Degradation of a model naphthenic acid, cyclohexanoic acid, by vacuum UV (172 nm) and UV (254 nm)/H2O2

Article abstract of DOI:10.1021/jp105727s

The mechanism of hydroxyl radical initiated degradation of a typical oil sands process water (OSPW) alicyclic carboxylic acid was studied using cyclohexanoic acid (CHA) as a model compound. By use of vacuum ultraviolet irradiation (VUV, 172 nm) and ultraviolet irradiation in the presence of hydrogen peroxide UV(254 nm)/H₂O₂, it was established that CHA undergoes degradation through a peroxyl radical. In both processes the decay of the peroxyl radical leads predominantly to the formation of 4-oxo-CHA, and minor amounts of hydroxy-CHA (detected only in UV/H₂O ₂). In UV/H₂O₂, additional 4-oxo-CHA may also have been formed by direct reaction of the oxyl radical with H_-O ₂. The oxyl radical can be formed during decay of the peroxyl-CHA radical or reaction of hydroxy-CHA with hydroxyl radical. Oxo- and hydroxy-CHA further degraded to various dihydroxy-CHAs. Scission of the cyclohexane ring was also observed, on the basis of the observation of acyclic byproducts including heptadioic acid and various short-chain carboxylic acids. Overall, the hydroxyl radical induced degradation of CHA proceeded through several steps, involving more than one hydroxyl radical reaction, thus efficiency of the UV/H ₂O₂ reaction will depend on the rate of generation of hydroxyl radical throughout the process. In real applications to OSPW, concentrations of H₂O₂ will need to be carefully optimized and the environmental fate and effects of the various degradation products of naphthenic acids considered.

Full text of DOI:10.1021/jp105727s

J. Phys. Chem. A 2010, 114, 12067–12074
12067
Degradation of a Model Naphthenic Acid, Cyclohexanoic Acid, by Vacuum UV (172 nm)
and UV (254 nm)/H2O2
Przemysław Drzewicz, Atefeh Afzal, Mohamed Gamal El-Din,* and Jonathan W. Martin
Department of CiVil and EnVironmental Engineering, 3-133 Markin/CNRL Natural Resources Engineering
Facility, UniVersity of Alberta, Edmonton, AB, T6G 2W2, Canada, and Department of Lab Medicine and
Pathology, DiVision of Analytical and EnVironmental Toxicology, Faculty of Medicine and Dentistry, UniVersity
of Alberta, Edmonton, AB, T6G 2G3, Canada
ReceiVed: June 21, 2010; ReVised Manuscript ReceiVed: September 22, 2010
The mechanism of hydroxyl radical initiated degradation of a typical oil sands process water (OSPW) alicyclic
carboxylic acid was studied using cyclohexanoic acid (CHA) as a model compound. By use of vacuum
ultraviolet irradiation (VUV, 172 nm) and ultraviolet irradiation in the presence of hydrogen peroxide UV(254
nm)/H2O2, it was established that CHA undergoes degradation through a peroxyl radical. In both processes
the decay of the peroxyl radical leads predominantly to the formation of 4-oxo-CHA, and minor amounts of
hydroxy-CHA (detected only in UV/H2O2). In UV/H2O2, additional 4-oxo-CHA may also have been formed
by direct reaction of the oxyl radical with H2O2. The oxyl radical can be formed during decay of the peroxyl-
CHA radical or reaction of hydroxy-CHA with hydroxyl radical. Oxo- and hydroxy-CHA further degraded
to various dihydroxy-CHAs. Scission of the cyclohexane ring was also observed, on the basis of the observation
of acyclic byproducts including heptadioic acid and various short-chain carboxylic acids. Overall, the hydroxyl
radical induced degradation of CHA proceeded through several steps, involving more than one hydroxyl
radical reaction, thus efficiency of the UV/H2O2 reaction will depend on the rate of generation of hy-
droxyl radical throughout the process. In real applications to OSPW, concentrations of H2O2 will need to be
carefully optimized and the environmental fate and effects of the various degradation products of naphthenic
acids considered.
1. Introduction
The mechanism of hydroxyl radical reaction with saturated
alicyclic carboxylic acids (i.e., NAs) has not been examined,
but the most common reaction of hydroxyl radical with saturated
aliphatic carboxylic acids is hydrogen abstraction from the
aliphatic chain.8 On the basis of established reaction rate
constants with aliphatic carboxylic acids in basic solution, it
was observed that hydroxyl radical preferentially abstracts
hydrogen from tertiary carbons, whereas the least reactive
hydrogens toward hydroxyl radical are those on primary carbon
atoms.8 Under acidic conditions, the deactivating effect of the
carboxylic acid group decreased from tertiary to primary carbon
in R position,9 and saturated aliphatic carboxylic acids were
oxidized principally at the ꢀ-carbon and the penultimate position.
Comparatively, under basic conditions the carboxylate group
increases the opportunity for delocalization of the unpaired
electron, thus favoring abstraction of hydrogen from the
R-carbon due to resonance stabilization of the resulting R-carbon
centered radical, albeit oxidation of carboxylic acids at the
penultimate position was also observed under basic conditions
as well; however, the oxidation at the penultimate position may
have been induced by the TiCl3/H2O2 system used for generation
of hydroxyl radical.9
Although there is no information about hydroxyl radical
degradation of alicyclic carboxylic acids to date, the mechanisms
of degradation of other organic compound by hydroxyl radical
in aqueous solution may be relevant to the mechanism of
decomposition. The general mechanism of hydroxyl radical
induced degradation of organic compounds was proposed by
von Sontang and Schuchmann.10 Generally, the hydroxyl radicals
react with an organic compound to produce an organic radical
which, in turn, reacts with molecular oxygen to form a peroxyl
Canada’s oil sands deposits represent one of the world’s
largest oil resources, representing an estimated 174 billion
barrels of oil in form of bitumen.1 For every barrel of oil
produced by the oil sands surface mining industry, more than 2
barrels of fresh water are used to extract the bitumen using an
alkaline hot water process.2 The resulting oil sands process-
affected water (OSPW) is recycled to reduce overall fresh water
use, but the result is that OSPW becomes highly saline and
acutely toxic to aquatic organisms, primarily due to high
concentrations of naphthenic acids (NAs). NAs are a naturally
occurring complex mixture of organic acids in bitumen and are
largely composed of saturated alicyclic carboxylic acids.3 They
are persistent in OSPW, with overall degradation half-lives
exceeding 13 years.4 With an expanding oil sands industry in
Northern Alberta, and increasing volumes of OSPW which must
eventually be reclaimed, the effective remediation of OSPW
has been a target of many studies since the 1980s.5
Naphthenic acids do not undergo significantly photolysis by
direct ultraviolet (UV) or solar light due to low absorption.6,7
In other produced waters, advanced oxidation processes (AOPs)
have been extensively used for removal of refractory or toxic
organic and inorganic pollutants, but to date the use of hydroxyl
radical for treatment of OSPW has received little attention. In
AOPs, hydroxyl radicals are generated in situ to react with
organic compounds in an effort to produce less toxic, or less
persistent, intermediates.
* To whom correspondence should be addressed. Phone: (780) 492-5124.
Fax: (780) 492-8198. E-mail: mgamalel-din@ualberta.ca.
10.1021/jp105727s  2010 American Chemical Society
Published on Web 10/25/2010

12068 J. Phys. Chem. A, Vol. 114, No. 45, 2010
Drzewicz et al.
radical. Oxygen is normally present in solution when advanced
oxidation processes such as ultraviolet irradiation in presence
of hydrogen peroxide (UV/H2O2), ozonation (O3), Fenton
processes (Fe(II)/H2O2), or their combination are applied for
decomposition of organic pollutants. Peroxyl radicals may
combine, forming short-lived tetroxide species (reaction 1)
which decomposes via a number of pathways (reactions 2-4)10
It is well known that H2O2 may react with organic radicals,
inducing secondary reactions.17 For example, for 2-propanol it
was found that the intermediate organic radicals reacted with
hydrogen peroxide to form acetone18
(CH3)2CHOH + •OH f (CH3)2•COH + H2O (11)
2RO•2 h RO4R
(1)
(CH3)•2COH + H2O2 f (CH3)2CO + •OH + H2O
(12)
Hydroxyl radicals are also generated by UV irradiation of
H2O2. In the UV/H2O2 processes, H2O2 is added in excess to
generate hydroxyl radicals over the entire process time.19 The
high concentrations of H2O2 in the UV/H2O2 process are more
likely to play important roles in the decomposition of organic
compounds and affect formation of byproduct than in the VUV
process.
In this study, cyclohexanoic acid (CHA) was selected as a
simple model NA compound. Although there are relatively few
NAs with less than 9 carbons in OSPW, NAs containing one
ring are a prominent class. The degradation of CHA was studied
by VUV and UV/H2O2. This allowed the mechanism of
hydroxyl radical induced decomposition of CHA to be inves-
tigated, based on identification of formed byproduct in both
systems, and the possible role of H2O2 in the decomposition
mechanism was evaluated by comparing byproducts and reaction
rate constants in the two systems.
The exact mechanism of these reactions is still controversial.
There is a dispute concerning whether the product-forming
process is sequential or concerted; however, the occurrence of
the reactions and formation of these products was reported by
many authors.10
In most studies on hydroxyl radical induced degradation of
organic compounds, reviewed by von Sontang and Schuch-
mann,10 radiolysis of water is used as a source of hydroxyl
radical. However, photolysis of water with light below 190 nm
can be also used as a convenient source of hydroxyl radical11
because the UV absorption of water steadily increases from
approximately 190 to 120 nm,12 and its electronic excitation
leads to its photolysis with the production of hydrogen atoms
and hydroxyl radicals
2. Experimental Section
2.1. Materials. CHA, cis-4-hydroxycyclohexanoic acid (cis-
4-hydroxy-CHA), and trans-4-hydroxycyclohexanoic acid (trans-
4-hydroxy-CHA) were of analytical grade purity (99+%) and
were purchased from TCI (Portland, OR, USA). All analyte
solutions were prepared in ultrapure water (resistivity ) 18 MΩ
× cm and TOC < 0.1 mg/L) prepared with a Millipore RO-
Milli-Q system equipped with a UV lamp, adjusted to pH 9
with carbonate free 50% sodium hydroxide (ACS purity).
Hydrogen peroxide (ACS purity, Fisher Scientific Co., Canada)
solutions were prepared from 30% w/w stock solution. Am-
monium acetate and methanol were purchased from Fisher
Scientific (LC-MS grade).
H2O + hν(VUV) f H• + •OH
(5)
The Xe-excimer lamp emission domain is 172 ( 14 nm.13
At 172 nm, the quantum yield for production of hydroxyl radical
is ΦHO ) 0.42;13 the quantum yield for the production of
•
hydrated electrons is lower than 0.05. Hydrogen atoms can be
quenched by molecular oxygen which is not reactive toward
hydroxyl radical9
2.2. VUV Irradiation. A dielectric barrier discharge Xe2*
excilamp (Model BD-P-barrier discharge portable, Institute of
High Current Electronics, Tomsk, Russia20) with dimensions
26.5 × 7.7 × 77 cm (length × width × height) including a
cooling fan was used for the VUV study. The lamp was placed
in a collimated beam apparatus designed specifically for VUV
experiments.21 During irradiation, the lamp was cooled with
gaseous nitrogen (100 mL/min), also functioning to prevent
formation of ozone from photolytic excited oxygen.10 Because
of low penetration of 172 nm light emitted by the lamp (less
than 70 µm in water22), the reactor was a 3 mL Suprasil quartz
cuvette with a stopper and a small magnetic stir bar. An
actinometry method, using methanol, was employed for calcu-
lating the fluence rate (irradiance) of the xenon excilamp at 172
H• + O2 f HO•2
(6)
•
The formed perhydroxyl radicals (HO2 ) are much less
reactive toward organic compounds than hydroxyl radical and
can decay by a recombination reaction to form hydrogen
peroxide (reaction 7) or water and oxygen (reaction 8)14
HO•2 + HO•2 f H2O2
(7)
(8)
HO•2 + OH f H2O + O2
•
•
nm23 to be 0.67 ( 5% mW/cm2, and the yield of OH was 1.4
In vacuum UV (VUV) photolysis of water, H2O2 concentra-
tions may build-up with very long irradiation times (reaction
7),14,15 or by decomposition of organic peroxyl radicals (reaction
3).10,16 However, the concentration of formed hydrogen peroxide
in these cases is very low and will generally not interfere in the
reaction process.
× 10-9 mol s-1 (Φ ) 0.42, photon flow )3.34 × 10-9 mol
s-1).21 The naturally dissolved oxygen was not removed from
solutions because oxygen quenched the formation of H•.
Contrary to γ-radiolysis, N2O cannot be added to scavenge H•
in a VUV system because N2O undergoes photolysis to N2 and
singlet oxygen.24,25

Degradation of a Model Naphthenic Acid
J. Phys. Chem. A, Vol. 114, No. 45, 2010 12069
TABLE 1: Chemical Formulae and Proposed Structure(s)
of Major By-Products Formed by VUV and UV/H2O2 at m/z
141 and 159
2.3. UV Irradiation in Presence of H2O2. The UV/H2O2
experiments were carried out using a collimated beam
apparatus from Calgon Carbon Corporation (Pittsburgh, PA,
USA). A 10-W low pressure UV lamp (Calgon Carbon,
Pittsburgh, PA, USA) was used for the UV/H2O2 experiments.
The main light emission of the low pressure lamp is at 254
nm. A radiometer (International Light Inc. Model IL 1400A)
equipped with a UV detector (International Light Inc. Model
SED240) and a neutral density filter (Model QNDS2),
calibrated at 5 nm intervals in the range of 200-400 nm,
was used to measure the irradiance at the surface of the water
samples. The average fluence rate in the solution and the
delivered fluence (UV dose) were calculated based on the
Bolton and Linden protocol26 and the spreadsheets available
at www.iuva.org for low pressure UV lamps for shallow (<2
cm) solutions. The fluence rate of the lamp was calculated
to be 13.69 mW/cm2 and the yield of •OH generation
8.2 × 10-6 mol s-1 (Φ ) 1, Petri dish area ) 28.26 cm2,
radiant energy of 1 mol of 254 nm photon ) 470.971 kJ
mol-1).
2.4. Analytical Methods. The analytical method was de-
signed for separation of CHA and two hypothesized degradation
products, cis-4- and trans-4-hydroxycyclohexanoic acid. Ana-
lytical high-performance liquid chromatography (HPLC)
columns (Luna C8, 5 µm, 150 × 3 mm, and 250 × 3 mm;
Phenomenex) were used for reversed-phase chromatography.
For routine analysis, an HPLC coupled to an ion trap mass
spectrometer (Varian 500-MS) with unit mass resolution was
used. For confirmation of empirical formulas, an HPLC coupled
with a time-of-flight mass spectrometer (Agilent 6220, 10,000
mass resolution) was used. Both mass spectrometers were
equipped with an electrospray interface operating in negative
ion mode. Chromatography was performed at 40 °C at a flow
rate 200 µL/min, and injection volumes were 20 µL. The mobile
phase consisted of 100% methanol and 4 mM ammonium
acetate with 0.1% acetic acid in aqueous solution. The concen-
tration of methanol was ramped from 40 to 80% over 20 min.
Chromatograms of carboxylic acid standards showed that
retention times were generally in order of decreasing polarity.
To achieve better resolution of more polar byproduct and their
isomers, a longer column (250 mm, otherwise identical to above)
was also applied. The same mobile phases were used, but the
initial concentration of methanol was 30%, and after 10 min it
was slowly ramped to 70% over 30 min. Heptanoic and octanoic
acid both eluted after 80 min; thus 80 min was the total run
time.
Ion chromatography with conductivity detection was used
to detect any short-chain carboxylic acids. Chromatography
was performed at 40 °C and injection volumes were 200 µL.
The mobile phase consisted sodium hydroxide solution at
flow rate 1 mL/min. The concentration of NaOH was ramped
from 2.5 to 5 mM over 15 min and then from 5 mM to 38.25
mM. The electrochemical suppressor current was 100 mA.
The AS11-HC (4 mm ×250 mm) with a guard column AG-
11 was used (Dionex, Sunnyvale, CA, USA). To remove
carbonate residue from the eluent, an anion trap column
(ATC-HC, Dionex, Sunnyvale, CA, USA) was used.
of concentrations used here, the kinetics of degradation were
examined. Direct photolysis follows zero-order kinetics,
whereas degradation processes based on generation of
hydroxyl radical should follow pseudofirst order kinetics.11,14,24
It was determined that at up to 20 mg/L of CHA (highest
concentration used in the experiments) in aerated aqueous
solutions, the degradation followed pseudo-first-order kinetic
decay. Under these conditions, the hydroxyl radicals are
present in solution in excess, and the process is independent
•
of the rate of OH generation from photolysis of water.
Furthermore, we observed that the addition of 80 µM tert-
butanol, a hydroxyl radical scavenger, significantly decreased
the effectiveness of CHA decomposition (40 µM). Taken
together, these results indicated that decomposition of CHA
•
in the investigated range was induced primarily by OH.
At an initial concentration of 20 mg/L CHA, 90%
degradation was achieved after 15 total minutes of irradiation.
In a study of the reaction products, HPLC high resolution
mass spectrometry indicated two major byproduct at m/z )
159.0737 and 141.06287 (the calculated m/z of CHA is
128.1690) (Table 1). These same products were also found
using the ion trap in autoscanning MS/MS mode, and we
examined the product ions of m/z ) 159.0737 and 141.06287
for structural elucidation, but the m/z 159 signal intensity
was too low to collect MS/MS spectra. However, retention
times of the m/z 159 degradation products identified here were
the same as those observed during UV/H2O2; thus MS/MS
spectra of m/z 159 were collected on the UV/H2O2 samples
(discussed in the next section).
The ion at m/z ) 141.06287 is characteristic of an oxo-
cyclohexanoic acid (oxo-CHA, Table 1), and in MS/MS the
product ion at m/z ) 97, representing a neutral loss of 44,
suggested the presence of a carboxylate group (Figure 1). The
formation of such a product suggested that after hydrogen
abstraction by hydroxyl radical (reaction 13) the resulting radical
reacted with oxygen to form peroxyl-CHA radical (reaction 14)
which may subsequently decay by recombination reactions
(reactions 15-17).
3. Decomposition of CHA by VUV
The study of hydroxyl radical decomposition of organic
compounds using VUV must be conducted at relatively low
concentration. This is because, at high concentration, direct
photolysis of organic compounds can interfere.6 To ensure
that direct photolysis of CHA was not occurring in the range

12070 J. Phys. Chem. A, Vol. 114, No. 45, 2010
Drzewicz et al.
Figure 1. Confirmed formation of 4-oxo-CHA (A) after 5 min VUV treatment and (B) after 60 min UV/H2O2 treatment of (red line) CHA and
(blue line) trans-4-hydroxyCHA, respectively. Chromatograms are for MS/MS transition m/z 141/97, characteristic of neutral loss of CO2 from
4-oxo-CHA.
Our further study of the VUV-induced degradation of trans-
4-hydroxy-CHA also showed formation of a byproduct at m/z
141 (Figure 1). In fact, this byproduct eluted at the same
retention time as the aforementioned oxo-CHA byproduct of
CHA degradation and produced the same m/z 97 product ion
in MS/MS. Therefore, both CHA and trans-4-hydroxy-CHA
The possible radical product formed in reaction 17 may have
produced a major oxo-CHA byproduct, and it is likely that the
undergone further peroxidation as the oxyl radical can undergo
VUV CHA byproduct was also 4-oxo-CHA. trans-4-Hydroxy-
R hydrogen shift (reaction 18) and then react with oxygen to
CHA was likely decomposed to oxo-CHA through reactions
form a peroxyl radical (reaction 19), which could decompose
19-21. Therefore, it is also likely that any 4-hydroxy-CHA
to oxo-CHA (reaction 20) as observed in the analogous
produced from CHA (i.e., as in reaction 15) would have been
decomposition of secondary alcohols.27 The decay of this
subsequently converted to 4-oxo-CHA. The rate constants of
peroxyl radical may also be induced by hydroxide anion
reaction for cis- and trans-4-hydroxy-CHA with hydroxyl radical
(reaction 21)27
(Table 2) are comparable to the literature rate constant of CHA
(k ) 5.5 × 109 L mol-1 s-1 at pH 98). It is also notable that
there was a statistically significant difference between the
hydroxyl radical rate constants of cis- and trans-4-hydroxy-
CHA. Unlike 4-oxo-CHA, which can be formed by all three
proposed degradation pathways of the peroxyl-CHA radical
(reactions 15-17), hydroxy-CHA can only be formed in
(reaction 15). Taken together, this may explain no detectable
yield of hydroxy-CHA from CHA and in these VUV experiments.

Degradation of a Model Naphthenic Acid
J. Phys. Chem. A, Vol. 114, No. 45, 2010 12071
TABLE 2: Comparison of Rate Constants Measured in VUV and UV/H2O2 at pH 9 Using Competitive Kinetics and
3
Cyclohexanoic Acid as the Reference (k ) 5.5 × 109 (L mol-1 s-1
)
rate constant (L mol-1 s-1), n ) 3
compound
VUV
UV/H2O2
4-trans-hydroxycyclohexanoic acid
4-cis-hydroxycyclohexanoic acid
5.5 ( (0.03) × 109
2.4 ( (0.42) × 109
4.1 ( (0.12) × 109
3.5 ( (0.35) × 109
5. Decomposition of CHA by UV/H2O2
be formed during CHA degradation by UV/H2O2 could also
undergo degradation in a mechanism involving reaction with
H2O2 (reaction 22). As in VUV, this may be a reason for no
detectable yield of hydroxy-CHA by UV/H2O2. Additionally,
our relative rate measurements showed that apparent rate
constants for the reaction of 4-hydroxy CHAs (cis and trans)
with hydroxyl radical were lower than those measured by VUV
(Table 2), thus indicating that H2O2 may play a role in the
decomposition mechanism of CHA at the concentrations
employed here. This is because the direct reaction of H2O2 with
organic compound decreases the number of hydroxyl radicals
generated in the UV/H2O2 process. It is germane to note that,
as was the case in VUV, there was a statistically significant
difference between the degradation rate constants for cis- and
trans-4-hydroxy-CHA by UV/H2O2 (Table 2).
The efficiency of the UV/H2O2 process depends on the
concentration of hydrogen peroxide, and two hydroxyl radicals
are formed during photolysis of H2O2.11 However, because of
the reaction of the produced hydroxyl radicals with H2O2, the
effective amount of hydroxyl radicals produced in the process
is lower than predicted by simple photolysis of H2O2.28
Additionally, H2O2 may also react directly with organic byprod-
uct, thus further decreasing the efficiency of hydroxyl radical
formation in the processes.17,29 Thus, it is necessary to optimize
the added amount of H2O2 in each application. In our system,
the highest efficiency of CHA degradation was obtained
when the molar ratio of H2O2:CHA was 23. The optimum ratio
was consistent for 1, 10, and 80 mg/L of CHA. It was also
confirmed in control experiments that CHA and hydroxy-CHA
did not react directly with H2O2. Furthermore, the absorption
spectra of these compounds showed absorption only below 210
nm, and compounds did not undergo photolysis at 254 nm.
Despite a higher rate of formation of hydroxyl radicals than
in VUV, 90% degradation of 20 mg/L CHA by UV/H2O2 was
only achieved after 80 min of treatment. The main byproduct,
as in VUV, appeared to be 4-oxo-CHA based on retention time
and MS/MS product ions (Figure 1). In the UV/H2O2 process,
4-oxo-CHA may be formed in reactions of H2O2 with organic
radicals. The oxyl radical of CHA formed in reactions 17 and
18 may react with H2O2 (reaction 22), as was observed in the
UV/H2O2 oxidation of 2-propanol18
At 20 mg/L of CHA or trans-4-hydroxy-CHA, no other
products (other than 4-oxo-CHA) were observed; thus a 50 mg/L
initial CHA concentration was used for product analysis, and
several byproducts were identified by MS/MS at m/z 159. These
were assigned as isomers of dihydroxy-CHA due to character-
istic m/z 97 and 141 product ions (Figure 2). The 97 m/z product
ion (neutral loss of 62, dihydroxyethane) may indicate com-
pounds containing vicinal hydroxyl groups or two hydroxyl
groups on the same carbon. Those products could be formed
by addition of hydroxyl radical to the hydroxy-CHA radical
formed in reaction 18 or addition of hydroxyl radical to the
double bond of enol tautomers of 4-oxo-CHA. The m/z 141
product (indicating neutral loss of 18, or H2O) may indicate
compounds containing one or more isolated hydroxyl groups.
Neutral loss of 18 was also observed in the MS/MS spectra of
our 4-hydroxy-CHA standards (m/z 143 > 125 transition). The
structure of the various dihydroxy-CHA isomers may be
elucidated, in part, by retention time. On the basis of analogy
to dihydroxybenzoic acids isomers separated by reverse-phase
chromatography,30 it is suggested that less polar dihydroxycy-
clohexanoic acids containing vicinal hydroxyl groups eluted later
than other dihydroxy isomers.
Study of the decomposition of 20 mg/L trans-4-hydroxy-CHA
by UV/H2O2 also showed formation of 4-oxo-CHA during the
process (Figure 1). This indicated that hydroxy-CHAs that may
Figure 2. Chromatograms confirming formation of byproduct with m/z 159 during UV/H2O2 decomposition of (A) 50 mg/L CHA and (B) 50 mg/L
4-trans-hydroxy CHA after 60 min. Transition 159 > 113 is characteristic of heptadioc acids, and transitions 159 > 141 and 159 > 97 are characteristic
of dihydroxy CHAs.

12072 J. Phys. Chem. A, Vol. 114, No. 45, 2010
Drzewicz et al.
Figure 3. UV/H2O2 degradation of 50 mg/L CHA (right axis, relative units, marked as 9) and formation of byproduct (left axis, peak area units):
(A) dihydroxy-CHAs and heptadioic acid, (B) oxo-CHAs, (C) hydroxy-CHAs, (D) total short-chain carboxylic acids by ion chromatography.
The m/z 159 > 113 transition in treated CHA chromatograms
indicated neutral loss of 46 (formic acid), which may also
suggest the presence of heptadioic acids as products. A proposed
explanation for formation of heptadioc acid may be reaction of
the oxyl radical with hydroxyl radical (reaction 23). The
formation of m/z 159 products over the process time is presented
in Figure 3.
cis-4-hydroxycyclohexanoic acid, respectively, Figure 4). The
production of these hydroxy-CHA products over the process
time is presented in Figure 3.
Formation of byproduct at m/z 159 was also observed in the
decomposition of 4-hydroxy-CHA. The m/z 159 > 141 transition
and neutral loss of 18 was characteristic of the aforementioned
dihydroxy-CHA isomers observed in the degradation of CHA.
However, unlike with CHA, here there were no dihydroxy
isomers at later retention times having the m/z 159 > 97
transition, characteristic of vicinal dihydroxy-CHA isomers. This
indicates that one of the pathways for degradation of 4-hydroxy-
CHA is hydroxylation of the hexane ring, presumably at carbon
position 1 or 2. Additionally, the early eluting compound having
the m/z 159 > 113 transition was found at the same retention
time here as in CHA degradation experiments, likely isomers
of heptadioc acid formed through reaction 23.
Other hydroxy-CHA isomers (i.e., other than 4-hydroxy CHA)
were also identified among the products of 50 mg/L CHA after
80 min of treatment. This was based on products with an m/z
143 > 125 transition (i.e., neutral loss of 18 or H2O). Seven
total hydroxy-CHA isomers were observed, corresponding to
all the possible isomers (1-hydroxy, and the cis and trans
isomers of 2, 3, and 4-hydroxy-CHA), but position 4 was the
most preferable based on peak area of peak 1 and 4 (trans- and
The formation of short-chain aliphatic carboxylic acids
containing from one to four carbons was observed by ion
chromatography. These could not be clearly resolved under the
chromatographic conditions, and complex chromatograms were
resultant, likely representing coeluting short-chain acids with
their respective oxo- or hydroxy- derivatives. The concentration

Degradation of a Model Naphthenic Acid
J. Phys. Chem. A, Vol. 114, No. 45, 2010 12073
SCHEME 1: Proposed Mechanism of Hydroxyl Radical
Induced Decomposition of CHA
Figure 4. Formation of hydroxy-CHA isomers during UV/H2O2 decomposi-
tion of 50 mg/L CHA. Chromatogram is for MS/MS transition m/z 143. Peaks
1 and 4 are trans- and cis-4-hydroxy-CHA, respectively.
of total short-chain aliphatic carboxylic acids (measured as
summary area of peaks) was increasing over the process time
process (Figure 3). These short-chain byproducts may be formed
from an initial reaction involving ꢀ-scission of the oxyl radical
(reaction 24).9 One can be assumed that formed radical in this
reaction is very unstable and quickly undergo degradation.
Formation of hydroxyl-CHA and oxo-CHA is predominant
in the process, and scission of the hexane ring and formation
of short-chain carboxylic acids follows formation of these
compounds. Additionally, concentration of heptadioic acids in
treated solution of CHA is very low contrary to treated solution
of hydroxy-CHA (Figure 2). It suggests that reactions 23 and
24 may explain formation of short-chain carboxylic acids.
6. Conclusion
The mechanism of hydroxyl radical initiated degradation of
a model OSPW alicyclic carboxylic acid revealed new infor-
mation about the products and mechanisms of degradation.
Contrary to predictions based on the reactivity of linear
saturated aliphatic acids, hydroxyl radical attack did not take
place preferentially at the R tertiary carbon position; rather
4-oxo-CHA was the major byproduct in VUV and UV/H2O2
systems, indicating that hydrogen atoms at the 4 position were
the most vulnerable toward abstraction, possibly due to steric
hindrance of the cyclohexane ring. Steric hindrance at the 4
position is also indicated by different reaction rate constants
for cis- and trans-4-hydroxy-CHA; however, this further
study with substituted CHA would confirm this hypothesis.
By comparison of VUV and UV/H2O2 processes, our results
also suggest that H2O2 can affect the reaction rate and yield
of the byproduct. In UV/H2O2, direct reaction of H2O2 with
organic radicals and organic compounds slowed down the
process.
Formation of oxo-, hydroxy-, and dixydroxy CHA isomers
was followed by formation of heptadioic and short-chain
carboxylic acids. It suggests that scission of the hexane ring
induced by reaction with hydroxyl radical is characteristic only
for the derivatives of CHA.
Degradation of CHA involves several reactions with hydroxyl
radical. Thus, presence of other constituents reacting with

12074 J. Phys. Chem. A, Vol. 114, No. 45, 2010
Drzewicz et al.
(12) Weeks, J. L.; G. M.; Meburn, A. C.; Gordon, S. Radiat. Res. 1963,
19, 559–567.
(13) Heit, G.; Braun, A. M. Water Sci. Technol. 1997, 35, 25–30.
(14) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J.
Phys. Chem. Ref. Data 1988, 17, 513–886.
hydroxyl radicals, other than naphthenic acid, in treated water
is important for efficiency of the process. Moreover, hydroxyl-
ated naphthenic acids may resemble sex steroids; thus the
endocrine activity of UV/H2O2 treated water should be con-
sidered. A proposed mechanism of the decomposition is depicted
in Scheme 1.
(15) Holroyd, R. A.; Bielski, H. J. J. Am. Chem. Soc. 1978, 100, 5796–
5800.
(16) Azrague, K.; Bonnefille, E.; Pradines, V.; Pimienta, V.; Olivieros,
E.; Maurette, M.-T. Photochem. Photobiology. Sci. 2005, 4, 406–408.
(17) Karpel vel Leitner, N.; Dore, M. J. Photochem. Photobiol. A:
Chemistry 1996, 99, 137–143.
(18) Burchill, C. E.; Jones, P. W. Can. J. Chem. 1971, 49, 4005–4016.
(19) Afzal, A.; Oppenla¨nder, T.; Bolton, J. R.; Gamal El-Din, M. Wat.
Res. 2010, 44, 278–286.
(20) Tarasenko, V.; Avdeev, S.; Erofeev, M.; Lomaev, M.; Sosnin, E.;
Skakun, V.; Shitz, D. Acta Phys. Pol., A 2009, 116, 576–578.
(21) Wang, D.; Oppenla¨nder, T.; Gamal El-Din, M.; Bolton, J. R.
Photochem. Photobiol. 2010, 86, 176–181.
(22) Braun, A. M.; Pintori, I. G.; Popp, H.-P.; Wakahata, Y.; Wo¨rner,
M. Wat. Sci. Technol. 2004, 49, 235–240.
(23) Oppenla¨nder, T.; Schwarzwalder, R. J. AdV. Oxid. Technol. 2002,
5, 155–163.
(24) von Sonntag, C.; Schuchmann, H.-P. Methods Enzymol. 1994, 233,
3–20.
(25) Tsuji, M.; Kamo, N.; Senda, M.; Kawahara, M.; Kawahara, T.;
Hishinuma, N. Jap. J. App. Phys. 2009, 48, 046002.
(26) Bolton, J. R.; Linden, K. G. J. EnViron Eng 2003, 129, 209–215.
(27) Bothe, E.; Schuchmann, M. N.; Schulte-Frohlinde, D.; von Sonntag,
C. J. Photochem. Photobiol. 1978, 28, 639–644.
References and Notes
(1) Alberta Energy and Utilities Board. Alberta’s reserVes 2004 and
supply/demand outlook 2005-2010, Report no. ST98-2005: Alberta Energy
and Utilities Board, Calgary, Alberta, 2005.
(2) Moschopedis, S. E.; Fryer, J. F.; Speight, J. G. Fuel 1977, 56, 109–
110.
(3) MacKinnon, M. D.; and Sethi, A. A comparison of the physical
and chemical properties of the tailing ponds at the Syncrude and Suncor
oil sands plants. In Proceeding of Our Petroleum Future Conference, Alberta
Oil Sands Technology and Research Authority (AOSTRA), Edmonton,
Alberta, 1993.
(4) Han, X.; MacKinnon, M. D.; Martin, J. W. Chemosphere 2009,
76, 63–70.
(5) Allen, E. W. J. EnViron Eng. Sci. 2008, 7, 123–138.
(6) Headley, J. V.; Du, J.-L.; Peru, K. M.; McMartin, D. W. J. EnViron.
Sci. and Health A 2009, 44, 591–597.
(7) McMartin, D. W.; Headley, J. V.; Friesen, D. A.; Peru, K. M.;
Gillies, J. A. J. EnViron. Sci. Health A 2004, 39, 1361–1383, 2004.
(8) Anbar, M.; Meyerstein, D.; Neta, P. J. Chem. Soc. (B) 1966, 8,
742–747.
(9) Hewgill, F. R.; Proudfoot, G. M. Aust. J. Chem. 1976, 29, 637–
647.
(10) von Sonntag, C.; Schuchmann, H.-P. Peroxyl radicals in aqueous
solutions: Peroxyl Radicals, Alfassi, Z. B. (ed.), John Wiley & Sons,
Chichester, England, 1997.
(28) Legrini, O.; Oliveros, E.; Braun, A. M. Chem. ReV. 1993, 93, 671–
698.
(29) Kislenko, V. N.; Berlin, A. A. Usp. Khim. 1991, 60, 949–981.
(30) Dan, L.; Zhiguo, S.; Changhai, W.; Ming, G. J. Chromatogr. Sci.
2009, 47, 345–348.
(11) Oppenla¨nder, T., Photochemical Purification of Water and Air,
Wiley-VCH, Weinnheim, 2003.
JP105727S