Elucidation of the 1,4-dioxane decomposition pathway at discrete ultrasonic frequencies

Article abstract of DOI:10.1021/es000928r

The sonolytic decomposition chemistry of the refractory compound 1,4-dioxane in aqueous solution has been investigated at four ultrasonic frequencies (205, 358, 618, and 1071 kHz). To maintain fully saturated solutions, argon and oxygen were used as sparge gases. Using a frequency of 358 kHz, the observed first-order kinetic rate constants for 1,4-dioxane destruction were highest with a sparge gas ratio of 75% Ar/25% O₂ (k = 4.32 ± 0.31 x 10^-4 s^-1) and lowest in the presence of pure argon (k= 8.67 ± 0.47 x 10^-5 s^-1). Ethylene glycol diformate, methoxyacetic acid, formaldehyde, glycolic acid, and formic acid were found to be the major intermediates of 1,4-dioxane degradation. A reaction mechanism involving these byproducts was proposed concerning primarily reactions with oxidizing species (·OH, ·OOH, ·O) in and near the interfacial region of the cavitation bubble. The highest observed first-order 1,4-dioxane decomposition rate occurred at 358 followed by 618, 1071, and 205 kHz. At each frequency, approximately 85% of the initial carbon is accounted for as the parent compound, as an intermediate, or as CO₂. The major byproducts formation was investigated at all four frequencies, and the results indicate that free radical mechanisms are significant over the entire range of frequencies studied. The sonolytic decomposition chemistry of the refractory compound 1,4-dioxane in aqueous solution has been investigated at four ultrasonic frequencies (205, 358, 618, and 1071 kHz). To maintain fully saturated solutions, argon and oxygen were used as sparge gases. Using a frequency of 358 kHz, the observed first-order kinetic rate constants for 1,4-dioxane destruction were highest with a sparge gas ratio of 75% Ar/25% O₂ (k = 4.32 ± 0.31 × 10^-4 s^-1) and lowest in the presence of pure argon (k = 8.67 ± 0.47 × 10^-5 s^-1). Ethylene glycol diformate, methoxyacetic acid, formaldehyde, glycolic acid, and formic acid were found to be the major intermediates of 1,4-dioxane degradation. A reaction mechanism involving these byproducts was proposed concerning primarily reactions with oxidizing species (·OH, ·OOH, ·O) in and near the interfacial region of the cavitation bubble. The highest observed first-order 1,4-dioxane decomposition rate occurred at 358 followed by 618, 1071, and 205 kHz. At each frequency, approximately 85% of the initial carbon is accounted for as the parent compound, as an intermediate, or as CO₂. The major byproducts formation was investigated at all four frequencies, and the results indicate that free radical mechanisms are significant over the entire range of frequencies studied.

Full text of DOI:10.1021/es000928r

Environ. Sci. Technol. 2000, 34, 3944-3953
pressure (37 mmHg at 25 °C) that are characteristic of 1,4-
Elucidation of the 1,4-Dioxane
Decomposition Pathway at Discrete
Ultrasonic Frequencies
dioxane (14). Moreover, the aforementioned treatment
methods do not destroy the target compound but rather
transfer it from one phase to another.
Fenton’s reagent (15) and γ-irradiation (16) have been
explored as possible destructive treatment methods for 1,4-
dioxane. Recently, Stefan and Bolton (17) have studied the
decomposition of 1,4-dioxane subject to irradiation with
ultraviolet light (UV) combined with H2O2 and have discov-
ered ethylene glycol diformate (EGD) as the primary inter-
mediate along with glyoxal, formaldehyde, and various other
short-chain organic acid byproducts. Similarly, studies of
heterogeneous photocatalysis (18, 19) have also revealed EGD
as the principal oxidation byproduct along with minor
amounts of formic, glycolic, and oxalic acids. Sonolysis is
also a potential destructive technique that has been effectively
applied to other ethers (20). Most of these processes exploit
the reactivity of the hydroxyl radical (•OH), a nonselective
oxidant that rapidly attacks organic compounds.
Ultrasonic irradiation has been investigated in the trans-
formation of organic pollutants such as phenols (21, 22),
chlorinated organics (23), and humic acids (24). Chemical
decomposition may be enhanced when an appropriate
ultrasonic frequency is employed (25, 26). Sonochemical
reactions result from acoustic cavitation, the rapid expansion
and collapse of gas- and vapor-filled bubbles. Temperatures
as high as 4300 K (27) and pressures approaching 1000 atm
(28) can exist within each bubble upon collapse. This transient
phenomenon in turn promotes compound degradation
through various possible mechanisms (29, 30): high tem-
perature reactions, including combustion; supercritical water
oxidation; and free radical oxidation. The last is important
when the pollutant does not partition significantly into the
vapor phase. A more complete discussion of aqueous
sonochemistry and environmental applications is given by
M I C H A E L A . B E C K E T T A N D I N E Z H U A *
School of Civil Engineering, Purdue University,
West Lafayette, Indiana 47907-1284
The sonolytic decomposition chemistry of the refractory
compound 1,4-dioxane in aqueous solution has been
investigated at four ultrasonic frequencies (205, 358, 618,
and 1071 kHz). To maintain fully saturated solutions, argon
and oxygen were used as sparge gases. Using a frequency
of 358 kHz, the observed first-order kinetic rate constants for
1,4-dioxane destruction were highest with a sparge gas
ratio of 75% Ar/25% O (k ) 4.32 ( 0.31 × 10-4 s-1) and
2
lowest in the presence of pure argon (k ) 8.67 ( 0.47 × 10-5
s-1). Ethylene glycol diformate, methoxyacetic acid,
formaldehyde, glycolic acid, and formic acid were found
to be the major intermediates of 1,4-dioxane degradation. A
reaction mechanism involving these byproducts was
proposed concerning primarily reactions with oxidizing
•
•
species (•OH, OOH, O) in and near the interfacial region
of the cavitation bubble. The highest observed first-order 1,4-
dioxane decomposition rate occurred at 358 followed by
618, 1071, and 205 kHz. At each frequency, approximately
85% of the initial carbon is accounted for as the parent
compound, as an intermediate, or as CO . The major
2
byproducts formation was investigated at all four frequencies,
and the results indicate that free radical mechanisms
are significant over the entire range of frequencies studied.
Hoffmann and co-workers (31), and
a comprehensive
discussion of ultrasound and sonochemistry can be found
by Crum (32) and Suslick (33).
There still remain many unanswered questions regarding
reaction mechanisms under a variety of ultrasonic conditions.
The research reported in this paper was directed toward
elucidating decomposition products and pathways for 1,4-
dioxane during sonolysis at discrete ultrasonic frequencies.
The viability of ultrasound and the chemical pathways
comprising sonochemical destruction of dioxane are ex-
plored. Ultrasonic frequency and sparge gas variation (argon
and oxygen) are also examined as a component of optimal
ultrasonic reactor design and operation.
Introduction
1,4-Dioxane (C4H8O2) is of environmental concern for as-
sorted reasons. It is currently employed as an organic solvent
in numerous chemical processes (1) and is classified as a
toxic chemical and hazardous pollutant by the U.S. Envi-
ronmental Protection Agency (2). Because 1,4-dioxane
persists in the natural environment and can migrate swiftly
through aquifers, its presence in both surface water and
groundwater is problematic (3, 4). The compound has been
detected in groundwater in Japan (5), Canada (6), and the
United States (7). Structurally related compounds (1,3-
dioxanes) have contaminated drinking water resources in
Spain (8). Dioxane is an ether and is therefore also of interest
because it is structurally related to oxygenated fuel additives,
which are groundwater contaminants in the United States
(9).
1,4-Dioxane is resistant to both aerobic (10) and anaerobic
biological processes (11). A current treatment technology
for removing 1,4-dioxane from contaminated waste streams
is by distillation, but this process is rather expensive (12).
Other treatment methods, including activated carbon (13)
and air stripping (12), provide inefficient removals due to
the high aqueous solubility (4.31 × 105 mg/ L) and low vapor
Experimental Materials and Methods
Ultrapure water (R ) 18 MΩ cm-1) was obtained through a
Barnstead Nanopure Ultrapure water system. Reagent so-
dium chloride (Fisher), grade potassium biphthalate (J. T.
Baker, Inc.), potassium iodide (VWR Scientific), ammonium
molybdate (Fisher Scientific), 1,4-dioxane (Sigma), ethyl-
benzene (Aldrich), and GC resolve grade hexane, and
methylene chloride (Fisher Scientific) were used as received.
Fresh solutions of 1.0 mM 1,4-dioxane were used for each
individual study. Acetic, methoxyacetic, glycolic, formic, and
oxalic acids and formaldehyde (Sigma) and 2,4-dinitro-
phenylhydrazine (Aldrich) were analytical reagent grade and
also used as received. Ethylene glycol diformate (EGD) was
purchased from Fluka and used as received.
All of the sonolytic experiments were performed with an
AlliedSignal URS ultrasonic transducer powered by an Allied
Signal R/ F generator (256 W maximum output). Four
* Corresponding author phone: (765)494-2409; fax: (765)496-1107;
e-mail: hua@ecn.purdue.edu.
9
3 9 4 4 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 18, 2000
10.1021/es000928r CCC: $19.00
2000 Am erican Chem ical Society
Published on Web 08/15/2000

TABLE 1. Formation of Principal Free Radical Species during
Ultrasonic Irradiation (33, 36-38)
Argon Sparge (M is an Inert Third Molecule)
H2O f •OH + •H
•OH + •H
M8 H2O
2•OH +
M8 H2O2
2•H
M8 H2
2•OH + T •O + H2O
2•O
M8 O2
(1)
9
(2)
(3)
9
9
(4)
(5)
9
(6)
Additional Reactions in the Presence of Oxygen
O2 f 2•O
(7)
(8)
FIGURE 1. 1,4-Dioxane degradation with various gas sparges ratios.
Reactor conditions: power intensity ) 5.1 W/cm2, frequency ) 358
kHz, C ) 1.0 mM, C ) concentration at time t. Results are averaged
O2 + •O f O3
o
t
from experiments performed in duplicate. Error bars represent the
95% confidence interval.
O2 + •H f •OOH (or OH + O)
(9)
•
•O + •OOH T •OH + O2
•O + H2O f 2•OH
(10)
(11)
(12)
2•OOH f H2O2 + O2
ultrasonic frequencies were employed during the experiments
(205, 358, 618, and 1071 kHz). The active acoustical vibration
area of the transducer was 25 cm2, and the output power of
the generator (indicated on the instrument itself) was 128 W.
Therefore, the effective transducer intensity during each
experiment was 5.1 W/ cm2 (power/ vibrational area). The
irradiated aqueous solutions were contained in a water-
jacketed glass reactor (AlliedSignal, Germany) with an inner
diameter of 6 cm (25), which was connected to a circulating
cooling bath (Fisher Scientific). While the maximum volume
for the glass reactor was 700 mL, the sonicated volume for
each experiment was 500 mL.
Kinetics experiments were performed in the batch mode
by sonicating an aqueous 500-mL solution of dioxane. The
temperature was monitored continuously and was main-
tained constant at 25 ( 1 °C. Four ports were located on the
reactor cover and were used to fully saturate the solution
with argon and/ or oxygen, to monitor liquid temperature,
and to withdraw samples periodically for analysis. Solutions
were sparged 20 min prior to the beginning of each sonication
at a gas flow rate of 100 mL/ min and were continuously
sparged throughout the entire run. The sparge gases included
argon and oxygen and varied according to the type of
experiment being performed.
To study the degradation kinetics of 1,4-dioxane, 2.5-mL
aliquots were taken initially and every 10 min for each
experiment. The samples were then extracted with a 3-mL
mixture of hexane and methylene chloride (80:20, v:v).
Ethylbenzene was used as an internal standard. The organic
extracts were analyzed by injection (1 µL) into a Hewlett-
Packard (HP) 5890 gas chromatograph equipped with a flame
ionization detector (FID) and an HP 7673 autosampler. The
GC contained a Vocol column from Supelco (30 m, 0.53 mm
i.d., 3.0 µm film thickness) and was programmed with an
initial oven temperature of 60 °C for 2 min followed by a
temperature ramp of 15 °C/ min for 4 min and holding at 120
°C for 1 min. The injection port and detector temperatures
were maintained at 150 and 250 °C, respectively.
FIGURE 2. Formation of sonolytic byproducts of 1,4-dioxane over
time. Reactor conditions: frequency ) 358 kHz, sparge gas ratio
) 75%/25% (Ar/O ), power intensity ) 5.1 W/cm2, C ) 1.0 mM.
2
o
Results are averaged from experiments performed in triplicate. Error
bars represent the 95% confidence interval.
EGD was identified using the solid-phase extraction
(SPME) method from Supelco combined with GC/ MS (17).
A 10-mL aliquot of sonicated sample was vigorously mixed
with a salted (NaCl) solution (10 mL/ 3.0 g, solution/ salt).
The SPME fiber (DVB/ CAR/ PDMS) was inserted into the
solution, and the compound was allowed to adsorb onto the
fiber for 20 min. The fiber was then inserted into an HP 5890
series II GC in tandem with an HP 5972 series mass
spectrometer. SPME was necessary in order to introduce a
detectable quantity of EGD into the mass spectrometer. EGD
was then quantified using a direct aqueous injection into the
GC/ FID with a J&W Scientific DB-624 column (30 m, 0.53
mm i.d., 3.0 µm film thickness). Using this method, EGD
eluted at 2.9 min.
All organic acids were detected and quantified using a
Dionex ion chromatograph (IC) equipped with a CD 20
conductivity detector (17). Methoxyacetic (retention time (rt)
) 3.6 min), acetic (rt ) 3.9 min), glycolic (rt ) 4.2 min), and
formic acids (rt ) 5.0 min) were all measured using an IonPac
9
VOL. 34, NO. 18, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 9 4 5

FIGURE 3. Reaction scheme of the initial degradation mechanism for 1,4-dioxane. Reactions took place under the following conditions:
sparge gas ratio ) 75%/25% (Ar/O ), power intensity ) 5.1 W/cm2, frequency ) 358 kHz, C ) 1.0 mM. The major intermediates observed
2
o
in this study are highlighted in boxes.
AS4A ion exchange column with an AG4A guard column in
a 2.5 mM sodium borate eluent (flow rate ) 1.5 mL/ min).
Oxalic acid (rt ) 3.7 min) was determined using an IonPac
AS11 ion-exchange column and AG11 guard with a 12 mM
NaOH eluent (flow rate ) 1.5 mL/ min).
Formaldehyde and glyoxal were derivatized with 2,4-
dinitrophenylhydrazine (DNPH) and then measured by high-
performance liquid chromatography (HPLC) (34). The glyoxal
and formaldehyde hydrazones eluted at 5.2 and 7.5 min,
respectively, when injected into a Varian 9012 HPLC with a
Varian 9050 variable wavelength UV-Vis detector. The
hydrazones were separated on a 4.6 × 250 mm Alltech
Econosil column (C18 reversed phase) with isocratic elution
(30/ 70 vol % water/ methanol solvent mixture) and a mobile-
phase flow rate of 1.0 mL/ min.
The total organic carbon (TOC) was determined with a
Shimadzu model TOC-5000A analyzer according to Standard
Methods (35). All experiments (analysis by GC/ FID, HPLC,
TOC) except the organic acid studies (IC detection) were
performed in duplicate, and two injections were made for
each sample. The organic acid investigations were performed
in triplicate with two injections made for each sample.
9
3 9 4 6 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 18, 2000

FIGURE 4. Reaction scheme for the sonolytic degradation of ethylene glycol diformate (EGD). Scheme continued from Figure 3. The tetroxide
intermediate mentioned here is structurally different than the one shown in Figure 3.
(reaction 2), or react with another •OH radical to form
Results and Discussion
hydrogen peroxide (reaction 3). The hydrogen atom is also
Ultrasonic reactor performance can be enhanced by the
known to recombine and form hydrogen gas (reaction 4)
appropriate choice of sparge gas and ultrasonic frequency,
(36). Reactions 5 and 6, however, are very unlikely to occur
among other parameters. A number of reports indicate that
in the absence of oxygen (36-38).
sparging with appropriate gases during sonolysis will sig-
nificantly impact the efficiency of bubble collapse during
Reactive gases such as O2 will also undergo numerous
reactions within the gaseous bubble phase (Table 1) (38).
acoustic cavitation (30, 36). Gases with a higher polytropic
index, K (K ) Cp/ Cv), and lower thermal conductivity will
lead to more intense conditions within a collapsing bubble
because less heat is dissipated to the surrounding aqueous
environment during the rapid implosion.
•
•
Sonolytic reactions of O2 lead to OH, •OOH, and O radical
species (reactions 7-11). These reactive intermediates formed
in the presence of O2, along with the thermolysis of water
provide additional radical species that attack the target
•
compound. In addition, the OOH radical can recombine to
During this implosion, a number of reactions may occur
within the cavitation bubble (see Table 1). The thermolysis
of H2O (reaction 1) into the hydroxyl radical and hydrogen
atom is one of the most prevalent reactions during sonication
of aqueous solutions (33). The resulting OH radical may
oxidize organic substrates, react with hydrogen to form water
produce hydrogen peroxide (reaction 12).
To determine the optimum sparging conditions for the
reactor system, sonication was performed employing dif-
ferent sparge ratios of Ar and O2. Ar is an inert gas, and the
high-temperature decomposition of O2 is well characterized
•
9
VOL. 34, NO. 18, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 9 4 7

FIGURE 5. Reaction scheme for the pathway to methoxyacetic acid and decomposition. Scheme continued from Figure 3.
(39). A primary objective of the experiments was to identify
byproducts and to propose a decomposition pathway, which
would have been more difficult in the presence of a complex
mixture such as air. Thus, these gases were also chosen in
order to simplify data interpretation during the byproduct
studies. Pseudo-first-order plots of 1,4-dioxane degradation
with varying ratios of sparge gas at a frequency of 358 kHz
are shown in Figure 1. 1,4-Dioxane degrades more quickly
in the presence of pure oxygen than with pure Ar, an inert
as.
To account for this difference, the polytropic index, K,
and thermal conductivity of both Ar and O2 must be
considered. The polytropic index, K, correlates to the heat
released upon gas compression and is higher for Ar than for
O2 (1.66 vs 1.41) (40). The thermal conductivity of O2 is higher
than that of Ar (48.1 vs 30.6 mW m-1 K-1) (25). This leads to
the conclusion that bubble implosion in the presence of Ar
leads not only to a higher overall internal temperature but,
during the acoustic cavitation process, less heat is dissipated
to the immediate surroundings by each bubble as well. While
higher cavitation temperatures should lead to more effective
compound decomposition, the results however appear to
show something different (Figure 1). This occurrence can be
explained by the production of additional radical species
during decomposition of O2 that can compensate for the
lower internal cavitation temperatures which arise from the
use of O2 over Ar. Nevertheless, it is clear that a combination
of the sparge gases O2 and Ar lead to higher degradation
rates than either gas individually; the highest degradation
rate occurs at a 75% Ar/ 25% O2 ratio. With the 75% Ar/ 25%
O2 sparge ratio, there is an optimum balance created between
the higher temperatures generated during acoustic cavitation
from Ar and the generation of active radical species from O2
that promotes the most effective conditions for compound
destruction. This approximate ratio has also been demon-
strated to be optimal by other investigators (36).
To elucidate the reaction mechanisms of 1,4-dioxane
degradation during sonolysis, the byproducts of 1,4-dioxane
9
3 9 4 8 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 18, 2000

TABLE 2. Total Carbon Balance at 20-Min Intervals during Sonolytic Degradation of 1,4-Dioxane as a Function of Timea
TOC measd
% C
accounted for
time (min)
dioxane
EGD
FA
MAA
F
GA
OA
(mg/L as C)b
0
20
40
60
80
1.00
0
0
0
0
0
0
0
48
100
89
86
81
80
83
83
0.663
0.451
0.271
0.160
0.082
0.046
0.055
0.134
0.186
0.233
0.250
0.219
0.198
0.264
0.401
0.433
0.556
0.618
0.061
0.109
0.140
0.153
0.161
0.172
0.053
0.093
0.136
0.148
0.149
0.122
0.001
0.001
0.002
0.004
0.008
0.011
45.5
43.8
44.0
42.8
40.5
39.0
0.017
0.039
0.064
0.080
0.099
100
120
a
All concentrations are in m M unless noted otherwise. The TOC rem aining in solution is determ ined by the TOC analyzer. Reactor conditions:
sparge gas ratio ) 75%/25% (Ar/O2), power intensity ) 5.1 W/cm 2, frequency ) 358 kHz, Co ) 1.0 m M. EGD, ethylene glycol diform ate; FA, form ic
acid; MAA, m ethoxyacetic acid; F, form aldehyde; GA, glycolic acid; OA, oxalic acid. Other carbon losses can be attributed by m ineralization to
b
CO2 (m ajor) and volatilization from sparging effects (m inor). Mineralization to CO2 is accounted for by the TOC m easurem ents.
were determined and quantified using a variety of analytical
methods as discussed previously. Compounds formed at 358
kHz are shown in Figure 2. After 120 min, 96% of the initial
1,4-dioxane concentration was depleted, and five major
reaction intermediates were detected: ethylene glycol di-
formate (EGD) (C4H6O4), methoxyacetic acid (C3H6O3), formic
acid (CH2O2), glycolic acid (C2H4O3), and formaldehyde
(CH2O). Formate is produced at the fastest rate followed by
EGD. Methoxyacetic acid and formaldehyde are formed at
similar rates followed by glycolic acid, which appears after
an induction period of 30 min. The pH of the solutions also
decreased from 7.50 ( 0.25 to 3.75 ( 0.25 during the course
of these experiments. This pH decrease is consistent with
the formation of organic acids as byproducts of 1,4-dioxane
degradation.
Investigations into 1,4-dioxane gas phase thermolysis
indicate that ethylene (C2H4), carbon monoxide (CO),
FIGURE 6. 1,4-Dioxane degradation at four ultrasonic frequencies.
Reactor conditions: power intensity ) 5.1 W/cm2, sparge gas ratio
) 75%/25% (Ar/O ), C ) 1.0 mM, T ) 25 °C. Error bars represent
hydrogen gas (H2), and formaldehyde (CH2O) are byproducts
resulting from direct C-C bond cleavage (41). During
sonolysis of 1,4-dioxane, formaldehyde was detected and
attributed to an oxidation pathway in these experiments.
Ethylene, H2, and CO were not detected. The possibility of
byproduct accumulation however as a consequence of direct
1,4-dioxane ring cleavage has not been completely ruled out
(Figure 3). If this cleavage were to occur though, it would
predominantly take place between the C-C bond due its
lower bond dissociation energy (BDE ) 79 kcal/ mol) as
compared to that of C-O (85 kcal/ mol) and C-H (BDE ) 93
kcal/ mol) (42).
2
o
the 95% confidence interval.
With the free radical mechanism, the reaction between
1,4-dioxane (I) and •OH initially begins with an H-abstraction
at any of the carbons (due to symmetry) leading to the 1,4-
dioxanyl radical (II) and water as shown in Figure 3 (47-49).
•
Reactions between the OOH radical and 1,4-dioxane will
lead to the same intermediate (II) along with the production
of H2O2. Subsequent reactions of the 1,4-dioxanyl radicals
with O2 at diffusion-controlled rates lead to the peroxyl
radicals (III), which normally undergo head-to-head termi-
In the following discussion, we postulate the major
pathways leading from 1,4-dioxane to its subsequent min-
eralization based on the byproducts observed during the
sonication of 1,4-dioxane. In view of the complexity of the
system, we propose a mechanism of sonolytic 1,4-dioxane
destruction that occurs predominantly with reactions in-
volving strongly oxidizing radical species (•OH,•OOH,•O) at
and near the interfacial region of the cavitation bubble.
Hydroxyl radicals have been specifically identified in
sonicated aqueous solutions via electron spin resonance (43).
Oxygen atoms (•O) (36, 39) and hydroperoxyl radicals (•OOH)
(36, 44), although not directly quantified or detected, have
been inferred from specific sonochemical reactions. The
reaction of 1,4-dioxane by other free radicals should not be
nation reactions (bimolecular rate constants >109 M-1 s-1
)
to form the tetroxide intermediate (IV) (50). Once formed,
this tetroxide intermediate rapidly decomposes to form R-oxyl
radicals (V) and molecular oxygen (17, 50). In the case of a
tetroxide derived from 1,4-dioxanyl peroxyl radicals, an
electrocyclic process leading to H2O2 and two carbonyl
compounds or a disproportionation reaction leading to a
2-hydroxy alcohol, a corresponding carbonyl compound, and
oxygen via the Russell mechanism (51) appear unlikely due
to stereochemical hindrance (17).
Upon formation, these R-oxyl radicals may undergo two
primary types of reactions: (i) â-C-C fragmentation leading
to radical VI or (ii) 1,2-H-shift leading to R-hydroxyl radicals
(X) (52). Further reactions involving radicals VI consume
oxygen and form peroxyl radicals (VII), which can then
dimerize to yield a tetroxide intermediate (VIII). The tetroxide
rapidly decomposes to ethylene glycol diformate (EGD) (IX)
and hydrogen peroxide (Figure 3). Maurino et al. (18) also
have evidence that the formation of EGD is indicative of a
sequential oxidative ring-opening mechanism.
•
•
ruled out, although hydrogen abstraction by OOH and O
atom with organic substrates may lead to byproducts that
are similar to those resulting from •OH radical attack. In
addition, the •O atom, due to its highly reactive nature in the
gaseous phase (38), is more likely to play an indirect role by
providing radical species (by reaction with oxygen and water
vapor) with longer lifetimes that can subsequently move to
the interfacial region and react with the target compound
(45, 46). All free radicals formed however may react with the
target compound or other molecules that are present within
or in close proximity to the cavitation bubble.
The further decomposition of EGD is initiated with
•
•
H-abstraction by OH or OOH attack (Figure 4). Transfor-
mation may proceed in the presence of O2 by forming another
9
VOL. 34, NO. 18, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 9 4 9

FIGURE 7. Formation of (A) ethylene glycol diformate (EGD) and (B) formaldehyde (F) over time. Conditions used: sparge gas ratio )
75%/25% (Ar/O ), power intensity ) 5.1 W/cm2, C ) 1.0 mM. Results are averaged from experiments performed in triplicate. Error bars
2
o
represent the 95% confidence interval.
tetroxide intermediate that can decompose via either a free
radical mechanism or a concerted mechanism (50). If the
free radical pathway takes place, two alkoxyl radicals (XII)
and O2 are formed. The alkoxyl radicals (XII) then undergo
â-scission and react to form aldehyde esters (XIV) and formic
acid (XV). The aldehyde esters are unstable and quickly
hydrolyze to formic acid (XV) and glycolaldehyde (XVI) (17).
Additional reactions involving glycolaldehyde (XVI) and free
radical species will lead to the rapid formation of glyoxal
(XVII) and glycolic acid (XVIII). Because glyoxal was measured
in trace concentrations during our investigations, we consider
this pathway to be minor. The concerted mechanism involves
the formation of H2O2 and an ester intermediate (XIII), which
can then hydrolyze to glycolic (XVIII) and formic acids (XV)
(17). It is not possible to further distinguish the relative
importance of each pathway based on H2O2 measurements
over time because H2O2 results from other reactions inde-
pendent of 1,4-dioxane decomposition (36, 53).
acid (MAA) as an intermediate (17). The researchers examined
the degradation of 1,4-dioxane in dilute aqueous solution by
irradiation with UV/ H2O2 and proposed that MAA was
eventually formed from an R-oxyl radical in a process
•
involving OOH as a reducing agent under reducing condi-
tions. Furthermore, in their experiments acetic acid was
formed as a byproduct of MAA in reducing conditions
•
involving OOH as well.
It is unlikely that reducing conditions exist in the
sonochemical reactor under the conditions reported in this
paper. Moreover, acetic acid was not detected during
sonication of aqueous MAA under the same conditions of
1,4-dioxane sonication. It is therefore hypothesized that the
formation of MAA occurs via the 1,2-H-shift pathway
mentioned previously (Figure 3). The bond cleavage, pro-
moted by the extreme conditions inside the cavitation bubble,
occurs between the C4-O bond (Figure 5). The oxygen can
then donate its unpaired electron to the carbon (denoted C1
in Figure 5) leading to the formation of a double bond between
the oxygen and C1. This ring opening leads to the creation
of the carboxyl radical XIX, which will, in a series of
Although there are previous investigations of 1,4-dioxane
decomposition due to oxidation, photolysis, or other de-
struction methods, only one study reports methoxyacetic
9
3 9 5 0 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 18, 2000

FIGURE 8. Formation of (A) methoxyacetic acid (MAA) and (B) formic acid (FA) over time. Conditions used: sparge gas ratio ) 75%/25%
(Ar/O ), power intensity ) 5.1 W/cm2, C ) 1.0 mM. Results are averaged from experiments performed in triplicate. Error bars represent
2
o
the 95% confidence interval.
undetermined oxidation steps involving free radicals possibly
involving the formation of a tetroxide intermediate, subse-
quently form MAA (XXI).
formic acid results from a number of pathways, which is
consistent with its presence at the highest concentration
among the intermediates (Figure 2).
A total carbon balance on 1,4-dioxane and its byproducts
was determined as follows:
MAA will then continue to react with free radicals through
H-abstraction at the methyl carbon followed by â-scission.
Formaldehyde (XX) and the formyl methyl radical (XXII) are
thus created. Formaldehyde is rapidly oxidized to formic acid,
which eventually oxidizes to CO2 (54). The formyl methyl
radical can then combine with •OH to form glycolic acid
(XVIII). Both formaldehyde and glycolic acid were detected
when MAA was sonicated alone, thereby giving support for
the proposed pathway in Figure 5.
[total carbon (mM as C)]t ) [1,4-dioxane]t +
[intermediates]t + [carbon mineralized]t (13)
∑
where the subscript t denotes concentrations measured after
t minutes of sonication. The total carbon mineralized
(converted to CO2) was determined using a TOC analyzer.
The results of the total carbon analysis at 358 kHz are shown
in Table 2. For the duration of the experiments, approximately
80-85% of the initial carbon is accounted for as the parent
compound, as an intermediate, or as CO2. Possible physical
loss mechanisms include vaporization; continuous sparging
is necessary to maintain gas saturation during the experi-
ments. Sparging studies in the absence of ultrasound were
performed on 1.0 mM solutions of each identified byproduct
for 2 h in order to ascertain volatilization losses. During that
Trace amounts of oxalate were measured during the
sonication of 1,4-dioxane but only appeared after the
formation of glycolic acid. The pathway for glycolic acid
oxidation to CO2 has previously been shown to produce
formate and oxalate as intermediates prior to mineralization
(17). In this study, when aqueous glycolic acid was sonicated,
formate and oxalate were detected in significant amounts,
thereby accounting for the production of oxalic acid. As is
evident in the proposed reaction schemes, the generation of
9
VOL. 34, NO. 18, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 9 5 1

time period, there was a maximum loss of 3-5% for each
compound. By assuming a 5% loss for each compound over
time, nearly 90% of the organic carbon can be accounted for.
The impact of frequency on sonolytic environmental
processes has been shown to be significant (55, 56). The
results have varied widely due to the use of different types
of ultrasonic emitters such as probes or plates and the
diversity of reactors incorporated (30, 55). The effect of
sonolysis on the degradation rate of 1,4-dioxane was studied
over a range of frequencies utilizing the same transducer
and reactor system for each experiment to keep the power/
area and power/ volume ratios consistent. The ultrasonic
reaction conditions included the predetermined optimal
sparge gas ratio of 75% Ar/ 25% O2 and a power intensity of
5.1 W/ cm2 at four frequencies. The fastest degradation occurs
at a frequency of 358 kHz followed by 618, 1071, and 205 kHz
(Figure 6).
ing with ultrasonic experimental observations (59). By
examining the reaction scheme leading to the formation of
methoxyacetic acid (Figure 3), it is possible that these extreme
conditions facilitate the scission of radicalX more extensively
after the 1,2-H-shift takes place at 205 kHz than during cavity
collapse at 1071 kHz. Although extreme conditions in
cavitation bubbles decrease with increasing frequency,
cavitation intensity at 358 and 618 kHz is still substantial
along with an increased number of these radicals capable of
reaching the gas phase via higher induced mass transfer rates.
At 1071 kHz, the mass transfer of solutes in to and out of the
cavitation bubble is greater than at lower frequencies, but
the energy available for bond cleavage upon bubble implosion
is diminished. This distinction accounts for overall higher
concentrations of methoxyacetic acid produced at 358 and
618 than at 205 and 1071 kHz. Nevertheless, the overall
conclusion from these byproduct results indicates that the
decomposition pathways at each frequency are analogous.
By using ultrasound, a refractory contaminant such as
1,4-dioxane can be decomposed into smaller short-chain
organics that can then be degraded more effectively by other
means such as biological processes. The results of this
investigation have shown that 1,4-dioxane is sonolytically
decomposed, leading to the formation of major intermediates
ethylene glycol diformate and methoxyacetic acid followed
by formaldehyde, formic acid, and glycolic acid. Free radical
pathways are significant over the entire range of frequencies
investigated (205, 358, 618, and 1071 kHz). Sparging gases
and frequency also play an important role in optimizing
sonochemical effects.
To explain this phenomenon, the following equation,
which reveals that the resonant size of an acoustic bubble
is inversely correlated to the emitted frequency, must be
considered (57):
Rr2 ) 3κP0/ Fωr2
(14)
where Rr is the resonant bubble radius, K (K ) Cp/ Cv) is the
polytropic index, P0 is the hydrostatic pressure, Fis the density
of the solution, and ωr is the resonant frequency. Lower
frequencies lead to a more violent collapse than higher
frequencies due to greater resonant bubble sizes. As the
frequency increases, bubble lifetimes are shorter, but there
are more cavitation events per unit of time, and the bubble
surface area to volume ratio is increased. The effect of this
Acknowledgments
•
is to increase the mass transfer of OH radicals into the
We thank the United States Department of Energy (DOE Grant
DE-FG07-96ER14710) and the Purdue Research Foundation
(Award 6902949) for funding these studies. Furthermore, we
thank Dr. Mihaela Stefan of the Siebens Drake Research
Institute for helpful advice and Dr. Changhe Xiao of the School
of Civil Engineering at Purdue University for analytical
laboratory assistance. Finally, we thank Wilmarie Flores for
her efforts in performing some of the experiments.
surrounding medium and concurrently increase diffusion of
volatile compounds into the bubble (56). At high frequencies,
the resonant bubble size is not large enough to produce
enough energy upon collapse to form sufficient numbers of
•OH radicals from water, and a point of diminishing returns
is reached. The frequency range proximal to 358 kHz appears
to optimize both energy from bubble implosion and con-
comitant mass transfer of reactive species in to and out of
the bubble.
Literature Cited
Since there have been few in depth investigations
comparing byproduct formation at several frequencies, the
byproducts of 1,4-dioxane decomposition were examined at
all frequencies (205, 358, 618, and 1071 kHz) in this study.
The formation of ethylene glycol diformate, methoxyacetic
acid, formic acid, and formaldehyde are shown in Figures 7
and 8. TOC measurements at 208, 618, and 1071 kHz were
also made. Carbon balance calculations accounted for
approximately 80-85% of all carbon, commensurate with
what was observed at 358 kHz. As seen in Figure 6, the
degradation rate of 1,4-dioxane is most rapid at 358 followed
by 618, 1071, and 205 kHz. The accumulation of each
intermediate is clearly the greatest at 358 kHz, consistent
with the parent compound decomposing most rapidly at
this frequency. The accumulation of most intermediates
follows the ordering of the 1,4-dioxane degradation rates.
However, production of ethylene glycol diformate, formal-
dehyde, and formic acid is slowest at 205 kHz.
The order of methoxyacetic acid accumulation (Figure
8A) is different however (358, 618, 205, followed by 1071 kHz).
Methoxyacetic acid accumulates more rapidly at 205 than at
1071 kHz during most of the experiment. The most violent
conditions and the hottest temperatures and pressures within
the cavitation bubble tend to decrease as the frequency is
increased. Evidence for this has been corroborated by both
studies on sonoluminescence (58), the emission of light
during the ultrasonic irradiation of a liquid, and acoustic
cavitation field predictions combining mathematical model-
(1) The Merck Index. An Encyclopedia of Chemicals, Drugs, and
Biologicals, 11th ed.; Budavari, S., Ed.; Merck and Co. Inc.:
Rahway, NJ, 1989.
(2) Handbook of Toxic and Hazardous Chemicals and Carcinogens,
3rd ed.; Sittig, M., Ed.; Noyes Publishers: Park Ridge, NJ, 1991;
p 1.
(3) Abe, A. Sci. Total Environ. 1999, 227, 41-47.
(4) Jackson, R. E.; Dwarakanath, V. Ground Water Monit. Remed.
1999, 19, 102-110.
(5) Yasuhara, A.; et al. Waste Manage. Res. 1999, 17, 186-197.
(6) Lesage, S.; Jackson, R. E.; Priddle, M. W.; Riemann, P. G. Environ.
Sci. Technol. 1990, 24, 559-566.
(7) Taylor, S. W.; Lange, C. R.; Lesold, E. A. Ground Water 1997, 35,
973-980.
(8) Romero, J.; Ventura, F.; Caixach, J.; Rivera, J.; Gode, L. X.;
Ninerola, J. M. Environ. Sci. Technol. 1998, 32, 206-216.
(9) Mormile, M. R.; Liu, S.; Suflita, J. M. Environ. Sci. Technol. 1994,
28, 1727-1732.
(10) Raj, C. B. C.; Ramkumar, N.; Siraj, A. H. J.; Chidambaram, S.
Process Saf. Environ. Prot. 1997, 75, 245-256.
(11) Adams, C. D.; Scanlan, P. A.; Secrist, N. D. Environ. Sci. Technol.
1994, 29, 1812-1818.
(12) Ware, G. W., Ed. p-Dioxane. In Reviews of Environmental
Contamination and Toxicology; Springer-Verlag: New York, p
113-121.
(13) Johns, M. M.; Marshall, W. E.; Toles, C. A. J. Chem. Technol.
Biotechnol. 1998, 71, 131-140.
(14) Lagrega, M. D.; Buckingham, P. L.; Evans, J. C. Hazard. Waste
Manage. McGraw-Hill: New York, 1994.
(15) Klecka, G. M.; Gonsior, S. J. J. Hazard. Mater. 1986, 13, 161-
168.
9
3 9 5 2 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 18, 2000

(16) Gehringer, P.; Matschiner, H. Water Sci. Technol. 1998, 37, 195-
201.
(40) CRC Handbook of Chemistry and Physics, 77th ed.; Lide, D. R.,
Ed.; CRC Press: Boca Raton, FL, 1996.
(17) Stefan, M. I.; Bolton, J. R. Environ. Sci. Technol. 1998, 32, 1588-
(41) Battin, F.; Marquaire, P. M.; Baronnet, F.; Come, G. M. J. Anal.
Appl. Pyrolysis 1989, 16, 345-354.
(42) Muller, C.; Scacchi, G.; Come, G. M. The Role of Data in Scientific
Progress; Glaeser, P. S., Ed.; Elsevier: Amsterdam, 1985.
(43) Makino, K.; Mossoba, M. M.; Riesz, P. J. Phys. Chem. 1983, 87,
1369-1377.
(44) Suslick, K. S.; Grinstaff, M. W. J. Am. Chem. Soc. 1990, 112, 7807.
1595.
(18) Maurino, V.; Calza, P.; Minero, C.; Pelizzetti, E.; Vincenti, M.
Chemosphere 1997, 35, 2675-2688.
(19) Hill, R. R.; Jeffs, G. E.; Roberts, D. R. J. Photochem. Photobiol.
A: Chem. 1997, 108, 55-58.
(20) Kang, J. W.; Hoffman, M. R. Environ. Sci. Technol. 1998, 32,
3194-3199.
(21) Kotronarou, A.; Mills, G.; Hoffmann, M. R. J. Phys. Chem. 1991,
95, 3630-3638.
(45) Fang, X.; Mark, G.; Von Sonntag, C. Ultrason. Sonochem. 1996,
3, 57-63.
(22) Trabelsi, F.; Ait-Lyazidi, H.; Ratsimba, B.; Wilhelm, A. M.; Delmas,
H.; Fabre, P. L.; Berlan, J. Chem. Eng. Sci. 1996, 51, 1857-1865.
(23) Catallo, W. J.; Junk, T. Waste Manage. 1995, 15, 303-309.
(24) Nagata, Y.; Hirai, K.; Bandow, H.; Maeda, Y. Environ. Sci. Technol.
1996, 30, 113-1138.
(25) Kang, J. W.; Hung, H. M.; Lin, A.; Hoffmann, M. R. Environ. Sci.
Technol. 1999, 33, 3199-3205.
(26) Entezari, M. H.; Kruus, P. Ultrason. Sonchem. 1994, 1, S75-S77.
(27) Didenko, Y. T.; McNamara, W. B., III; Suslick, K. S. J. Phys. Chem.
A 1999, 103, 10783-10788.
(28) Suslick, K. S.; Kemper, K. A.; Flint, E. B. Proc. IEEE Ultrason.
Symp. 1993, 2, 777-783.
(46) Hart, E. J.; Fischer, C.-H.; Henglein, A. J. Phys. Chem. 1986, 90,
5989-5991.
(47) Eibenberger, J.; Schulte-Frohlinde, D.; Steenken, S. J. Phys. Chem.
1980, 84, 704-710.
(48) Maurer, T.; Hass, H.; Barnes, I.; Becker, K. H. J. Phys. Chem. A
1999, 103, 5032-5039.
(49) Platz, J.; Sehested, J.; Mogelberg, T.; Nielsen, O. J.; Wallington,
T. J. J. Chem. Soc., Faraday Trans. 1997, 93, 2855-2863.
(50) Schuchmann, M. N.; Von Sonntag, C. J. Phys. Chem. 1982, 86,
1995-2000.
(51) Russell, G. A. J. Am. Chem. Soc. 1957, 79, 3871-3877.
(52) Nese, C.; Schuchmann, M. N.; Steenken, S.; Von Sonntag, C. J.
Chem. Soc., Perkin Trans. 2 1995, 1037-1044.
(29) Hua, I.; Hochemer, R. H.; Hoffmann, M. R. J. Phys. Chem. 1995,
99, 2335-2342.
(30) Hua, I.; Hoffman, M. Environ. Sci. Technol. 1997, 31, 2237-
(53) Henglein, A.; Kormann, C. Int. J. Radiat. Biol. 1985, 48, 251-
2243.
258.
(31) Hoffmann, M. R.; Hua, I.; Hochemer, R. Ultrason. Sonochem.
1996, 3, S163-S172.
(54) Carey, F. A. Organic Chemistry, 3rd ed.; McGraw-Hill: New York,
1996.
(32) Crum, L. A., Mason, T. J., Reisse, J. L., Suslick, K. S., Eds.
Sonochemistry and Sonoluminescence; NATO ASI Series Series
C: Mathematical and Physical Sciences Vol. 524; Kluwer
Academic Publishers: Dordrecht, The Netherlands, 1999; p 404.
(33) Suslick, K. S. Ultrasound: Its Chemical, Physical and Biological
Effects; Suslick, K. S., Ed.; VCH Publishers: New York, 1988.
(34) Fung, K.; Grosjean, D. Anal. Chem. 1981, 53, 168-171.
(35) APHA, AWWA, and WEF. Standard Methods for the Examination
of Water and Wastewater, 19th ed.; United Book Press: Baltimore,
MD, 1995.
(36) Hart, E. J.; Henglein, A. J. Phys. Chem. 1985, 89, 4342-4347.
(37) Hart, E. J.; Henglein, A. J. Phys. Chem. 1987, 91, 3654-3656.
(38) Colussi, A. J.; Weavers, L. K.; Hoffmann, M. R. J. Phys. Chem. A
1998, 102, 6927-6934.
(55) Cum, G.; Galli, G.; Gallo, R.; Spadaro, A. Ultrasonics 1992, 30,
267-270.
(56) Hung, H.-M.; Hoffmann, M. R. J. Phys. Chem. A 1999, 103, 2734-
2739.
(57) Leighton, T. G. The Acoustic Bubble; Harcourt Brace & Com-
pany: Orlando, FL, 1994.
(58) Walton, A. J.; Reynolds, G. T. Adv. Phys. 1984, 33, 595-660.
(59) Laborde, J. L.; Bouyer, C.; Caltagirone, J. P.; Gerard, A. Ultrasonics
1998, 36, 581-587.
Received for review January 21, 2000. Revised manuscript
received June 27, 2000. Accepted June 29, 2000.
(39) Fischer, C. H.; Hart, E. J.; Henglein, A. J. Phys. Chem. 1986, 90,
1954-1956.
ES000928R
9
VOL. 34, NO. 18, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 9 5 3