Chemosphere 73 (2008) 1708–1715

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Degradation of tert-butyl formate and its intermediates by an ozone/UV process

Temesgen Garoma *, Mirat D. Gurol, Lalitha Thotakura, Olufisayo OsiboduDepartment of Civil, Construction, and Environmental Engineering, San Diego State University, 5500 Campanile Drive, San Diego, CA 92182-1324, United States

a r t i c l e i n f o

Article history:Received 23 July 2008Received in revised form 9 September 2008Accepted 10 September 2008Available online 16 October 2008

Keywords:tert-Butyl formate (TBF)Ozone/UV processOxidationDegradation pathways

0045-6535/$ - see front matter � 2008 Elsevier Ltd. Adoi:10.1016/j.chemosphere.2008.09.039

* Corresponding author. Tel.: +1 619 594 0957; faxE-mail address: tgaroma@mail.sdsu.edu (T. Garom

a b s t r a c t

In this paper, the oxidation of tert-butyl formate (TBF) in aqueous solution by an ozone/UV process wasdescribed. The oxidation process was investigated experimentally in a semibatch reactor. The results ofthe study indicated that the ozone/UV process was very effective in oxidizing TBF. tert-Butyl alcohol(TBA), hydroxy-iso-butyraldehyde (HiBA), acetone, formaldehyde, and formic acid were identified asmajor primary intermediates during the oxidation of TBF. About 90% organic carbon balance wasobtained indicating that most reaction intermediates have been identified and quantified. Some of theprimary intermediates were also oxidized in the ozone/UV system. Accordingly, HiBA, acetone, formalde-hyde, and formic acid were the primary intermediates of TBA oxidation. The oxidation of acetone in theozone/UV system generated formaldehyde, pyruvaldehyde, acetic acid, formic acid as primary interme-diates. It was also observed that the reaction intermediates formed during the oxidation of TBF react wellin the ozone/UV system and complete mineralization could be achieved by the process.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

tert-Butyl formate (TBF) is a major intermediate produced dur-ing the biodegradation of methyl tert-butyl ether (MTBE) (Mormileet al., 1994). It is also produced during the oxidation of MTBE bychemical oxidation (Damm et al., 2002) and advanced oxidationprocesses (Barreto et al., 1995; Chang and Young, 2000; Stefanet al., 2000; Acero et al., 2001; Burbano et al., 2002; Garoma andGurol, 2006). TBF is also detected in exhaust emissions from carsusing reformulated gasoline containing MTBE (Froines et al.,1998). Earlier studies on the biodegradability of MTBE indicatedthat it degrades very slowly under aerobic conditions (Mormileet al., 1994; Yeh and Novak, 1994). Recent studies reported higherrates for MTBE biodegradation under reducing conditions (Bradleyet al., 2001; Finneran and Lovley, 2001; Somsamak et al., 2001)than in aerobic environment. It was also reported that the degrada-tion of MTBE was significantly reduced by the presence of TBF.Francois et al. (2003) reported that about 6 and 17 mg L�1 of TBFin about 20 mg L�1 MTBE solution decreased the degradation rateof MTBE by approximately 30% and 42%, respectively.

Although studies on animals indicated that MTBE and one of itsdegradation byproduct, tert-butyl alcohol (TBA), are suspected aspossible human carcinogens (Cirvello et al., 1995; Mehlman,2002), TBF has not been tested for human health effects. Exposurestudies at 700 ppm and higher levels showed significant respira-tory depression in animals (Froines et al., 1998). TBF is also likelyto have irritant effects on human respiratory tract and eyes be-

ll rights reserved.

: +1 619 594 8078.a).

cause other aliphatic esters, such as isobutyl formate and n-butylformate, are known to exhibit these effects (Sandmeyer andKirwin, 1981).

The widespread detection of MTBE in drinking water sources(surface water and groundwater) and treated water has causedgreat concern in the scientific communities and public healthagencies because MTBE is suspected as a possible human carcino-gen (Mehlman, 2002). Consequently, 25 states including California,which accounts for over 30% of US’s MTBE consumption, havepassed legislation that would completely ban or partially restrictthe use of MTBE in gasoline (EPA, 2004). However, MTBE that hasalready been released into the environment will likely continueto degrade groundwater and soils for several years to come be-cause MTBE is known to degrade slowly by microorganisms (Mor-mile et al., 1994; Yeh and Novak, 1994). In addition, thebiodegradation rate for TBF is about four times less than that ofMTBE (Francois et al., 2003), and therefore, it will likely to accumu-late in groundwater. Thus, TBF must be removed from groundwa-ter before it can be used for domestic applications.

The main objective of this study is to investigate the applicationof an ozone/UV process for the removal of TBF from dilute aqueoussolution, identify the degradation products, and propose degrada-tion pathways for TBF. Ozone/UV process has shown to be a viableoption for the removal of organic pollutants from water. With thedevelopment of large scale ozone generators, capable of generatingozone directly from atmospheric air, and energy efficient UV sys-tems, the process has been getting increasing interest. The majoradvantage of the ozone/UV process is attributed to hydroxyl radical(OH�) which is generated as a result of ozone decomposition andattacks indiscriminately almost any organic compounds with a

T. Garoma et al. / Chemosphere 73 (2008) 1708–1715 1709

very high reaction rate constant (Haag and Yao, 1992) as opposedto aqueous ozone (Hoigne and Bader, 1983a,b). In addition, the OH�

oxidizes a wide variety of organic pollutants into innocuous inter-mediates and end products.

The decomposition of aqueous ozone in pure water is initiatedby its reaction with OH�, and this reaction leads to the productionof radicals that propagate the decomposition process by chains ofradical–radical or radical–solute reactions and produce OH� (Staeh-elin and Hoigne, 1982)

O3 !OH�

OH� ð1Þ

The above process is known to occur very slowly at low pH lev-els (Staehelin and Hoigne, 1982) and the decomposition processcan be further accelerated with UV radiation. In aqueous phase,the photolysis of ozone produces H2O2 as primary product (Peytonand Glaze, 1988), which then can be photolyzed into OH� directlyand can react with aqueous ozone to produce OH� as shown below.The principal reactions involved in an ozone/UV system in naturalwaters were summarized in an earlier publication; and interestedreaders are referred to Table 2 of the publication (Garoma andGurol, 2004).

O3 þH2O!hvH2O2

þO3 ! OH�

!hvOH�

(ð2Þ

2. Methods and materials

2.1. Approach

The experimental setup for the study (Fig. 1) consists of a quartzreactor placed inside a concentric photochemical chamber thatcontains 16 removable low-pressure mercury lamps. The reactionvessel has an internal diameter of 13 cm and was filled with 3 Lof solution during experiments. The lamps emit UV light primaryat 254 nm at a power rated (out-of-ballast) as 2.2 W per lamp asspecified by the manufacturer. A glass diffuser was used to spargeozone gas into the solution at a constant flow rate of 1.5 L min�1.

The reactor was operated in a semibatch mode and the contentswere stirred continuously using a magnetic stirrer. For each exper-iment, appropriate numbers of UV lamps were inserted into thephotochemical chamber and solution of TBF or intermediates pre-

FlowMeter

S

TemperatureProbe

Sample

To Exhaust

UV LampReaction Vessel

UV Chamber

FlowMeter

 

 

Fig. 1. Experime

pared at desired concentration was placed inside the reactor. Attime zero, ozone gas was introduced into the reactor at desiredflow rate and UV lamps were switched on. The experiments wereconducted at room temperature and the increase in temperaturedue to UV light exposure was less than 2 �C. Aliquot samples werewithdrawn periodically and analyzed for residual concentration ofTBF, reaction intermediates, and total organic carbon (TOC). Thesamples were analyzed on the same day since TBF is known tohydrolyze to TBA (Church et al., 1999; Stefan et al., 2000). Kineticdata indicate that the half-life for hydrolysis of TBF is the slowestat neutral pH and 22 �C which is estimated to be 5 d (Churchet al., 1999). In the current study, the solution pH was adjustedto 7.0 at the start of each experiment by using dilute solutions ofsulfuric acid (H2SO4) and sodium hydroxide (NaOH). Therefore,no significant TBF hydrolysis to TBA could occur in the time scaleof the experiment and pH ranges used in this study.

2.2. Materials

All chemicals and reagents used in the research were analyticalgrade. TBF and TBA were purchased from Sigma–Aldrich. Acetonewas purchased from J.T. Baker chemical company. Other chemicalsand reagents used in this study were purchased either from Sigma–Aldrich, Fisher Scientific, or J.T. Baker chemical companies. Ozonegas was generated from oxygen using Osmotic OREC V-10 ozonegenerator, and the amount of ozone produced in the oxygen gaswas controlled by changing power input to the generator.

2.3. Analyses

The influent and effluent ozone gas concentration was mea-sured at 258 nm by a Shimadzu UV-1601 spectrophotometer thatwas calibrated by the Potassium Iodide method (APHA, 1998).Ozone concentration in the liquid phase was determined by the In-digo method (Bader and Hoigne, 1981). The intensity of UV lampsas received in the photochemical reactor was measured by themethod of potassium ferrioxalate actinometry (Hatchard and Par-ker, 1956). In this study, 4 and 8 lamps were used and the incidentUV light intensities measured in the experimental reactor were1.35 and 2.73 W L�1, respectively.

For the quantification of TBF, TBA, and acetone, aqueous sam-ples were analyzed according to EPA Method 5030 by using a

OxygenTank

Ozone Generator

FlowMeter

UV  pectrophotometer

To Exhaust

 

 

ntal setup.

1710 T. Garoma et al. / Chemosphere 73 (2008) 1708–1715

Perkin Elmer gas chromatograph (GC) equipped with flame ioniza-tion detector. The GC was connected to purge and trap concentra-tor (Tekmar–Dohrmann 3000 sample concentrator) and interfacedto Peak Simple workstation. The purge and trap unit was interfacedwith Tekmar ALS-2016 purge autosampler and to an automaticsample heater. Aqueous samples were injected to the purge andtrap, and then purged by the autosampler to the concentrator.From the concentrator, samples were desorbed to a GC column.The column was a DB-624: 75 m � 0.53 mm i.d. with 5 lm film(J&W Scientific). Helium was used as a carrier gas at 15 mL min�1,while hydrogen and air were used to fuel the detector. The injectorand detector were operated at temperatures of 170 �C and 300 �C,respectively. The initial oven temperature was 40 �C, which waskept isothermal for 13 min. Subsequently it was ramped at a rateof 25 �C min�1 to 200 �C and kept again isothermal for 2 min.

Aldehydes were derivatized in aqueous phase by using penta-fluorobenzylhydroxylamine according to the standard methods(APHA, 1998). The derivatized aldehydes were then analyzedaccording to EPA Method 1666 using HP 6890 GC with HP 5973MS detector. The GC method was as follows: 1.0 lL splitless injec-tion, injector temperature 175 �C, detector temperature 250 �C,and oven temperature 40 �C for 2 min, then 5 �C min�1 ramp to220 � C, hold for 2 min, ramp to 250 �C at 20 �C min�1, hold for2 min. The column was an RTX-MS5 30 m � 0.25 mm i.d.,

0.0

0.2

0.4

0.6

0.8

1.0

1.2

TBF

(mM

)

[TBF]o=1.03 mM(O3)inf=39 mg/L(O3)eff=15 mg/LIo=2.73 W/L[HCO3

-]=2 mMpH=7.0Qg=1.5 L/min

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 10 20 30

Time (mi

TBF

(mM

)

[TBF]o=1.03 mM(O3)inf=39 mg/L(O3)eff=15 mg/LIo=2.73 W/L[HCO3

-]=2 mMpH=7.0Qg=1.5 L/min

a

b

Fig. 2. TBF oxidation and accumulation and decay of intermediates. (a) Accumulatintermediates.

0.25 lm film thickness (Restek). Hydroxy-iso-butyraldehyde(HiBA) was quantified by using isobutyraldehyde since it was notcommercially available (Stefan et al., 2000; Acero et al., 2001).

Organic acids were analyzed using a Dionex 500 Ion Chromato-graph according to EPA Method 300. The method requires AS11analytical and AG11 guard columns, isocratic and gradient-stageseparation modes, and conductivity detector and an anion trap col-umn to remove anionic contaminants from the eluent. Solutions ofNaOH prepared at 1 and 100 mM and deionized water were usedas an eluent at a flow rate of 1.0 mL min�1. During a typical run cy-cle, fist the AS11 column was equilibrated with 0.1 mM NaOH forten min, and then the sample is injected and the eluent was keptisocratic at 0.1 mM for five more min. Subsequently, the eluent isramped to 30 mM NaOH in 30 min. Before the next run cycle starts,the column was regenerated with 40 mM of NaOH for 5 min. ForTOC analysis, a TOC-5000A total organic carbon analyzer with anintegrated ASI-5000A autosampler was used.

3. Results and discussion

3.1. TBF degradation and accumulation and decay of intermediates

Fig. 2a and b shows the degradation of TBF and reaction inter-mediates formed during an ozone/UV treatment of 1.03 mM

0.0

0.1

0.2

0.3

0.4

Prod

ucts

(mM

)

TBF

TBA

HiBA

AC

FAD

FOA

40 50 60

nutes)

0.0

0.1

0.2

0.3

0.4

Prod

ucts

(mM

)

TBF

PYD

ACA

PYA

OXA

ion and decay of primary intermediates. (b) Accumulation and decay of other

T. Garoma et al. / Chemosphere 73 (2008) 1708–1715 1711

(105 mg L�1) TBF solution. The solution was exposed to an incidentUV light intensity of 2.73 W L�1 and influent ozone gas concentra-tion of 39 mg L�1. The effluent ozone concentration was 15 mg L�1.The solution pH was adjusted to 7.0 at the start of the experimentby using dilute solutions of H2SO4 and NaOH. The solution was alsobuffered at 2 mM of HCO�3 using solution of sodium bicarbonate(NaHCO3) to avoid large change in pH during the course of theexperiment. However, it was observed that the pH increased to avalue of about 8.5 at the end of the experiment. As CO2 is producedin the oxidation process, pH is expected to decrease. The increasein pH is most likely due to the stripping of CO2 caused by the ozonegas bubbling into the solution. The scavenging of CO2 by OH� couldbe another reason for the observed increase in pH. In this study, noattempt was made to investigate the effect of pH on the rate of TBFremoval. In an earlier study, it was observed that the removal effi-ciency of oxalic acid in an ozone/UV system increased with a de-crease in pH (Garoma and Gurol, 2005). From a practical point ofview, however, lower pH may not be applicable in the actual de-sign of an ozone/UV process because at lower pH values corrosioncan be a problem.

Since TBF reacts with OH� at a high rate constant of 1.2 � 109

M�1 s�1 (Chang and Young, 2000), it decayed rapidly followingpseudo-first-order kinetics with an observed rate constant of1.61 � 10�3 s�1 for an influent ozone gas concentration of

0.0

0.2

0.4

0.6

TBA

(mM

)

[((I[pQ

0.0

0.2

0.4

0.6

0 10 20 30

Time (mi

TBA

(mM

)

[((I[pQ

a

b

Fig. 3. TBA oxidation and accumulation and decay of intermediates. (a) Accumulatintermediates.

39 mg L�1 and incident UV light intensity of 2.73 W L�1. In the cur-rent study, the ozone/UV system was designed to result in an aque-ous ozone concentration below the minimum detectable limit of5 lg L�1 by the Indigo method (Bader and Hoigne, 1981). There-fore, it can be inferred that the reaction of OH� with TBF and itsreaction intermediates is the main mechanism through which theyare oxidized in the ozone/UV process.

Several reaction intermediates identified during the degrada-tion of TBF, including TBA, HiBA, acetone (AC), pyruvaldehyde(PYD), formaldehyde (FAD), pyruvic acid (PYA), acetic acid(ACA), formic acid (FOA), and oxalic acid (OXA) are shown inFig. 2a and b. The time profiles indicate that TBA, HiBA, acetone,formaldehyde, and formic acid are primary intermediates of TBF,Fig. 2a. The accumulation of acetone is the highest among the ma-jor primary intermediates indicating that it is also formed fromthe degradation of the other primary intermediates, such as TBAand HiBA. The time profiles for formaldehyde and formic acid alsoshow that they are primary intermediate formed during the oxi-dation of TBF and further oxidation of primary intermediates,such as TBA, HiBA, and acetone. In addition, the results presentedinFig. 2a and b shows that the reaction intermediates formed dur-ing the oxidation of TBF react well in the ozone/UV process indi-cating that complete mineralization could be achieved in theprocess.

0.0

0.1

0.2

0.3

Prod

ucts

(mM

)

TBA

HiBA

AC

FAD

FOA

AC]o=0.50 mMO3)inf=21 mg/LO3)eff=10 mg/L

o=1.35 W/LHCO3

-]=2 mMH=7.0

g=1.5 L/min

40 50 60

nutes)

0.0

0.1

0.2

0.3

Prod

ucts

(mM

)

TBA

ACA

PYA

OXA

PYD

AC]o=0.50 mMO3)inf=21 mg/LO3)eff=10 mg/L

o=1.35 W/LHCO3

-]=2 mMH=7.0

g=1.5 L/min

ion and decay of primary intermediates. (b) Accumulation and decay of other

0.0

0.2

0.4

0.6

0.8

1.0

Ace

tone

(mM

)

0.0

0.1

0.2

0.3

0.4

Prod

ucts

(mM

)

AC

FAD

PYD

ACA

FOA

[AC]o=0.85 mM(O3)inf=39 mg/L(O3)eff=16 mg/LIo=2.73 W/L[HCO3

-]=2 mMpH=7.0Qg=1.5 L/min

0.0

0.2

0.4

0.6

0.8

1.0

0 10 20 30 40 50 60

Time (minutes)

Ace

tone

(mM

)

0.0

0.1

0.2

0.3

0.4

Prod

ucts

(mM

)

AC

PYA

OXA

[AC]o=0.85 mM(O3)inf=39 mg/L(O3)eff=16 mg/LIo=2.73 W/L[HCO3

-]=2 mMpH=7.0Qg=1.5 L/min

a

b

Fig. 4. Acetone oxidation and accumulation and decay of intermediates. (a) Accumulation and decay of primary intermediates. (b) Accumulation and decay of otherintermediates.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

Time (minutes)

TOC

(mM

)

0 10 20 30 40 50 60

Calculated TOC

Measured TOC

[TBF]o=1.03 mM(O3)inf=0.82 mM(O3)eff=0.26 mMIo=5.8x10-6 E/L-s[HCO3

-]=2 mMpH=7.0Qg=1.5 L/min

Fig. 5. Total organic carbon balance.

1712 T. Garoma et al. / Chemosphere 73 (2008) 1708–1715

T. Garoma et al. / Chemosphere 73 (2008) 1708–1715 1713

3.2. Degradation of primary intermediates

TBA and acetone were also oxidized in the ozone/UV processand their degradation products were identified. In Fig. 3a and b,the reaction intermediates formed during the oxidation of0.5 mM (37 mg L�1) of TBA by using an influent ozone gas of21 mg L�1 and incident UV light intensity of 1.35 W L�1 are pre-sented. The effluent ozone concentration was measured as15 mg L�1. At the start of the experiment, bicarbonate and pH ofthe solution were adjusted to 2 mM and 7.0, respectively. The timeprofiles of the reaction intermediates reveal that HiBA, acetone,formaldehyde, and formic acid are primary intermediates of TBAdegradation, Fig. 3a. These chemicals were also identified as theprimary intermediates of the parent compound TBF. This is ex-pected because the two chemicals share similar structure withthe only difference being the presence of a hydroxyl group (–OH)in TBA and a formyl group (–CHO) in TBF. Intermediates formedas a result of further oxidation of the primary intermediates ofTBA include pyruvaldehyde, acetic acid, pyruvic acid, and oxalicacid Fig. 3b.

Fig. 4a and b shows the decay of acetone and the reaction inter-mediates formed during an ozone/UV treatment of acetone. Theinitial concentration of acetone, influent ozone gas, bicarbonate,and pH of the solution were 0.85 mM (49 mg L�1), 39 mg L�1,2 mM, and 7.0, respectively. The solution was exposed to an inci-dent UV radiation of 2.73 W L�1 and the resulting effluent ozonegas concentration was 16 mg L�1. Formaldehyde, pyruvaldehyde,acetic acid, formic acid were identified as primary intermediates,Fig. 4a. Pyruvic acid and oxalic acid are formed as a result of furtheroxidation of the primary intermediates, Fig. 4b.

3.3. Total organic carbon (TOC) balance

In Fig. 5, the material balance for organic carbon during the oxi-dation of TBF by the ozone/UV process is presented. To check theorganic carbon balance, the molar concentration of organic carboncontained in TBF and its intermediates were added together andcompared with the TOC experimentally measured. The molar con-centration of organic carbon associated with each compound wascalculated by multiplying their molar concentration with the num-

3

|

|3

3

HC

OCHOCCH

HC

−−|

|3

C

CCH

C

−

Tert-butyl Formate (TBF) Tert-butyl Alc

3C HC3CH

O||

−− CCH

O ||

3 −−Acetone (AC) Pyruvaldehy

C HO3CH

O ||

−− CH

O ||

−−Acetic Acid (ACA) Formaldehy

OHCCHO

O O || ||

−− −Oxalic Acid (OXA)

Fig. 6. Structural formula of TBF a

ber of carbons contained in their structure. The results indicatethat about 90% organic carbon balance was achieved between thecalculated and measured TOC values. The discrepancy, especiallyin the 10–40 min region, could be due to the existence of a possibleintermediate that is not detected. However, the organic carbon bal-ance obtained is reasonable taking into account the complex nat-ure of radical reactions taking place in the ozone/UV system.

In the ozone/UV system, TOC removal followed pseudo-first-or-der kinetics with an observed rate constant of 4.9 � 10�4 s�1 for aninfluent ozone gas concentration of 39 mg L�1 and incident UVlight intensity of 2.73 W L�1. On the basis of this rate constant,95% TOC reduction could be achieved in approximately 100 minand 99% TOC reduction in about 150 min.

3.4. Degradation pathways of TBF

The structural formula for TBF and its intermediates identifiedare presented in Fig. 6. The degradation of TBF could be initiatedby an OH� attack at the formyl group (–CHO) or one of the equiva-lent methyl groups (–CH3) and could lead to the formation of TBAand HiBA, respectively. Additionally, abstraction of two methylgroups or methyl and formyl group in TBF could lead to the gener-ation of acetone. The methyl ðCH�3Þ and formyl (CHO�) radicalsformed, as a result of these abstractions, lead to the formation offormaldehyde and formic acid through series of reaction interme-diates (Stevens et al., 1972; Rabani et al., 1974), and upon oxida-tion, formaldehyde leads to formic acid (Chin and Wine, 1994).The degradation pathways proposed above agree with the timeprofiles of intermediates presented in Fig. 2a. These chemicalswere also identified as primary intermediates when TBF was oxi-dized using different advanced oxidation processes (Barreto et al.,1995; Stefan et al., 2000; Acero et al., 2001).

3.5. Degradation pathways for intermediates

The degradation of TBA can be initiated by an OH� attack at anyof the equivalent methyl groups or hydroxyl group and lead to theformation of HiBA, acetone, formaldehyde, and formic acid. Thetime profile presented in Fig. 3a also supports the degradationpathways proposed for TBA. Similarly, other studies using different

3

3

H

OH

H

−

3

|

|

3

HC

OHCCHO

HC

−−

ohol (TBA) Hydroxy-iso-Butyraldehyde (HiBA)

HC

O ||

− OHCCCH

O O || ||

3 −− −de (PYD) Pyruvic Acid (PYA)

H

H C HO

O ||

−−de (FAD) Formic Acid (FOA)

nd its reaction intermediates.

Fig. 7. Proposed degradation pathways of TBF.

1714 T. Garoma et al. / Chemosphere 73 (2008) 1708–1715

advanced oxidation processes (Stefan et al., 2000; Acero et al.,2001) reported HiBA, acetone, and formaldehyde as the primaryintermediates of TBA oxidation. No product study was conductedfor HiBA since it is not commercially available. Its degradationcould be initiated by an OH� attack at the methyl, carbonyl or hy-droxyl group and may lead to the formation of acetone, pyruvalde-hyde, formaldehyde, and formic acid as primary intermediates.

An OH� attack at either of the equivalent methyl groups in ace-tone can lead to the formation of pyruvaldehyde, acetic acid, form-aldehyde, and formic acid as primary intermediates. Theexperimental results presented in Fig. 4a support the proposeddegradation pathways. The chemicals were also identified as theprimary intermediates of acetone degradation in UV/H2O2 process(Stefan and Bolton, 1999).

The degradation of other reaction intermediates, such as pyruv-aldehyde, formaldehyde, acetic acid, pyruvic acid, and oxalic acidcan be initiated by OH� attack at methyl, carbonyl (–CHO), carboxyl(–COO–), or hydroxyl group. The carbonyl and carboxyl radicalsformed, by the abstraction process, can lead to the generation ofcarbon dioxide (Ilan and Rabani, 1976). The time profiles of formicacid in Figs. 2a, 3a and b show that it continued to tail during thecourse of oxidation unlike other primary intermediates formedfrom TBF, TBA, and acetone. Formic acid reacts with OH� at a veryhigh rate constant of 2.2 � 109 M�1 s�1 (Baxendale and Khan,1969) compared with that of TBF, TBA, acetone, and formaldehyde,and organic acids formed during the oxidation of TBF. The tailingcould be due the formation of formic acid from further oxidationof secondary or tertiary intermediates, such as pyruvaldehyde,pyruvic acid, and acetic acid. Fig. 7 summarizes the proposed deg-radation pathways for TBF.

3.6. Applicability of study

In this study, TBF was spiked in deionized water and the effectof scavengers of OH� other than TBF, reaction intermediates, andcarbonate ions was negligible. However, for real samples the effectof other scavengers needs to be taken into account. Further, thestudy was limited to water systems that do not contain UV absorb-ing species or target chemicals (and intermediates) that do not ab-sorb UV radiation.

4. Summary

This study investigated the removal of tert-butyl formate (TBF)from dilute aqueous solution using an ozone/UV process and iden-tified intermediates formed during the oxidation of TBF. tert-butylalcohol (TBA), hydroxy-iso-butyraldehyde (HiBA), acetone, formal-dehyde, and formic acid were identified as primary intermediates,whereas pyruvaldehyde, acetic acid, pyruvic acid, and oxalic acidwere generated in the system as a result of further oxidation pro-cesses. About 90% organic carbon balance was obtained during theoxidation of TBF indicating that most reaction intermediates havebeen identified and quantified. In addition, to establish the degra-dation pathways for TBF, some of the primary intermediates wereoxidized in the ozone/UV system. Accordingly, HiBA, acetone,formaldehyde, and formic acid were the primary intermediates ofTBA degradation. The oxidation of acetone in the ozone/UV systemgenerated formaldehyde, pyruvaldehyde, acetic acid, formic acid asprimary intermediates.

The ozone/UV system was very effective in degrading TBF. Itwas observed that the reaction intermediates formed during theoxidation of TBF react well in the ozone/UV system and completemineralization could be achieved by the process.

References

Acero, J.L., Haderlein, S.B., Schmidt, T.C., Suter, M.J.F., Von Gunten, U., 2001. MTBEoxidation by conventional ozonation and the combination ozone/hydrogenperoxide: efficiency of the processes and br bromate formation. Environ. Sci.Technol. 35, 4252–4259.

APHA, 1998. In: Clescerl, L., Greenberg, A., Eaton, A. (Eds.), Standard Methods forExamination of Water and Wastewater. Washington, DC.

Bader, H., Hoigne, J., 1981. Determination of ozone in water by the indigo method.Water Res. 15, 449–456.

Barreto, R.D., Gray, K.A., Anders, K., 1995. Photocatalytic degradation of methyl-tert-butyl ether in TiO2 slurries – a proposed reaction scheme. Water Res. 29, 1243–1248.

Baxendale, J.H., Khan, A.A., 1969. Pulse radiolysis of nitrosodimethylaniline inaqueous solution. Int. J. Radiat. Phys. Chem. 1, 11–24.

Bradley, P.M., Chapelle, F.H., Landmeyer, J.E., 2001. Methyl t-butyl ethermineralization in surface-water sediment microcosms under denitrifyingconditions. Appl. Environ. Microbiol. 67, 1975–1978.

Burbano, A.A., Dionysiou, D.D., Richardson, T.L., Suidan, M.T., 2002. Degradation ofMTBE intermediates using Fenton’s reagent. J. Environ. Eng. – ASCE 128, 799–805.

Chang, P.B.L., Young, T.M., 2000. Kinetics of methyl tert-butyl ether degradation andby-product formation during UV/hydrogen peroxide water treatment. WaterRes. 34, 2233–2240.

Chin, M., Wine, P.H., 1994. A temperature dependent competitive kinetics study ofthe aqueous-phase reaction of OH radicals with formate, formic acid, acetate,acetic acid, and hydrated formaldehyde. In: Helz, G.R., Zepp, R.G., Crosby, D.G.(Eds.), Aquatic and Surface Photochemistry. Lewis Publishers, pp. 85–96.

Church, C.D., Pankow, J.F., Tratnyek, P.G., 1999. Hydrolysis of tert-butyl formate:kinetics, products, and implications for the environmental impact of methyltert-butyl ether. Environ. Toxicol. Chem. 18, 2789–2796.

Cirvello, J.D., Radovsky, A., Heath, J.E., Farnell, D.R., Lindamood, C., 1995. Toxicityand carcinogenicity of t-butyl alcohol in rats and mice following chronicexposure in drinking-water. Toxicol. Ind. Health 11, 151–165.

Damm, J.H., Hardacre, C., Kalin, R.M., Walsh, K.P., 2002. Kinetics of the oxidation ofmethyl tert-butyl ether (MTBE) by potassium permanganate. Water Res. 36,3638–3646.

EPA, 2004. State Actions Banning MTBE (Statewide), EPA-420-B-04-009,Washington, DC, <http://www.epa.gov/mtbe/420b04009.pdf>.

Finneran, K.T., Lovley, D.R., 2001. Anaerobic degradation of methyl tert-butylether (MTBE) and tert-butyl alcohol (TBA). Environ. Sci. Technol. 35, 1785–1790.

Francois, A., Garnier, L., Mathis, H., Fayolle, F., Monot, F., 2003. Roles of tert-butylformate, tert-butyl alcohol and acetone in the regulation of methyl tert-butylether degradation by Mycobacterium austroafricanum IFP 2012. Appl. Microbiol.Biotechnol. 62, 256–262.

Froines et al., 1998. An evaluation of the scientific peer-reviewed research andliterature on the human health effects of MTBE, its metabolites, combustionproducts and substitute compounds: In: Report to the Governor and Legislatureof the State of California, Sacramento, California, <https://tsrtp.ucdavis.edu/public/mtbe/mtberpt/vol2.pdf>.

Garoma, T., Gurol, M.D., 2004. Degradation of tert-butyl alcohol in dilute aqueoussolution by an O3/UV process. Environ. Sci. Technol. 38, 5246–5252.

Garoma, T., Gurol, M.D., 2005. Modeling aqueous ozone/UV process using oxalicacid as probe chemical. Environ. Sci. Technol. 39, 7964–7969.

T. Garoma et al. / Chemosphere 73 (2008) 1708–1715 1715

Garoma, T., Gurol, M.D., 2006. Oxidation of methyl tert-butyl ether inaqueous solution by an ozone/UV process. J. Environ. Eng. – ASCE 132, 1404–1412.

Haag, W.R., Yao, C.C.D., 1992. Rate constants for reaction of hydroxyl radicalswith several drinking-water contaminants. Environ. Sci. Technol. 26, 1005–1013.

Hatchard, C.G., Parker, C.A., 1956. A new sensitive chemical actinometer II.Potassium ferrioxalate as a standard chemical actinometer. Proc. R. Soc. Lond.,Ser. A, Math. Phys. Sci. A235, 518–536.

Hoigne, J., Bader, H., 1983a. Rate constants of reactions of ozone with organic andinorganic-compounds in water 1. Non-dissociating organic-compounds. WaterRes. 17, 173–183.

Hoigne, J., Bader, H., 1983b. Rate constants of reactions of ozone with organic andinorganic-compounds in water 2. Dissociating organic-compounds. Water Res.17, 185–194.

Ilan, Y., Rabani, J., 1976. Some fundamental reactions in radiation-chemistry –nanosecond pulse-radiolysis. Int. J. Radiat. Phys. Chem. 8, 609–611.

Mehlman, M.A., 2002. Carcinogenicity of methyl-tert butyl ether in gasoline. Ann. N.Y. Acad. Sci. 982, 149–159.

Mormile, M.R., Liu, S., Suflita, J.M., 1994. Anaerobic biodegradation of gasolineoxygenate – extrapolation of information to multiple sites and redoxconditions. Environ. Sci. Technol. 28, 1727–1732.

Peyton, G.R., Glaze, W.H., 1988. Destruction of pollutants in water with ozone incombination with ultraviolet-radiation 3. Photolysis of aqueous ozone. Environ.Sci. Technol. 22, 761–767.

Rabani, J., Klugroth, D., Henglein, A., 1974. Pulse radiolytic investigations ofOHCH2O2 radicals. J. Phys. Chem. 78, 2089–2093.

Sandmeyer, E.E., Kirwin Jr., C.J., 1981. Industrial toxicology. Esters. Patty’s Ind. Hyg.Toxicol. 2A, 2259–2412 (third revis. ed.).

Somsamak, P., Cowan, R.M., Haggblom, M.M., 2001. Anaerobic biotransformation offuel oxygenates under sulfate-reducing conditions. Fems. Microbiol. Ecol. 37,259–264.

Staehelin, J., Hoigne, J., 1982. Decomposition of ozone in water – rate of initiation byhydroxide ions and hydrogen-peroxide. Environ. Sci. Technol. 16, 676–681.

Stefan, M.I., Bolton, J.R., 1999. Reinvestigation of the acetone degradationmechanism in dilute aqueous solution by the UV/H2O2 process. Environ. Sci.Technol. 33, 870–873.

Stefan, M.I., Mack, J., Bolton, J.R., 2000. Degradation pathways during the treatmentof methyl tert-butyl ether by the UV/H2O2 process. Environ. Sci. Technol. 34,650–658.

Stevens, G.C., Hart, E.J., Clarke, R.M., 1972. Radiolysis of aqueous methane solutions.J. Phys. Chem. 76, 3863–3867.

Yeh, C.K., Novak, J.T., 1994. Anaerobic biodegradation of gasoline oxygenates insoils. Water Environ. Res. 66, 744–752.