields of HOCl and OCl
reported
as a unique species. In addition, the concentration of free chlo-
rine may have an influence on the quantum yield because of
possible chain reactions. Therefore, the main objective of this
research was to investigate the effects of pH and concentration
of free chlorine in solution on the overall photolysis quantum
yield at 254 nm. Based on the experimental results, a model
was derived for predicting the overall quantum yield of free
chlorine, which can be applied to practical operations.
Experimental
Materials and methods
All chemicals used in this research were analytical reagent
grade and deionized water (chlorine demand free) was used for
all analytic determinations. Fresh sodium hypochlorite solution
(5.65 6%, Fisher Scientific Co., Canada) was diluted to pro-
duce the samples of free chlorine in all experiments. Other com-
pounds used in this research were stock reagent grade chemi-
cals.A quasi collimated beam UV apparatus (Model PSI-1-120,
Calgon CarbonCorporation, USA) and a 10 W low-pressureUV
lamp (Ster-L-RayGermicidal Lamp, model G12T6L 15114,At-
lantic Ultraviolet Corporation, USA) were used to generate a
nearly parallel beam of UV light at 254 nm. The UV irradiance
at the center of the UV beam on the surface of the free chlorine
sample was measured by a radiometer (Model IL 1400A with
an SED240 Detector, International Light Inc.), which was cal-
ibrated at 254 nm. An Ultraspec 2000 UV-Visible spectropho-
tometer (Pharmacia Biotech, Fisher Scientific Co., Canada) and
either a 10.0 or a 50.0 mm path length quartz cell (Fisher Scien-
tific Co., Canada) were used to determine the UV absorbance
of samples. In addition, some glassware (chlorine demand free)
and other routine equipments, such as a pH meter, electronic
balance, and magnetic stirrer, were also used in this research.
UV exposure and irradiance measurements
Two aliquots of free chlorine samples were used in each run:
one was used for exposure to UV light, and the other was used as
a dark control sample. The former sample was placed under the
open end of the plastic collimating tube of the UV collimated
beam apparatus in the center of the beam, and the latter sample
was placed in the dark on a bench that was far away from the
UV apparatus. The surface of the solution to be exposed was
about 300 mm from the UV lamp. The UV exposure time of the
© 2007 NRC Canada
Feng et al. 279
Table 1. Comparison of reported molar absorption coefficients of HOCl and
OCl
.
ε
235
/M
1
cm
1
ε
292
/M
1
cm
1
ε
254
/M
1
cm
1
HOCl OCl
OCl
HOCl Reference
368 155 121 Nowell and Hoigné (1992b)
100 350 58 62 Morris (1966)
105 60 Thomsen et al. (2001)
362 Chen (1967)
101±2 365±866±159±1 This work
Note: These data were determined at approximately 25
C.
free chlorine solution was controlled by means of a pneumatic
shutter located at the top of the collimating tube beneath the UV
lamp, and was activated using a stopwatch. During the exposure
period,a3mm× 12 mm Teflon
-coated stir bar was added
into each of these two samples to provide sufficient stirring. At
the beginning and end of the UV exposure, the concentration
of free chlorine in each sample was measured in order to deter-
mine the amount of free chlorine decomposed due to UV light
exposure. The incident irradiance was measured by position-
ing the radiometer vertically, such that the reference calibration
marker on the radiometer detector head was placed at the same
level as the top surface of the sample solution. During these
experiments, two similar (internal diameter 31.2 mm) Pyrex
beakers were used as the photolysis reactor and the control re-
actor. In addition, the beaker exposed to UV light was covered
with an opaque cap, containing a 14.2 mm diameter hole in
the center of the cover to let UV light into the sample in the
beaker. This was done to avoid any interaction of the UV beam
with the sides of the beaker. To allow for the same conditions
as the UV irradiated sample, a similar cap was also used for the
dark control sample. About 25 mL of the free chlorine sample
was added into the beaker for each run. Figure 2 is a schematic
diagram of the UV exposure test apparatus.
Quantum yield
The quantum yield, , is defined as the number of moles of
product formed or reactant removed per einstein
4
of photons
absorbed. Therefore, the quantum yield of free chlorine de-
composed can be determined from
[5] =
moles of free chlorine decomposed
einsteins absorbed at 254 nm
where “free chlorine” includes OCl
and (or) HOCl present in
the samples. In the pH range of 5 to 10 and in the absence of
free ammonia and amines, free chlorine exists primarily as an
equilibrium mixture of OCl
and HOCl. When the concentra-
tion of free chlorine was lower than 50 mg Cl/L it was measured
by the DPD method in this research (APHA et al. 1995). Other-
wise, the concentration was determined by the UV absorbance
of samples at an appropriate wavelength divided by the molar
absorption coefficient of the free chlorine at this wavelength.
4
One einstein is a mole (6.023 × 10
23
) of photons.
Fig. 2. Schematic of the collimated beam apparatus.
UV light
Low pressure UV lamp
Small beaker
Stir plate
Water
sample
Opaque cap with a
small hole
Collimated beam
Then, the amount of free chlorine decomposed can be calculated
according to
[6] moles of free chlorine decomposed
=
(
C
bi
V
bi
C
ai
V
ai
)
(
C
bc
V
bc
C
ac
V
ac
)
V
mi
V
mc
where, C is the concentration of free chlorine samples in M; V is
the volume of samples in L; the subscripts of “i” and “c” are the
“irradiated sample by UV light” and “dark control sample”; the
subscripts of “b” and “a” are the “before” and “after” exposed
to UV light; and “m means mean. The “einsteins absorbed at
254 nm” can be determined from
[7] einsteins absorbed at 254 nm
=
E × PF × A × t × (1 R) × (1 10
A
254
)
U
λ
where, E is the irradiance (W/cm
2
) measured at the center of
sample position; PF is Petri Factor defined as the average irra-
diance over the surface of sample exposed to UV light divided
by the irradiance at the center; A is the area of sample surface
directly exposed to UV light (i.e., the cross-sectional area of the
hole on the opaque cap covered on the sample beaker, cm
2
); t is
© 2007 NRC Canada
280 J. Environ. Eng. Sci. Vol. 6, 2007
Fig. 3. Absorption spectra of HOCl and OCl
. The absorbance of
HOCl and OCl
was measured, respectively, at pH 5 and pH 10
(temperature 21 ± 2
C).
250 300 350 400
0
100
200
300
400
500
OCl
-
HOCl
Wavelength (nm)
Molar absorption coefficient (M cm )
-1 -1
the exposure time (s); R is the reflection coefficient (0.025) at
the sample surface; A
254
is the mean absorbance of the sample
at 254 nm during the exposure time; and U
λ
is the energy of
one einstein of photons at 254 nm (471527.6 J/einstein).
Fluence (UV dose)
The fluence (UV dose) F was determined according to the
protocols specified by Bolton and Linden (2003). Specifically,
this involved converting the incident fluence rate (irradiance)
E
o
, as determined using the radiometer, to the average fluence
rate E
avg
utilizing the various correction factors specified in
that protocol.
Results and discussion
Absorption spectra and molar absorption coefficients of
HOCl and OCl
The absorption spectra and molar absorption coefficients of
free chlorine are very important to this research. The reason
is that the components of free chlorine mainly depend on pH.
According to the equilibrium of reaction 2, for free chorine
samples at pH 5 the dominant species is HOCl, whereas at pH
10, the dominant species is OCl
. Therefore, the amount of
free chlorine decomposed by UV light at different pH could
be different due to the different free chlorine species and their
different absorption spectra. The absorption spectra of HOCl
and OCl
are shown in Fig. 3.
The absorption spectra show that there is an absorption peak
at about 236 nm for HOCl; whereas for OCl
, the peak is at
about 292 nm. Using more than 15 concentrations overthe range
of 0 to 1000 mg Cl/L, the molar absorption coefficient obtained
by regression at 236 nm for HOCl is 101 ± 2M
1
cm
1
.In
the same manner, the molar absorption coefficient at 292 nm
for OCl
was determined to be 365 ± 8M
1
cm
1
. When the
concentration of free chlorine was more than 50 mg Cl/L, it was
measured in this research by the UV absorbance of samples at
236 or 292 nm divided by the molar absorption coefficient at
Fig. 4. Quantum yield of HOCl at pH 5 (temperature 21 ±
2
C): (a) Concentration < 71 mg Cl/L, and (b) Concentration
>71 mg Cl/L.
0 102030405060708090
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Experimental points
Regression line
95% confidence level
Concentration of HOCl (mg Cl
L
-1
0 200 400 600 800 1000 1200 1400 1600
0.0
1.0
2.0
3.0
4.0
5.0
6.0
(a)
(b)
)
Quantum yield of HOCl
the corresponding wavelength. In addition, the molar absorption
coefficients of HOCl and OCl
at 254 nm were also determined
to be 59 ± 1 and 66 ± 1M
1
cm
1
, respectively. These values
were used for the calculation of fluence absorbed by free chlo-
rine samples. Compared with the values listed in Table 1, the
results obtained in this research are very consistent with data
reported by Morris (1966), Chen (1967), and Thomsen et al.
(2001).
Quantum yields of free chlorine
The quantum yields of free chlorine (HOCl and OCl
) were
determined from Figs. 4(a)–4(b) and Fig. 5. The quantum yield
of HOCl is essentially constant at 1.0 ± 0.1 as long as the con-
centration is less than 71 mg Cl/L; however, for concentrations
from 71 to 1350 mg Cl/L, the quantum yield increases linearly
with a slope of 0.0025 (mg Cl/L)
1
, such that at 1350 mg Cl/L,
the quantum yield is 4.5 ± 0.2. In contrast, the quantum yield
of OCl
is virtually independent of concentration at 0.9 ± 0.1
when the concentration ranges from 3.5 to 640 mg Cl/L.
As reported by other researchers (Oliver and Carey 1977;
Buxton and Subhani 1972a), the concentration dependence of
the photolysis of HOCl can be explained by a chain reaction
© 2007 NRC Canada
Feng et al. 281
Fig. 5. Quantum yield of OCl
at pH 10 (temperature 21 ±
2
C).
0 100 200 300 400 500 600 700
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Experimental points
Regression
95% Confidence level
Concentration of OCl
-
(mg Cl L
-1
)
Quantum yield of
OCl
-
initiated by the reactions
HOCl + (UV photons) →·OH Cl[8a]
·OH + HOCl H
2
O OCl[8b]
where the ·Cl and (or) ·OCl species may participate in further
reactions that deplete the HOCl.
5
For example, the ·Cl atom may react with HOCl by
[9] ·Cl + HOCl HCl OCl
Chain termination may involve the reactions
·Cl OCl + H
2
O 2HOCl[10a]
2 ·OCl + H
2
O HCl + HClO
3
[10b]
The experimental results of this research indicate that the
above reactions have a greater effect on the quantum yield of
HOCl at higher concentrations than at lower concentration. At
pH 10, ·OH radicals can be generated by the reactions
OCl
+ (UV photons) →·O
Cl[11a]
·O
+ H
2
O →·OH + OH
[11b]
The ·OH radical reacts with OCl
by electron transfer with
a rate constant of 8.8 × 10
9
M
1
s
1
(Buxton and Subhani
1972b) to generate ·OCl. However, at pH 10, it appears that ·Cl
and (or) ·OCl species do not initiate any chain reactions, since
the quantum yield is independent of concentration at this pH.
A model for the prediction of the quantum yield of free
chlorine decomposed by UV light at 254 nm
According to the above discussion, free chlorine exists pri-
marily as an equilibrium mixture of hypochlorous acid (HOCl)
and its conjugate base (OCl
)withapK
a
of 7.5. Based on
5
No rate constant is available for reaction 8(b); however, it is likely that
this reaction proceeds with a rate constant near diffusion controlled
(10
10
M
1
s
1
).
the equilibrium of reaction 2, the fraction of HOCl existing in
free chlorine samples can be expressed as a function of pH as
follows:
[12] f =
[HOCl]
[OCl
]+[HOCl]
=
1
1 + 10
pHpK
a
where, f is the fraction of HOCl present in free chlorine sam-
ples.
From the absorption spectra of HOCl and OCl
(Fig. 3), the
molar absorption coefficient of HOCl at 254 nm is different
from that of OCl
. In addition, the quantum yields of HOCl
and OCl
and their tendency to change along with their con-
centrations are different from each other. Assuming that the
photolysis of HOCl and OCl
are two independent processes
and that the interactions between these two processes are neg-
ligible, the overall quantum yield of free chlorine (HOCl and
OCl
) decomposed by UV irradiation at 254 nm at a given pH
value may be estimated by the following equation:
[13]
overall
=
HOCl
HOCl
+ (1 f)ε
OCl
HOCl
+
(1 f)ε
OCl
HOCl
+ (1 f)ε
OCl
OCl
where ε
HOCl
, is the molar absorption coefficient of HOCl at
254 nm; ε
OCl
is the molar absorption coefficient of OCl
at 254 nm; f is the fraction of HOCl present in free chlorine
samples;
HOCl
is the quantum yield of HOCl;
OCl
is the
quantum yield of OCl
;
overall
is the overall quantum yield of
aqueous chlorine at a given pH condition.
In the above equation, it needs to be emphasized that the
quantum yield of HOCl is not a constant. As discussed in the
above section, it is about 1.0 ± 0.1 when the concentration is
less than 71 mg Cl/L. Otherwise, it will be a function of its
concentration as shown in Fig. 4(b). Therefore, when eq. [13]
is used for the prediction of quantum yield of free chlorine
decomposed, the quantum yield of HOCl corresponding to its
concentration should be employed. In addition, the parameters
(molar absorption coefficient, quantum yield, etc.) were deter-
mined at 254 nm, so this model should only be used for the
condition of low pressure UV lamps.
Validation of the quantum yield model
As discussed in the above section, the model (eq. [13]) was
developed based on the assumption of no interactions between
the photolysis processes of HOCl and OCl
. Because only a
unique component, HOCl or OCl
, is present in free chlorine
samples at pH 5 or 10, the assumption of the model is met
under these critical conditions and, obviously, the model can
work very well for the quantum yield prediction. However, at
intermediate pH values, free chlorine samples always consist
of both HOCl and OCl
. Therefore, several experiments were
performed in order to examine if the model can be applied at
different pH values. Figures 6(a)–6(d) show that the quantum
yields predicted by the model at various pH values are con-
sistent with those measured by the experiments. These figures
also indicate that the assumptions of the model are essentially
© 2007 NRC Canada
282 J. Environ. Eng. Sci. Vol. 6, 2007
Fig. 6. Comparison of experimental results and data calculated by the model at various concentrations: (a) 71 mg Cl/L;
(b) 140 mg Cl/L; (c) 284 mg Cl/L; and (d) 670 mg Cl/L (temperature 21 ± 2
C).
0.0
0.4
0.8
1.2
1.6
2.0
2.4
(a)
(b)
pH
5678910
Experimental Plot
Calculated by Model
(d)
5678910
0.0
0.5
1.0
1.5
2.0
2.5
3.0
(c)
Quantum yield
correct. Hence, this model should be appropriate to estimate the
overall photolysis quantum yield of free chlorine under various
pH conditions.
Effect of water quality on the quantum yield
It should be stressed that the quantum yields discussed in
the above paragraphs were determined in DI water. Actually,
the drinking water quality varies greatly from place to place
and because the strong oxidant, hydroxyl radical (·OH), is pro-
duced in the photolysis of free chlorine, the materials dissolved
in drinking water may have an important effect on the quantum
yield. In this research, methanol, simple organic compound to
model dissolved organic matter, was added to DI water. The
quantum yields of free chlorine were then studied as a func-
tion of the concentration of methanol. Figure 7 shows that the
amount of free chlorine (HOCl) decomposed depends linearly
on the concentration of methanol at pH 5, such that the quan-
tum yield is over 50 for a methanol concentration of 120 mM.
Clearly at pH 5, the product of the oxidation of methanol by ·OH
(the ·CH
2
OH radical) initiates a chain reaction involving HOCl
to cause the increase in the quantum yield. By contrast, the
quantum yields of OCl
obtained at pH 10 are always around
1.2 ± 0.2 when the methanol concentration increases from 20
to 50 mM. Compared with the quantum yield (0.9 ± 0.1) ob-
tained in DI water at pH 10, these results indicate that methanol
has little impact on the decomposition of OCl
. In practice, the
quantum yield of free chlorine could be affected by dissolved
organic matter in the drinking water and more chlorine could
be decomposed at low pH rather than at high pH.
Fig. 7. Dependence of quantum yield of free chlorine (HOCl) on
the concentration of methanol at pH 5 (temperature 21 ± 2
C).
0 20 40 60 80 100 120 140
0
10
20
30
40
50
60
70
Experimental points
Regr. of the experimental data
95% Confidence intervals
Concentration of methanol (mM)
Quantum yield
Dependence of free chlorine decomposition on the
fluence
When the concentration of free chlorine is not very high
(<70 mg Cl/L), the quantum yield of free chlorine can be as-
sumed to be independent of the concentration of free chlorine.
Therefore, the rate of free chlorine destruction follows a first
order kinetics and usually can be expressed as
ln
[C]
o
[C]
t
= k
1
t[14a]
© 2007 NRC Canada
Feng et al. 283
where [C]
o
and [C]
t
are the free chlorine concentration before
and after exposure to UV light for a time of t min; k is the
first-order constant in units of min
1
. Because the time-based
first-order rate constant (k) can be affected by the experimental
conditions (e.g., irradiance, absorbance, path length, etc.) and
is difficult to reproduce, eq. [14a] can be modified to
ln
[C]
o
[C]
F
= k
1
F[14b]
where [C]
o
and [C]
F
are the chlorine species concentration
before and after being exposed to the UV fluence of F ; k
1
is
the fluence-based first-order rate constant (Bolton and Stefan
2003). Here k
1
only depends on the quantum yield and molar
absorption coefficient of the UV light absorber (see eq. [15]).
So, theoretically k
1
is a constant for a certain UV light absorber
and can be easily reproduced.
[15] k
1
=
ε ln(10)
10U
λ
where ε is the molar absorption coefficient (M
1
cm
1
)of
the UV light absorber and other terms are defined as above.
According to eq. [14b], one can plot the logarithm (base e)
of the free chlorine reduction vs. fluence absorbed to get the
fluence-based first-order constant k
. Then, the quantum yield
can be calculated by eq. [15].
As an example, Figures 8(a) and 8(b) show that the loga-
rithmic (base e) reduction of free chlorine increases linearly
with fluence and the fluence-based first-order rate constant k
is
2.9 × 10
5
m
2
J
1
in both cases, when the concentration of free
chlorine is about 20 mg Cl/L. Based on the fluence-based rate
constant, the quantum yields estimated according to eq. [15]
were about 1.0. This result is very close to the value obtained
by the method given in the above section on “quantum yield”.
Figures 8a and 8b also indicate that only a slight amount (1%)
of free chlorine was decomposed when the concentration of
free chlorine was about 20 mg Cl/L, and the UV fluence was
400 J/m
2
, a fluence that is often specified in UV disinfection
regulations. Because DI water was used in the research, it is pos-
sible that a larger fraction of free chlorine could be destroyed
in drinking water as discussed in the above section. Generally,
the destruction of free chlorine depends not only on the fluence,
but also on the concentration of free chlorine, water quality, and
the pH. Data on this destruction are presented in Table 2.
Conclusions
The molar absorption coefficients of HOCl and OCl
at
254 nmobtained in this researchare 59 ± 1and 66 ± 1M
1
cm
1
,
respectively. These results are very consistent with other re-
ported results (Morris 1966; Chen 1967; and Thomsen et al.
2001). In DI water, the concentration of free chlorine and pH are
important factors affecting the photolysis of free chlorine sam-
ples. At low concentrations (<71 mg Cl/L), the quantum yields
of HOCl and OCl
are approximately constant at
1.0 ± 0.1 and 0.9 ± 0.1, respectively, as obtained in this re-
search. However, at higher concentrations (>71 mg Cl/L) the
Fig. 8. Logarithmic (base e) reduction of free chlorine as a
function of UV fluence (UV dose): (a) 71 mg Cl/L free chlorine
at pH 5; (b) 71 mg Cl/L free chlorine at pH 10.
0.00
0.04
0.08
0.12
0.16
0.20
0.24
y = 2.85E-05*x
R
2
= 0.998
0 2000 4000 6000 8000
0.00
0.04
0.08
0.12
0.16
0.20
0.24
Experimental points
Regr. of the experimental data
y = 2.86E-05*x
R
2
= 0.999
Photodecomposition of free chlorine (–ln(C/C ))
0
(a)
(b)
Fluence (UV dose) (J m
-2
(
Table 2. Percent of free chlorine destructed at the fluence of
400Jm
2
.
Characteristics of water sample
Percent of chlorine
species reduced
71 mg Cl/L free chlorine in DI water at
pH 5
1.1%
71 mg Cl/L free chlorine in DI water at
pH 10
1.1%
213 mg Cl/L free chlorine in DI water at
pH 5
2.8%
213 mg Cl/L free chlorine in DI water at
pH 10
1.2%
213 mg Cl/L free chlorine and 1.6 g/L
methanol in DI water at pH 10
1.3%
213 mg Cl/L free chlorine and 1.1 g/L
methanol in DI water at pH 5
18%
213 mg Cl/L free chlorine and 0.7 g/L
1,4-dioxane in DI water at pH 5
11%
quantum yield of HOCl increases linearly with the concentra-
tion at a rate of 0.0025 (mg Cl/L)
1
. Compared with HOCl, it
was found that the concentration has an insignificant effect on
the quantum yield of OCl
over the investigated concentration
range of 3.5 to 640 mg Cl/L in this research.A model (eq. [13])
© 2007 NRC Canada
284 J. Environ. Eng. Sci. Vol. 6, 2007
was established for the prediction of overall quantum yield of
free chlorine decomposed by UV irradiation at 254 nm in DI
water. Compared with the experimental results measured at var-
ious pH values, the predictions calculated by this model agree
quite well with the measured quantum yields. In addition, the
effect of water quality is very significant on the quantum yield
of HOCl, but not on the quantum yield of OCl
. The addition
of methanol at pH 5 can cause a large increase in the quantum
yield of HOCl; whereas, at pH 10, the effect of methanol addi-
tion was minimal. The decomposition of free chlorine depends
on the pH, concentration of free chlorine, fluence, and water
quality. At fluence of 400 J/m
2
, the decomposition of free chlo-
rine is very slight (1%) in DI water when the concentration
of free chlorine is not very high (20 mg Cl/L).
Acknowledgements
The authors wish to express their appreciation to Natural
Sciences and Engineering Research Council of Canada for the
funding of this research. We also thank Dr. Steve Stanley of
EPCOR Utilities Inc. for his kind support of this Project.
References
APHA, AWWA, and WEF. 1995. Standard methods for the examina-
tion of water and wastewater. 19th ed. DPD colorimetric method.
American Public Health Association,Washington, D.C. 4500-Cl.
Bolton, J.R., and Linden, K.G. 2003. Standardization of methods for
fluence (UV dose) determination in bench-scale UV experiments.
J. Environ. Eng. 129: 209–216.
Bolton, J.R., and Stefan, M.I. 2003. Fundamental photochemical ap-
proach to the concepts of fluence (UV dose) and electrical energy
efficiency in photochemical degradation reactions. Res. Chem. In-
termed. 28: 857–870.
Bolton, J.R., Dussert, B., Bukhari, Z., Hargy, T., and Clancy, J.L. 1998.
Inactivation of cryptosporidium parvum by medium-pressure ultra-
violet light in finished drinking water. Proc. AWWA 1998 Annual
Conference, Dallas, TX, Vol. A. pp. 389–403.
Bukhari, Z., Hargy, T.M., Bolton, J.R., Dussert, B., and Clancy, J.L.
1999. Medium pressure UV light for oocyst inactivation. J. Am.
Water Works Assoc. 91: 86–94.
Buxton, G.V., and Subhani, M.S. 1972a. Radiation chemistry and pho-
tochemistry of oxychlorine ions: . photodecomposition of aqueous
solutions of hypochlorite ions. J. Chem. Soc., Faraday Trans. 1. 68:
958–969.
Buxton, G.V., and Subhani, M.S. 1972b. Radiation chemistry and pho-
tochemistry of oxychlorine ions: I. radiolysis of aqueous solutions
of hypochlorite and chlorite ions. J. Chem. Soc., Faraday Trans. 1.
68: 947–957.
Chen, T.H. 1967. Spectrophotometric determination of microquanti-
ties of chlorate, chlorite, hypochlorite and chloride in perchlorate.
Anal. Chem. 39: 804–813.
Clancy, J.L., Bukhari, Z., Hargy, T.M., Bolton, J.R., Dussert, B., and
Marshall, M.M. 2000. Comparison of medium- and low- pressure
ultravioletlight for inactivation of cryptosporidium parvum oocysts.
J. Am. Water Works Assoc. 92: 97–104.
EPCOR. 2004. Water treatment in Edmonton: water treat-
ment plants [online]. Available from http://www.epcor.com/
EPCOR+Companies/EPCOR+Water+Services/Treatment+Plants/
E+L+Smith/On-Line+Tour.htm [cited May 31 2004].
Morris, J.C. 1966. The acid ionization constant of HOCl from 5 to 35
.
J. Phys. Chem. 70: 3798–3805.
Nowell, L.H., and Hoigné, J. 1992a. Photolysis of aqueous chlorine
at sunlight and ultraviolet wavelengths: I. degradation rates. Water
Res. 26: 593–598.
Nowell, L.H., and Hoigné, J. 1992b. Photolysis of aqueous chlorine at
sunlight and ultraviolet wavelengths: . hydroxyl radical production.
Water Res. 26: 599–605.
Oliver, B.G., and Carey, J.H. 1977. Photochemical production of chlo-
rinated organics in aqueous solutions containing chlorine. Environ.
Sci. Technol. 11: 893–895.
Örmeci, B, Ducoste, J.J., and Linden, K.G. 2005. UV disinfection of
chlorinated water: impact on chlorine concentration and UV dose
delivery. J. Water Supply: Res. Technol. – AQUA. 54: 189–199.
Sawyer, C.N., McCarty, P.L., and Parkin, G.F. 1994. Chemistry for
environmental engineering, 4th Edition. McGraw-Hill, Inc., New
York, N.Y.
Snoeyink, V.L., and Jenkins, D. 1980. Water chemistry. John Wiley &
Sons, Inc., New York, N.Y.
Thomsen, C.L., Madsen, D., Poulsen, J.A., Thogersen, J., Jensen,
S.J.K., and Keiding, S.R. 2001. Femtosecond photolysis of aqueous
HOCl. J. Chem. Phys. 115: 9361–9369.
Zheng, M., Andrews, S.A., and Bolton, J.R. 1999. Impacts of medium-
pressure UV and UV/H
2
O
2
on THM and HAA formation in pre-
UV chlorination drinking water. [CD-ROM]. Proceedings Water
Quality Technology Conference, Tampa, Fla., 31 Oct–4Nov1999.
© 2007 NRC Canada
    • ". Time profiles of MPT byproducts in (a) deionized water, (b) surface water, (c) tap water, and (d) filtered water during the UV/chlorination reaction (experimental conditions; [MTP]0 = 100 ng/L, [Cl2] = 5 mg/L, UV–C intensity = 4.4 mW/cm 2 , pH 7, 20 • C). (Fig. 4(d)). Feng et al. [28] proposed two chain reactions related to DOM in water by OH and Cl radicals during the UV/chlorination reaction (Eqs. (14)–(17)). "
    [Show abstract] [Hide abstract] ABSTRACT: Metoprolol (MTP), a hypertension depressor, has been increasingly detected even after conventional water treatment processes. In this study, the removal of MTP was compared using chlorination (Cl2), UV-C photolysis, and UV/chlorination (Cl2/UV) reactions. The results showed that the UV/chlorination reaction was most effective for MTP removal. MTP removal during UV/chlorination reaction was optimized under various conditions of UV intensity (1.1-4.4mW/cm(2)), chlorine dose (1-5mg/L as Cl2), pH (2-9), and dissolved organic matter (DOM, 1-4mgC/L) using a two-level factorial design with 16 experimental combinations of the four factors. Among the factors examined, DOM scavenging by OH radicals was the most dominant in terms of MTP removal during UV/chlorination reaction. The established model fit well with the experimental results using to various water samples including surface waters, filtered and tap water samples. The optimized conditions (UV intensity=4.4mW/cm(2), [Cl2]=5mg/L, pH 7, and [DOM]=0.8-1.1mgC/L) of the model removed more than 78.9% of MTP for filtered water samples during UV/chlorination reaction. Using LC-MS/MS, five byproducts of MTP (molecular weight: 171, 211, 309, 313, and 341, respectively) were identified during UV/chlorination reaction. Based on this information, the MTP transformation mechanism during UV/chlorination was suggested. Our results imply that applying UV/chlorination process after filtration stage in the water treatment plant (WTP) would be the most appropriate for effective removal of MTP. Copyright © 2014 Elsevier B.V. All rights reserved.
    Full-text · Article · Dec 2014
    • "), when electrolysis are carried out at chloride 51 containing medium at specific pH conditions. Moreover, the 52 combination of the chloro oxidant species generation and UV 53 irradiation can also result in the formation high reactive species 54 [13]. The possible generation of organochlorine byproducts is the 55 main drawback; however, as showed by Malpass et al. [14], no 56 toxicity was observed in samples of the herbicide Atrazine treated 57 with the photo-assisted electrochemical method using a DSA 1 58 electrode at chloride or sulfate containing media. "
    [Show abstract] [Hide abstract] ABSTRACT: Phthalic acid esters are synthetic compounds used as additives in plastic structures. As these chemicals are physically bound, they can contaminate humans affecting the endocrine system. The present study focused on the photo-assisted electrochemical (PAE) method to degrade a phthalic acid ester, such as dimethyl phthalate–DMP, using a one-compartment filter press flow cell and a commercial DSA® electrode. The variables investigated to optimize the DMP compound removal were NaCl concentration, current density, and temperature. The efficiency towards the DMP concentration abatement and total organic carbon (TOC) removal were the analytical parameters analyzed. The highest rates of DMP and TOC removals were attained at high NaCl concentrations and current densities, due to the generation of high reactive oxidants in the PAE method that seemed to suppress mass transport limitations. The temperature had no significant influence on the DMP removal. Moreover, depending on the nature of the intermediate compounds generated, the PAE method could enable a rapid treatment option before disposal of contaminated effluents containing DMP, and possibly other endocrine disruptors. Despite the high levels of current efficiency attained, more work is required to optimize the electrical energy consumption, especially with respect to the choice of the UV lamp.
    Full-text · Article · Jun 2014
    • "Free chlorine photolysis in the UV photoreactor [30]. "
    [Show abstract] [Hide abstract] ABSTRACT: Hybridization of a UV-C photoreactor with an electrochemical cell led to the discovery of a distinctive synergistic mechanism for the degradation and mineralization of phenol when free chlorine-producing anodes were utilized in the presence of chloride. After 6 h, photochemical (UV) treatment resulted in 29% phenol mineralization and electrochemical (EL) treatment resulted in 35% and 52% mineralization using boron-doped diamond (BDD) and ruthenium oxide on titanium (DSA-Cl2) anodes, respectively. However, the photo-assisted electrochemical (UVEL) process removed 88% and 96% of total organic carbon (TOC) after 6 h with BDD and DSA-Cl2 anodes, respectively. The hybrid UVEL process generated highly reactive hydroxyl and chlorine radicals by the photolysis of in situ electrogenerated free chlorine. As a result, the UVEL treatment produced fewer chlorinated by-products compared to EL alone. UVEL with DSA-Cl2 had the highest TOC removal after 6 h (96%), one of the highest mineralization rates (0.38 h−1), and the lowest energy demand per order TOC removed (EEO = 104 kW h m−3 order−1) without any residual toxicity after dechlorination. Generally, conventional advanced oxidation processes (UV/O3 and UV/H2O2) were most effective at mineralization up to 50%, but UVEL appeared more cost effective for mineralization beyond 75%.
    Full-text · Article · Mar 2014
Show more