Decomposition of Trichlorobenzene Isomers by Co-Grinding with CaO

Article abstract of DOI:10.1246/bcsj.76.1919

Two types of trichlorobenzene (TCB) isomers, 1,2,3- and 1,3,5-, were subjected to grinding in air with CaO in a planetary ball mill to investigate their decomposition phenomena. Results showed that TCB samples were decomposed by the co-grinding to form calcium chloride hydrate and amorphous carbon as the main products, with other minor compositions such as carbon dioxide and water. There was no significant difference in the decomposition phenomena between the two isomers. Radicals were generated during the decomposition process, predominantly due to the rupture of C-Cl bonds in the molecule. Supported by the simulation data, we concluded that dehydrochlorination during TCB decomposition process to form hydrochloride (HCl, further to CaOHCl) exhibited the lower possibility, suggesting dechlorination rather than dehydrochlorination as the plausible decomposition pathway.

Full text of DOI:10.1246/bcsj.76.1919

Ó 2003 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 76, 1919–1925 (2003) 1919
Decomposition of Trichlorobenzene Isomers by Co-Grinding with CaO
ꢀ
;y
yy
y
Yasumitsu Tanaka, Qiwu Zhang, Koichi Mizukami, and Fumio Saito
Material Solution Development Dept. Recording Media Company, SONY Sendai Technology Center,
-4-1, Sakuragi, Tagajo 985-0842
yInstitute of Multidisciplinary Research for Advanced Materials (IMRAM), Tohoku University,
-1-1, Katahira, Aoba-ku, Sendai 980-8577
3
2
yyIndustrial Technology Institute, Miyagi Prefectural Government, 2-2, Akedohri, Izumi-ku, Sendai 981-3206
Received April 21, 2003; E-mail: zhangqw@tagen.tohoku.ac.jp
Two types of trichlorobenzene (TCB) isomers, 1,2,3- and 1,3,5-, were subjected to grinding in air with CaO in a
planetary ball mill to investigate their decomposition phenomena. Results showed that TCB samples were decomposed
by the co-grinding to form calcium chloride hydrate and amorphous carbon as the main products, with other minor com-
positions such as carbon dioxide and water. There was no significant difference in the decomposition phenomena be-
tween the two isomers. Radicals were generated during the decomposition process, predominantly due to the rupture of
C–Cl bonds in the molecule. Supported by the simulation data, we concluded that dehydrochlorination during TCB de-
composition process to form hydrochloride (HCl, further to CaOHCl) exhibited the lower possibility, suggesting de-
chlorination rather than dehydrochlorination as the plausible decomposition pathway.
The incineration of chlorinated organic compounds or halo-
gen-containing substances has some potential to form hazard-
ous compounds such as harmful gases and dioxins. This has
initiated interest in the development of decomposition technol-
ogies for such chlorinated materials and wastes. Several meth-
ods have been proposed, but inherent problems with safety, re-
liability, operational costs and the like persist. Recently,
Zhang et al. have proposed a method for decomposing such
halogenated organic compounds: namely, a mechanochemical
method based on grinding with an inorganic material such as
CaO. The decomposition is induced through the solid-state re-
namely CaOHCl as product, can be obtained than from any
other chlorinated aromatic compounds, if the so-called dehy-
drochlorination occurred.
Such phenomena were elucidated further by a computation-
al chemical process, which is a useful tool for understanding
intermediate states and unstable conditions during chemical
reactions. In addition, this process is advantageous for han-
dling hazardous and toxic substances as well as directly ana-
8–12
lyzing the phenomena which are happening.
There have
been two typical methods proposed, namely a classical me-
chanics method and a quantum mechanics method. And the
quantum mechanics method is classified into a non-experimen-
tal method and a semi-empirical molecular method. The for-
mer is a kind of non-empirical method called an ab initio
method, that enables us to reveal reaction mechanisms which
are related to various electronic and excitation states between
two components such as metallic and organic ones without any
1–5
action between the organic and inorganic compounds, and
when the chlorinated organic compound is, for example,
ground with CaO, the chlorines are transformed into inorganic
chloride. All the same, the mechanochemical method is a non-
thermal process, similar to those for the photo-induced and ra-
6,7
diation-induced catalytic methods.
However, a decomposi-
13
tion mechanism based on a mechanochemical method remains
to be determined. It is easy to imagine that such a decompo-
sition of chlorinated organics on the surfaces of inorganic ad-
ditives would be very complex so that it would be difficult to
interpret the phenomena observed and the reaction mechanism
by one simple pathway. In order to simplify the problem, dis-
cussion on the reaction mechanism was focused on whether it
proceeded through a dechlorination or a dehydrochlorination,
since such a controversy was met before during our work on
the decomposition of chlorobiphenyl, a model sample of
PCB and dioxin, and needs further investigation. Here, tri-
chlorobenzenes (C6H3Cl3, TCB) were chosen as the target
sample for several reasons. A solid state of TCB did not limit
the amount used, which occurred in the case of PCB of liquid
state, because a dry grinding is preferred. Another important
reason came from the result that a greater amount of HCl,
experimental parameters.
However, the method is not al-
ways suitable for a compound with large molecular weight be-
cause it requires a lengthy time for calculation. In contrast, the
latter method, called a semi-empirical molecular method, is
suitable for compounds with large molecular weight like or-
ganic molecules, while it can also be successfully used for in-
14–16
organic compounds.
The purpose of this paper is to provide information on a
TCB decomposition reaction based on both experimental re-
sults and computational chemical simulations. As a result,
the decomposition of TCB is interpreted through a pathway
of dechlorination rather than dehydrochlorination. It is expect-
ed that a combination of computational chemical simulation
and practical experimental method will prove useful for this
kind of study.
Published on the web October 15, 2003; DOI 10.1246/bcsj.76.1919

1
920 Bull. Chem. Soc. Jpn., 76, No. 10 (2003)
Decomposition of Trichlorobenzene
Experimental
Experimental Procedure and Characterizations. 1,3,5-TCB
sample (Tokyo Kasei Kogyo, Ltd., Japan) and 1,2,3-TCB sample
Wako Pure Chemicals Ltd., Japan) were used. The TCB samples
were mixed with the CaO powder at a 1:12 molar ratio. The CaO
(
sample was prepared by heating calcium hydroxide (Ca(OH)2
ꢁ
(
Wako Pure Chemical Industries, Ltd. Japan)) at 800 C for 2 h
in air in an electric furnace. The mixture of 2 g (0.43 g TCB +
.57 g CaO) was ground by a planetary ball mill (Fritsch Pulveri-
1
sette-7, Germany) in air at 700 rpm for various periods of time.
The mill is composed of pair of pots made of zirconia; each pot
3
has an inner volume of 45 cm . Seven zirconia balls of 15 mm
diameter were charged in the pot together with the mixture, and
then the mill was operated. To prevent excess heating, the milling
was stopped for 15 min after every 15 min grinding operation.
The ground samples were characterized as follows: X-ray dif-
fraction (XRD) (RAD-B system, Rigaku, Japan) using Cu-Kꢀ ra-
diation was used to identify phases formed in the ground product.
Raman spectra (Labspec Raman spectrograph, Horiba) were re-
corded at room temperature with a helium laser beam at 632 nm
line. Samples for ESR analysis were charged in a quartz tube
of 5 mm diameter, and measurements were carried out with an
X-band ESR spectrometer (Bruker ESP-380E).
The gas composition was analyzed by GC/MS. The gas was
drawn by a syringe just after stopping the mill and inserted into
the spectrometer for analysis using a Hewlett-Packard gas chro-
matograph (model 6890) equipped with a mass selective detector
Fig. 1. Yield of TCB remaining in the ground mixture as a
function of grinding time.
(
model 5973). The ground samples were also washed with water,
acetone, or toluene as follows: 0.5 g samples were agitated in 100
mL distilled water or 50 mL acetone or toluene with a magnetic
stirrer for 30 min to extract chemicals into the solution. After
filtration, the filtrates were subjected to ion chromatography (IC)
(LC10series, Shimadzu Co. Ltd.) and GC/MS analysis (the same
spectrometer).
Computational Chemical Work. The semi-empirical molec-
ular orbital method was used during the computational chemical
calculation. The calculation software was winMOPAC ver. 3.0
(
semi-empirical Hamiltonian.
FUJITSU). PM3 (Parametric Method-3) was adapted for the
Original structures of the mole-
17,18
cules were calculated by geometry optimization methods. The
same method was adapted for the calculation of bond energy, ex-
cept for the aimed bond. The length of the bond was fixed for 0.02
nm longer than the original structure, and the difference from orig-
inal internal energy was defined for each ‘‘bond energy’’. This dif-
ference also represents the strength of the bond. Furthermore, the
reaction barrier was calculated by the same procedure. The length
of the bond was stretched until it was 0.2 nm longer than the orig-
inal structure, and the internal energy was calculated every 0.02
nm. The reaction barrier was defined as the maximum difference
from the original internal energy. In this paper, the highest reac-
tion energy until 0.2 nm was defined for each ‘‘barrier energy’’. If
the difference energy did not converge until 0.2 nm, the reaction
energy was equal to the energy of the structure whose bond length
was 0.2 nm longer than the original structure, defined as the ‘‘final
energy’’.
Fig. 2. Yield of water-soluble Cl in the ground mixture as a
function of grinding time.
Results and Discussion
ing the prolonged grinding. Finally, the yield reaches 0% at 6
h grinding. This indicates that the TCB samples can be de-
composed under this grinding condition within 6 h. Between
the two TCBs, there exists little difference in decomposition
yield and other experimental results, although this data is not
Experimental Part. Figure 1shows the yield of the re-
maining TCBs in the ground samples as a function of the
grinding time. As shown in the figure, the yields decrease rap-
idly with the grinding time in the early stage, and level off dur-

Y. Tanaka et al.
Bull. Chem. Soc. Jpn., 76, No. 10 (2003) 1921
shown here. Therefore, the following results resulting from
the 1,3,5-TCB sample are presented.
original TCB sample has been observed, together with other
very small peaks. When the grinding time is extended to 6
h, only little peaks due to benzaldehyde (C6H5CHO) and
1,3,5-TCB are detected from the spectrum. Zero percent to
the original TCB was obtained by the calculation based on
the peak area, indicating that the decomposition of TCB sam-
ple has been almost completely achieved by this time. The re-
sults obtained by using acetone as solvent are almost the same
as the above results.
From the spectrum of 1h ground sample, several small
peaks are observed, suggesting the existence of some inter-
mediate phases from the TCB decomposition. Since the effect
of tiny impurities from the solvent itself can not be ruled out
completely, it is more reasonable to directly analyze the com-
positions of gases in mill pot to get the information about the
intermediate phases. Figure 4 shows the GC/MS spectra of
the gases collected from the pot just after grinding. It is very
interesting to note that many compounds such as TCB, tetra-
chlorobenzene (TeCB), dichlorobenzene (DCB), chloroben-
zene (CB), toluene (C6H5CH3) and benzene (C6H6), are ob-
served from the gas at 1h grinding. And these phases have
disappeared when the grinding time is extended to 6 h, where
only a little peak due to TCB is observed to remain. The anal-
ysis using different columns showed that gases with low
molecular weight, such as H2O, ethane (C2H6), ethylene
Figure 2 shows the yield of water-soluble chloride formed in
the ground product as a function of grinding time. The yield is
defined as (Clm/Clt), where Clm denotes the concentration of
Cl ions in the filtrate measured and Clt is the concentration
of Cl in the TCB sample. As has been indicated in the figure,
the yield increases rapidly with an increase in grinding time
and reaches about 95% within 6 h grinding. It is clear that
the organic chlorine has been changed into an inorganic wa-
ter-soluble compound, consistent with the results shown in
Fig. 1.
Figure 3 shows GC/MS spectra of the substances extracted
by toluene from the TCB ground samples. In the spectrum of
the sample ground for 1h, a strong peak corresponding to the
Fig. 4. GC/MS spectra of gases collected from the mill pot
after 1h and 6 h grinding. A: Trichlorobenzene, B: Ben-
zene, C: Toluene, D: Chlorobenzene, E: Dichlorobenzene,
F: Tetrachlorobenzene.
Fig. 3. GC/MS spectra of the substances extracted from
ground products by toluene. A: Trichlorobenzene, G:
Benzaldehyde.

1
922 Bull. Chem. Soc. Jpn., 76, No. 10 (2003)
Decomposition of Trichlorobenzene
(C2H4) (only 1h ground sample), carbon dioxide (CO 2) and
methane (CH4), were formed irrespective of the grinding
time. Since no gas pressure has been observed, these gases
should be present in tiny amounts. Except the tiny amount
of TCB, all other chlorine-containing intermediates have been
found to disappear by the prolonged grinding up to 6 h. Both
GC/MS data sets shown in Figs. 3 and 4 confirm the existence
of so many intermediate phases, suggesting that the mechano-
chemical process of decomposition is very complex, possibly
with various pathways. Fortunately, it separates chlorine from
benzene ring at the beginning of decomposition and these in-
termediate compounds have disappeared from the sample by
prolonged grinding, indicating that they are also decomposed
together with the TCB during the further grinding operation.
Figure 5 shows the Raman spectra of the TCB ground
À1
samples. The Raman shifts at 1570 and 1310 cm bands from
are observed and the peak intensity
19–21
amorphous carbon
gets higher with the increase in grinding time, indicating the
occurrence of carbonization during the decomposition
process. The carbon formation has been observed from the
mechanochemical decompositions of organic polymers such
22
as polytetrafluoroethylene (PTFE) and hexabromobenzene
(
2
HBB), where hydrogen does not exist. In this case, although
hydrogen exists in TCB, 6 h grinding operation also changes
the TCB into amorphous carbon. The spectrum of 1h ground
sample exhibits a very high background. This may be attrib-
uted to the existence of organic compositions which function
as the impurities to CaO to raise the fluorescence for Raman
measurement. Prolonged grinding to 6 h results in a low
and flattened baseline due to the less fluorescence, indicating
that almost no organic impurities remain in the sample. Both
results by Raman analysis and water-soluble chlorine measure-
Fig. 6. ESR spectra of 1h and 6 h ground samples.
ment have shown that organic TCB has been mineralized com-
pletely by forming inorganic chloride and amorphous carbon.
Figure 6 shows the ESR spectra of the ground samples.
Radical signals are clearly observed and their intensity increas-
es with an increase in grinding time. Although detailed inter-
pretation on the compositions of the radical needs further re-
search work, the related information on the radicals formed
in a similar way reported previously could be taken as
1,2,23
references.
The broad peak around 3475–3500 B/G could
be attributed to organic radicals and the sharp ones may result
from the trapped free electrons inside inorganic oxide. The
concentrations of both radicals rose up with the progress of
grinding; especially the former became higher than the latter.
It is supposed that radicals are induced by grinding on CaO
surfaces and subsequently the charge transfer occurs to organic
TCB. And C–Cl bonding in TCB tends to decompose to pro-
duce some intermediates or products. It can be said that the
mechanochemical decomposition is related with the formation
of radicals.
The above data have clearly shown that the decomposition
of TCB is achieved by the mechanochemical reaction with
CaO. Because various types of chlorinated compounds may
be involved during the practical application of the mechano-
chemical process, a deeper understanding toward the decom-
position mechanism is required in order to assure the effective-
ness of the process to any type of compound. From the above
data, it seems that the pathway of decomposition involves sev-
eral possibilities. Here the problem is considered based on
whether it is dechlorination or dehydrochlorination.
If the TCB sample is presumed to decompose mechano-
chemically through the dehydrochlorination pathway as fol-
lows:
Fig. 5. Raman spectra of the mixtures ground for 1h and 6 h.

Y. Tanaka et al.
Bull. Chem. Soc. Jpn., 76, No. 10 (2003) 1923
exchange is induced mechanochemically to form Ca(OH)2 and
CaSO4 0.5H2O (gypsum hemihydrate), as is clear by phase
ꢂ
identification. This can help one to understand why the state
of chloride is amorphous and the formation of carbon occurs.
The existence of radicals also offers more evidence that
dechlorination is more reasonable than dehydrochlorination.
The dechlorination pathway means the separation of C–Cl
bonding, leaving free radicals in the ground sample. The de-
hydrochlorination pathway implies that both C–Cl and C–H
bondings are disconnected simultaneously to form HCl absor-
bed by CaO as CaOHCl and an intermediate phase with a
structure similar to benzyne. This will leave less chance for
the formation of free radicals.
Furthermore, based on the supposed dehydrochlorination
pathway, it is difficult to interpret the formations of other inter-
mediate phases such as TeCB, DCB, CB, toluene, benzene and
gases with low molecular weight, such as H2O, C2H6, CO2,
CH4. The joint cut-off of chlorine and hydrogen from benzene
ring leads to the possible intermediate phases of C6H2Cl2 and
C6H1Cl1. The formations of the above compounds seem to
support the notion of a separate dechlorination of Cl rather
than the joint disconnection of H and Cl from the benzene
ring. Therefore, a dechlorination pathway rather than a dehy-
drochlorination pathway is reasonable. This will be discussed
again based on the following results from computational chem-
ical simulations.
Fig. 7. XRD patterns of 1h and 6 h ground samples.
C6H3Cl3 þ CaO ! C6H2Cl2 þ CaOHCl
C6H2Cl2 þ CaO ! C6H1Cl1 þ CaOHCl
C6H1Cl1 þ CaO ! 6C þ CaOHCl
ð1Þ
ð2Þ
ð3Þ
Simulation Part Based on Computer Chemistry. To
confirm this experimental data, we used a semi-empirical
molecular orbital method to simulate the TCB decomposition
process. Due to the difficulty in calculating the direct bonding
between Cl and Ca, the purpose of the calculation was focused
on the difference in easiness of C–Cl separation from that of
the joint separations of C–Cl and C–H to form HCl, without
considering the effect of CaO existence.
And three steps of dehydrochlorination may result in the
formation of carbon, possibly explaining the reason for the car-
bon occurrence. In fact, the mechanochemical decomposition
of poly(vinyl chloride) has been observed to occur through a
dehydrochlorination pathway and the crystalline phase of
Table 1shows the ‘‘bond energy’’, ‘‘barrier energy’’ and ‘‘fi-
nal energy’’ values of a 1,3,5-TCB model molecule. The bond
energy is defined as the increasing internal molecular energy
24,25
CaOHCl has been detected in the ground sample.
Howev-
er, the formation of crystalline CaOHCl here has not been ob-
served from the XRD analysis results shown in Fig. 7, suggest-
ing that the chloride formed is a amorphous one. In this case,
XRD patterns show only CaO existence both in 1h and 6 h
ground samples. The peak intensity of 6 h ground sample is
smaller than that of 1h, indicating the progress of mechano-
chemical reaction and decomposition of TCB. The amorphous
state of chloride can not offer direct information about prod-
ucts.
Taking together the analytical results, we can make some
deductions about the compositions of the ground products.
When Cl is separated from C, it tends to absorb CaO. On
the other hand, the O of CaO tends to form combination with
TCB, via hydrogen bonding for example. The exchange be-
tween the combination results in the Ca chloride and an organ-
ic phase consisting of C, H and O. Ca chloride has a strong
tendency to absorb water to form a hydrate. On the other hand,
the organic phase with C, H and O tends to dehydrate during
grinding. As a result, Ca chloride hydrate and amorphous car-
bon are obtained as the most stable state through water
exchange. It may be difficult to imagine such a water ex-
change between organic and inorganic compounds. In fact,
Table 1. 1,3,5-Trichlorobenzene Internal Energy
Bond
energy
kJ mol
Barrier
energy
/kJ mol
Final
energy
/kJ mol
Possibility
products
À1
À1
À1
/
C–C bond
1
2
3
4
C–2C
C–3C
C–4C
C–5C
64
64
64
64
64
64
699
699
699
699
699
699
497
497
497
497
497
497
7 nucleus
7 nucleus
7 nucleus
7 nucleus
7 nucleus
7 nucleus
5C–6C
6C–1C
C–H bond
1
3
5
C–7H
C–9H
C–11C
44
44
44
—
—
—
771
771
771
C–Cl bond
2
4
6
C–8Cl
C–10Cl
C–12Cl
29
29
29
—
—
—
558
558
558
À1
if CaO and CaSO4 2H2O (gypsum) are ground together, water
1,3,5-TCB: original energy 16.3 kJ mol
.
ꢂ

1
924 Bull. Chem. Soc. Jpn., 76, No. 10 (2003)
Decomposition of Trichlorobenzene
Table 2. 1,2,3-Trichlorobenzene Internal Energy
Bond
energy
kJ mol
Barrier
energy
/kJ mol
Final
energy
/kJ mol
Possibility
products
À1
À1
À1
/
C–C bond
1
2
3
4
5
6
C–2C
C–3C
C–4C
C–5C
C–6C
C–1C
64
64
63
63
63
63
—
751251
711
592
592
—
761
249
390
390
745
7 nucleus
7 nucleus
C–H bond
1
3
5
C–7H
C–9H
C–11C
44
44
44
771
771764
771
380
380
HCl
HCl
Fig. 8. Internal energy in C–C, C–H and C–Cl bonds in the
model molecule of 1,3,5-TCB calculated by the semi-em-
pirical method as a function of the stretched length be-
tween neighboring elements.
C–Cl bond
2
4
6
C–8Cl
C–10Cl
C–12Cl
29
29
29
—
—
—
564
573
564
when each bond distance (C–H, C–Cl and C–C) is stretched
À1
À1
until 0.02 nm. The bond energy of C–Cl is 29 kJ mol , which
1,3,5-TCB: original energy 23.6 kJ mol
.
is the lowest one of all. This means that the C–Cl bond is the
weakest bond, where the decomposition likely started. This
simulation data also reveal that the decomposition is due to the
dechlorination because the energy of C–H is higher than C–Cl.
Furthermore, the 1,3,5-TCB model molecule has no chance to
release HCl from the structure, even if chlorines in the 1,3,5-
TCB model molecule are located at an ideal position. There
is a chance for Cl, H or hydrocarbons to combine with another
benzene to form TeCB, TCB, DCB and other chemicals.
Figure 8 shows the bond energy for C–C, C–H and C–Cl in
the 1,3,5-TCB model molecule as a function of stretched
length. In general, the binding force between the two-neighbor
elements is weak when the internal energy is low at the same
stretched length. This means that the molecule is chemically
unstable. The figure indicates that the C–Cl bond is located
in the lowest side in the range of stretched length. This is very
consistent with the experimental results shown in Figs. 1and 2.
In contrast, Table 2 shows internal energy values of 1,2,3-
TCB as a function of stretched length. Final energies of 1C–
Fig. 9. Internal energy in C–C, C–H and C–Cl bonds in the
model molecule of 1,2,3-TCB calculated by the semi-em-
pirical method as a function of the stretched length be-
tween neighboring elements.
7H and 3C–9H are lower than barrier energies. This means
that Cl may separate together with H to form HCl. When
7
7
H and 9H are stretched continuously, HCl (consisting of
H–12Cl, 9H–10Cl) can be formed at the long stretched length
of around 0.18 nm. However, product energies of HCl are too
big for practically forming HCl for the 1,2,3-TCB model
molecule. Therefore, it is reasonable to say that the 1,2,3-
TCB decomposition is caused by the release of Cl from the
benzene ring structure of the molecule. Figure 9 shows calcu-
lated potential energy between each two neighboring elements,
such as C–Cl, C–H and C–C in the model molecule of 1,2,3-
TCB, as a function of the stretched length when extended by
certain external force. It is also found from the figure that
the C–Cl bond is the weakest one of all, implying that the de-
composition starts from this bond in the model molecule.
ena experimentally. In addition, the simulation on the poten-
tial energy of each bond in the TCB model molecule was con-
ducted based on the computational chemical simulation. The
results are summarized as follows:
1) The TCB can be decomposed mechanochemically by the
grinding with CaO through dechlorination, while is in-
duced by the charge transfer generating radicals.
2) The final product is composed of Ca chloride hydrate and
amorphous carbon, with minute amounts of several kinds
of stable gases.
3) The simulation on the potential energy of the TCB model
molecule reveals that the molecule is decomposed by
dechlorination. This is consistent with the experimental
result.
Conclusions
Trichlorobenzene was ground in air with CaO by using a
planetary ball mill to investigate the decomposition phenom-

Y. Tanaka et al.
Bull. Chem. Soc. Jpn., 76, No. 10 (2003) 1925
The authors are grateful to Dr. Tadaaki Ikoma, Tohoku
University, for his kind help with the radical measurement as
well as for invaluable discussions.
Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson,
P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. Cioslowki, J. V. Ortiz, B. B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.
Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe, P.
M. W. Gill, B. G. Johnson, W. Chen, M. W. Wong, J. L. Andres,
M. Head-Gordon, E. S. Replogle, and J. A. Pople, ‘‘GAUSSIAN
98 (Revision A.10),’’ Gaussian, Inc., Pittsburgh, PA (1998).
14 C. J. Chen and J. W. Bozzelli, J. Phys. Chem. A, 103, 9731
(1999).
References
1
Q. W. Zhang, F. Saito, T. Ikoma, S. Tero-kubota, and K.
Hatakeda, Environ. Sci. Technol., 35, 4933 (2001).
Q. W. Zhang, H. Matsumoto, F. Saito, and M. Baron, Che-
mosphere, 48, 787 (2002).
2
3 H. Mio, S. Saeki, J. Kano, and F. Saito, Environ. Sci. Tech-
nol., 36, 1344 (2002).
4 G. Gao, S. Doppiu, M. Monagheddu, R. Orru, M. Sannia,
and G. Cocco, Ind. Eng. Chem. Res., 38, 3218 (1999).
15 M. A. Noble, S. Gul, C. S. Verma, and K. Brocklehurst,
Biochem. J., 351, 723 (2000).
16 F. Ragot, V. Ivanov, J. Wery, A. Garcia, D. L. Perry, G.
Ouvrard, and E. Faulques, Synth. Met., 120, 773 (2001).
17 J. J. P. Stewart, J. Comput. Chem., 10, 209 (1989).
18 J. J. P. Stewart, Special issue —MOPAC— a semi-empiri-
cal molecular-orbital program, J. Comput.-Aided Mol. Des., 4, 1
(1990).
5
S. A. Rowlands, A. K. Hall, P. G. MaCormick, R. Street, J.
Hart, G. F. Ebell, and P. Donecker, Nature, 367, 223 (1994).
T. M. EL-Morsi, W. R. Budakowski, A. S. Abe-EL-Azis,
and K. J. Friesen, Environ. Sci. Technol., 34, 1018 (2000).
G. A. Zacheis, K. A. Gray, and P. V. Kamat, J. Phys.
Chem. B, 103, 2142 (1999).
6
7
19 H. Wilhelm, M. Lelaurain, E. McRae, and B. Humber, J.
Appl. Phys., 84, 6552 (1998).
8 E. A. Nantsis and W. R. Carper, J. Mol. Struct.: THEO-
CHEM, 468, 51( 19 98).
20 H. Wakayama, J. Mizuno, Y. Fukushima, K. Nagano, T.
Fukunaga, and U. Mizutani, Carbon, 37, 947 (1999).
21F. Salver-Disma, A. Du Pasquier, J. M. Tarascon, J. C.
Lassegues, and J. N. Rouzaud, J. Power Sources, 81–82, 291
(1999).
22 Q. W. Zhang, H. Matsumoto, and F. Saito, Chem. Lett.,
2001, 148.
23 T. Ikoma, Q. W. Zhang, F. Saito, K. Akiyama, and S. T.
Kubota, Bull. Chem. Soc. Jpn., 74, 2303 (2001).
24 Q. W. Zhang, F. Saito, K. Shimme, and S. Masuda, J. Soc.
Powder Technol., Jpn., 36, 468 (1999).
9 K. A. Nguyen and R. Pachter, J. Chem. Phys., 114, 10757
2001).
(
1
0
Y. Ohkatsu, S. Ishikawa, and E. Tobita, Polym. Degrad.
Stab., 67, 541(2000).
M. G. Organ, V. Dragan, M. Miller, R. D. J. Froese, and J.
D. Goddard, J. Org. Chem., 65, 3666 (2000).
Y. Shinohara, S. Mae, D. Shouro, and T. Nakajima, J. Mol.
Struct:. THEOCHEM, 497, 1(2000).
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
1
1
1
2
1
3
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A.
Montgomery, R. E. Stratmann, J. C. Burant, S. Dapprich, J. M.
Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C.
25 S. Saeki, J. Kano, F. Saito, K. Shimme, S. Masuda, and T.
Inoue, J. Mater. Cycles Waste Manage., 3, 20 (2001).