Photocatalysis of Chloroform Decomposition by Tetrachlorocuprate (II) on Dowex 2-X8

Article abstract of DOI:10.1111/php.12336

Heterogenized on a polystyrene anion exchange resin and in the presence of oxygen, CuCl24 catalyzes the photodecomposition of chloroform at wavelengths above 345 nm with greater efficiency than an equivalent amount in homogeneous solution. The reaction is proposed to proceed in two stages, the first stage yielding CCl4 and HO2 as products, the second consisting of a chain reaction resulting from the CuCl2 4-catalyzed photodissociation of CCl4, yielding phosgene with CCl3 radicals as chain carriers. Photodecomposition is retarded by added Cl, CH3CN, C6H12 or C2H5OH, which is ascribed to the displacement of CHCl3 molecules from the vicinity of the copper by attraction to the polystyrene matrix or to the alkylammonium cation sites.

Full text of DOI:10.1111/php.12336

Photochemistry and Photobiology, 2014, 90: 1234–1242
Photocatalysis of Chloroform Decomposition by Tetrachlorocuprate (II) on
Dowex 2-X8
Brent M. Harvey and Patrick E. Hoggard*
Department of Chemistry, Santa Clara University, Santa Clara, CA, USA
Received 28 April 2014, accepted 11 August 2014, DOI: 10.1111/php.12336
ABSTRACT
gas phase, which takes place only at wavelengths below approxi-
mately 260 nm, have been worked out reasonably well (8).
Hydrogen abstraction by each of the photogenerated radicals leads
to the formation of phosgene and hydrogen chloride.
Heterogenized on a polystyrene anion exchange resin and in
the presence of oxygen, CuCl24ꢀ catalyzes the photodecompo-
sition of chloroform at wavelengths above 345 nm with
greater efficiency than an equivalent amount in homogeneous
solution. The reaction is proposed to proceed in two stages,
the first stage yielding CCl4 and HOꢀ2 as products, the second
consisting of a chain reaction resulting from the CuCl24ꢀ-cata-
lyzed photodissociation of CCl4, yielding phosgene with CCl3
radicals as chain carriers. Photodecomposition is retarded by
added Clꢀ, CH3CN, C6H12 or C2H5OH, which is ascribed to
the displacement of CHCl3 molecules from the vicinity of the
copper by attraction to the polystyrene matrix or to the alky-
lammonium cation sites.
hv
ꢁ
ꢁ
CHCl3 ! CHCl2 þ Cl
ð2Þ
ð3Þ
ð4Þ
ð5Þ
ð6Þ
ð7Þ
ꢁ
CHCl2 þ CHCl3 ! CH2Cl2 þ CCl3
ꢁ
Cl þ CHCl3 ! HCl þ CCl3
ꢁ
ꢁ
CCl3 þ O2 ! CCl3OO
ꢁ
ꢁ
2CCl3OO ! 2CCl3O þ O2
ꢁ
ꢁ
CCl3O ! COCl2 þ Cl
INTRODUCTION
The tetrachlorocuprate(II) ion acts as a homogeneous catalyst for
the photodecomposition of dichloromethane (1). In that system,
the photocatalytically active species is not CuCl24ꢀ, but rather
Cu2Cl26ꢀ, the two species being in rapid equilibrium in CH2Cl2
(1).
Trichloromethyl radicals formed by hydrogen abstraction react
with oxygen to yield trichloromethylperoxy radicals (9–11).
Their self-reaction results in trichloromethoxy radicals (12),
which in the gas phase are unstable with respect to dissociation
to phosgene and chlorine atoms (13). Eqs. (4)–(7) constitute a
radical chain with the net stoichiometry
2CuCl24ꢀ ꢀ Cu2Cl62ꢀ þ 2Clꢀ
ð1Þ
1
2
CHCl3 þ O2 ! COCl2 þ HCl
ð8Þ
During the course of photodecomposition, net reduction to
copper(I) is observed, which is later reversed. The reversal is
attributed to a buildup of oxidizing species such as the hydroper-
oxide CHCl2OOH (1).
In solution, the much higher rates of collision may make
hydrogen abstraction by trichloromethoxy radicals competitive
with unimolecular dissociation.
When heterogenized on a Dowex anion exchange resin,
CuCl24ꢀ loses its ability to catalyze the photodecomposition of
CH2Cl2 (2), possibly because the equilibrium with Cu2Cl26ꢀ is
unfavorable on the surface of the anion exchange resin. We find,
however, that heterogenized CuCl24ꢀ does catalyze the photode-
composition of chloroform, which is the subject of this study.
The thermal decomposition of chloroform yields almost exclu-
sively HCl and CCl2 (3), the least endergonic of the possible
decomposition products (3). Studies of chloroform photolysis,
however, have invariably found that the primary photochemical
step is carbon–chlorine bond fission (4–6). This is due to excita-
tion to an n(Cl)?r* (C–Cl) excited state (7), which leads to C–Cl
bond homolysis with essentially unit efficiency (4). The remaining
steps in the decomposition of CHCl3 by direct photolysis in the
CCl Oꢁ þ CHCl ! CCl OH þ
CCl
3
ð9Þ
ꢁ
3
3
3
From theoretical calculations by Sun and Bozzelli, DH° for this
reaction in the gas phase is approximately ꢀ65 kJ molꢀ1 (14).
Trichloromethanol is, however, unstable with respect to dissocia-
tion into hydrogen chloride and phosgene (15), thus hydrogen
abstraction yields a stoichiometry identical with that of Eq. (8).
The absorption spectrum of CuCl24ꢀ in the blue and near-UV
consists principally of ligand-to-metal charge transfer (LMCT)
bands (16,17). This suggests chlorine atom dissociation as a
potential mechanism for the initiation of the radical chain repre-
sented by Eqs. (4)–(7).
hv
CuCl24ꢀ ! CuCl23ꢀ
ð10Þ
þ Clꢁ
This would lead to the stoichiometry of Eq. (8) and the pre-
diction that the yields of HCl and COCl2 would be
*Corresponding author email: phoggard@scu.edu (Patrick E. Hoggard)
© 2014 The American Society of Photobiology
1234

Photochemistry and Photobiology, 2014, 90 1235
approximately equal. Given a reoxidation of Cu(I) to complete a
catalytic cycle, however, the ratio of HCl to COCl2 could
change. One possibility, not, however, observed during the
homogeneous catalysis of dichloromethane decomposition by
(Et4N)2CuCl4 in solution (1), is the reduction of chloroform by
the predominant copper(I) species, here taken to be CuCl23ꢀ for
convenience:
HCl in the photolysate was measured by adding aliquots (between 10
and 100 lL) to 3.00 mL portions of tetraphenylporphyrin (H2TPP, Fron-
tier Scientific, chlorin-free) dissolved in destabilized chloroform. The
absorbance increase at 446 nm, with the aid of extinction coefficients for
H2TPP and H4TPP2+ in chloroform (21,22), was used to determine the
concentration of [H4TPP2+
remaining porphyrin nitrogens:
] formed through the protonation of the
2HCl þ H2TPP ꢀ fH4TPP2þ; 2Clꢀg
ð15Þ
CuCl23ꢀ þ CHCl3 ! CuCl2ꢀ þ CHCl
ð11Þ
ꢁ
2
The intermediate H3TPP+ is unstable to disproportionation and can be
neglected (23), simplifying the analysis. Although the equilibrium con-
stant for Eq. (15) is very large, the reaction is not quite stoichiometric
(24). While readings can, in principle, be corrected for this by taking
measurements with two different ratios of photolysate to H2TPP (24),
deviations in duplicate measurements with the same amount of photoly-
sate were greater than the calculated corrections; therefore, corrections
were not applied and instead repeated measurements with different
amounts of photolysate were simply treated as independent measures of
the HCl concentration.
Phosgene concentrations in the photolysate were measured by adding
methanol (50 lL) to the TPP solution containing an aliquot of the pho-
tolysate, immediately following the HCl determination. Alcoholysis of
phosgene to form methyl chloroformate was generally complete within
2 min and the liberated HCl was measured by means of the additional
H4TPP2+ formed. Alcoholysis of the second chlorine, i.e. of methyl chlo-
roformate, is much slower and did not interfere with measurements made
within 30 min.
4
Following hydrogen abstraction by the resulting dichloromethyl
radical as in Eq. (3), the overall stoichiometry would be little chan-
ged, but some quantity of dichloromethane would be expected.
If, on the other hand, reoxidation were effected by a hydro-
peroxide, CCl3OOH, e.g. (18), some alteration in the ratio of
HCl to COCl2 would be expected.
2ꢀ
CuCl23ꢀ þ CCl3OOH ! CuCl3ðOHÞ þ CCl3O
ð12Þ
ꢁ
2ꢀ
CuCl3ðOHÞ þ HCl ! CuCl24ꢀ þ H2O
ð13Þ
ð14Þ
ꢁ
ꢁ
CCl3O ! COCl2 þ Cl
A parallel analysis for phosgene was conducted by adding an aliquot
from the photolysate to a 3.00 mL solution of approximately 0.01 M
Bu4NI in CHCl3 from which ethanol had been removed. After 2 min, the
absorbance at 365 nm was used to determine the concentration of I3ꢀ
(25), arising from the reaction (26)
Reoxidation by this mechanism would produce phosgene with
no increase in hydrogen chloride. This might or might not be
noticeable, depending on the length of the radical chain repre-
sented by Eqs. (4)–(7).
COCl2 þ 2Iꢀ ! CO(g) þ I2 þ 2 Clꢀ
ð16Þ
Given these considerations, our initial hypothesis was that
chloroform would be photodecomposed in a process initiated by
CuCl24ꢀ as in Eq. (10), producing HCl and COCl2 at approxi-
mately equal rates by the chain process of Eqs. (4)–(7), but with
potentially more COCl2 should Cu(I) be reoxidized primarily by
hydroperoxide. We also expected the side products to be domi-
nated by C2Cl6 from the self-termination of the CCl3 radical
chain carriers in Eqs. (4)–(7).
and the subsequent formation of I3ꢀ by reaction of the iodine with iodide
in solution. Triiodide reduces a number of potential species in solution
aside from phosgene, including peroxides; thus, a higher result than that
from the TPP test would indicate the presence of oxidizing species other
than phosgene.
UV–visible spectra were recorded with a Cary 50 spectrophotometer.
UV–visible reflectance spectra were recorded on an Ocean Optics USB4-
UV-VIS spectrophotometer with an Ocean Optics R400-7 reflectance
probe, using the Clꢀ form of the resin as a baseline. GC-MS measure-
ments were carried out with a Shimadzu QP-5000 instrument with an
Agilent DB624 column (30 m, 0.32 mm 9 1.8 lm cyanopropylphenyl/
dimethyl polysiloxane film). The oven start temperature was 40°C and a
20°C minꢀ1 linear temperature gradient was applied to a final temperature
of 240°C. A 10:1 split ratio was applied to sample injections. Species
were identified from their mass spectra and peak areas were used to deter-
mine concentrations. Before injection, 50 lL of ethanol, containing a
standard concentration of xylene as an internal standard, was mixed with
500 lL of the photolysate. Ethyl chloroformate, formed from the alcohol-
ysis of phosgene, was then used as the analytical standard in place of
phosgene. Without added ethanol, phosgene was readily identifiable in
the gas chromatogram, as was its complete disappearance and conversion
to ethyl chloroformate in the presence of ethanol. Analyses had to be
undertaken within a few hours to avoid further alcoholysis to diethyl
carbonate.
MATERIALS AND METHODS
Chloroform (J. T. Baker, ACS Grade) was washed seven times with an
equal volume of water to remove the ethanol stabilizer, after which 4A
molecular sieves were added to remove dissolved water.
Because the chloride form of Dowex 2-X8 catalyzes the photodecom-
position of chloroform (19), it was desirable to remove Clꢀ entirely when
heterogenizing CuCl24ꢀ. The resin was first converted to the perchlorate
form, which is photocatalytically inactive (19), by passing a 1M aqueous
solution of NaClO4 through a column containing the chloride form of the
resin (Bio-Rad AG2-X8,) until the eluate was chloride-free by AgNO3
test, then rinsing with water and air-drying. The tetrachlorocuprate cata-
lyst was made by dissolving 0.54 g of (Et4N)2CuCl4 (Sigma-Aldrich) in
50 mL of acetonitrile, adding 1.00 g of the ClOꢀ4 form of Dowex 2-X8,
and stirring for 24 h, at which time the liquid phase was nearly colorless.
After filtering the resin and rinsing with CH3CN, the resin was soaked in
chloroform for 1 hr to remove traces of acetonitrile, then filtered and air-
dried. Based on the stated resin capacity of 3.2 mequiv per dry gram
(Clꢀ form) (20), which corresponds to 2.7 mequiv gꢀ1 in the perchlorate
form, approximately 87% of the cationic resin sites were occupied by
CuCl24ꢀ ions. The CuCl24ꢀ loading was 1.1 mmol per gram of dry resin.
Photodegradation experiments were generally performed with 1.0 mL
portions of chloroform with a stirred suspension of the CuCl24ꢀ/Dowex
catalyst in a 1 cm fused silica cuvette. A balloon containing air or O2
was affixed to the cuvette with a plastic pipet. Light from an Oriel 350-
W mercury lamp was passed through a 345 nm longpass filter and direc-
ted to the sample slightly defocused, so as to irradiate most of the sam-
ple. A fan was used to maintain the temperature in the cuvette at
22 ꢂ 2°C.
RESULTS AND DISCUSSION
Reaction products
When chloroform was exposed to broadband irradiation
(k > 345 nm) in the presence of the CuCl24ꢀ/Dowex catalyst,
HCl and COCl2 accumulated in solution, although it can be
assumed that both escaped into the gas phase to some extent.
Figure 1 shows a typical 30-min experiment. A short induction
period is evident, following which COCl2 increased steadily,
while HCl formation, from the beginning slower than that of

1236 Brent M. Harvey and Patrick E. Hoggard
Table 1. Product yields following a 180-min irradiation (k > 345 nm,
350-W Hg lamp) of a suspension of 30 mg of CuCl24ꢀ/Dowex in 1 mL
of CHCl3 exposed to air.
Product
Yield, lmol
HCl
10
260
345
360
150
235
3
COCl2 (GCMS)
COCl2 (TPP test)
COCl2 (Iˉ test)
COCl2 in balloon*
CCl4
C2Cl6
CH2Cl2
1
*TPP test.
are produced during irradiation and build up in solution along
with phosgene.
The longer irradiation confirms the decline in the rate of HCl
production intimated in Fig. 1, while the rate of COCl2 forma-
tion increased further, compared to the rate during the first
30 min. Unexpectedly, given the presumed mechanism from
Eqs. (3)–(7), a large amount of CCl4 was also found, far exceed-
ing the amount of HCl. On the other hand, there were only trace
quantities of C2Cl6, the expected termination product from the
CCl3 radicals that were presumed to be chain carriers. These
results are in distinct conflict with our initial hypothesis of chlo-
Figure 1. Products formed during the broadband (k
> 345 nm, 350-W
Hg lamp) irradiation of 2 mL of chloroform with 29 mg of CuCl42ꢀ
/
Dowex exposed to air. Solid lines are fits to Eqs. (28) and (29) in text.
rine atom photodissociation from CuCl42ꢀ
.
COCl2, appeared to level off. The considerably greater rate of
COCl2 formation might be consistent with peroxide or hydroper-
oxide reoxidation of Cu(I), and certainly rules out reoxidation by
chloroform as a major pathway. However, the COCl2 to HCl
ratio was much higher than would be predicted for decomposi-
Control experiments
Table 2 shows a comparison of chloroform decomposition yields
in the presence of the CuCl24ꢀ form of the Dowex 2-X8 resin,
the perchlorate form from which it was prepared, and the chlo-
ride form, which has been shown to be photocatalytically active
(19), all under irradiation above 345 nm. While the perchlorate
form showed no activity whatever, the resin in the chloride form
was even more active than the resin with CuCl24ꢀ. This differ-
ence may be attributable to the considerably higher concentration
of anion sites per gram in the chloride form, but the complex
relationship between the quantity of a photocatalyst and its activ-
ity (27) permits only the broadest comparisons. No decomposi-
tion of chloroform occurred in the presence of CuCl24ꢀ/Dowex in
the dark. There was also no decomposition of pure chloroform
under irradiation for 60 min in the absence of a catalyst.
tion initiated by chlorine atom photodissociation from CuCl42ꢀ
.
A 3-hr irradiation was undertaken in order to be able to detect
minor products by GC-MS, and the results are shown in Table 1,
which also includes a comparison of the three methods used to
determine the phosgene yield.
While the TPP and triiodide tests gave similar estimates of
the phosgene concentration in solution, the concentration by GC-
MS was approximately 25% smaller. It is possible that the dis-
crepancy is due to the evaporation of phosgene during sample
handling and further evaporation into the head space of the sam-
ple vial while waiting to be processed. The high volatility of
COCl2 in chloroform was tested by expressing into methanol the
contents of an air balloon that had been connected to the sample
during irradiation, converting much of the phosgene to methyl
chloroformate and releasing HCl in the process. This served only
to establish a lower limit for the COCl2 in the vapor phase, since
it can be assumed that some of the phosgene escaped without
reacting with methanol. As shown in Table 1, the vapor phase
contained on the order of half as much phosgene as was present
in the chloroform solution.
Reflectance spectra
Because Cu2Cl26ꢀ, in equilibrium with CuCl24ꢀ as in Eq. (1), was
found to be the photocatalytically active species in homoge-
neously catalyzed CH2Cl2 photodecomposition, reflectance spec-
The discrepancies between the phosgene concentration deter-
mined by GCMS and by the TPP test were much smaller for
shorter irradiation times, which is consistent with a greater loss
of volatile phosgene into the head space from solutions of high
COCl2 concentration (ca. 0.5M after the 3-hr radiation reported
in Table 1) during irradiation and sample handling for the analy-
ses (1–2 hr). On the other hand, the difference between the TPP
and iodide tests was persistent, often greater than that shown in
Table 1. Since the iodide test responds to many oxidizing
species, it can be concluded that oxidants other than phosgene
Table 2. Product yields (in lmol) following 60-min irradiation
(k > 345 nm, 350-W Hg lamp, O2 balloon) of a suspension in 1 mL
CHCl3 of 31 mg of Dowex 2-X8 resin in the CuCl24ꢀ, ClO4ꢀ, or Clꢀ
form.
HCl
COCl2*
CuCl24ꢀ
ClO4−
Clꢀ
6.3
0.0
35
322
0
489
*Methanol/TPP test.

Photochemistry and Photobiology, 2014, 90 1237
Table 3. Product yields (in lmol) following 60-min irradiation
(k > 345 nm, 350-W Hg lamp) of a suspension of 31 mg CuCl24ꢀ/Dowex
in 1 mL CHCl3 with an air or an oxygen balloon.
21% O2
100% O2
HCl
COCl2*
1.0
68
6.3
348
*Iodide test.
Table 4. Product yields (in lmol) following 60-min irradiation (k
>
345 nm, 350-W Hg lamp, O2 balloon) of a suspension of 31 mg CuCl42ꢀ
Dowex in 1 mL CHCl3 with and without the addition of acetonitrile.
/
no CH3CN
3% CH3CN
HCl
6.3
322
348
0.3
32
115
COCl2†*
COCl2
*Methanol/TPP test. †Iodide test.
1.00 mL of chloroform, no detectable amount of either HCl or
Figure 2. Reflectance, Kubelka–Munk function (60,61), of CuCl42ꢀ
/
phosgene was obtained (< 0.1 lmol). This result was unex-
Dowex, relative to Clꢀ/Dowex.
pected, because Cl atoms from the dissociation of CuCl42ꢀ
,
Eq. (10), would be expected to abstract hydrogen from both
chloroform and cyclohexane, the C–H bond enthalpies being
similar in these two molecules (28,29). No bicyclohexyl or other
cyclohexyl products were seen by GC-MS; thus, it appears that
the quenching is not a result of radical scavenging by C6H12.
Photodecomposition was also completely quenched when irra-
diation was carried out on stabilized chloroform, containing
approximately 1% ethanol. Again, despite the well-known role
of ethanol as a radical scavenger, quenching was not attributable
to this mechanism, since the products from hydrogen abstraction
from the a-carbon, primarily acetaldehyde (30), but potentially
2,3-butanediol (31), were not found.
tra were taken both before and after irradiation to determine
whether any detectable amount of Cu2Cl26ꢀ might be present. In
CH2Cl2 solution, CuCl24ꢀ was characterized by a peak at
409 nm, while Cu2Cl26ꢀ had an almost equally intense peak at
475 nm (1). By reflectance (Fig. 2) we could see that the resin
absorbed strongly between 380 and 440 nm, but could not find
the exact peak position. Nevertheless, there was no evidence of
an absorbance peak near 475 nm, and it appears that all the cop-
per was in the form of CuCl24ꢀ. This did not change after irradia-
tion; thus, if Cu2Cl26ꢀ does participate in the process, its
existence is transitory.
Oxygen dependence
Photodecomposition in the presence of acetonitrile
Although the mechanism described by Eqs. (4)–(7) is clearly at
odds with the experimental product distribution, the formation of
the trichloromethylperoxy radical as in Eq. (5) could still be a
key step. If so, a dependence of the overall photodecomposition
rate on the rate of Eq. (5), and thus on the photostationary state
concentration of CCl3, would lead to the prediction that as the
partial pressure of O2 increased, the photodecomposition rate
would be higher, asymptotically approaching the condition in
which essentially all CCl3 radicals reacted with oxygen and addi-
tional O2 yielded no increase in rate.
Experimentally, as seen in Table 3, photodecomposition pro-
ceeded much more rapidly under an atmosphere of pure oxygen
than under air. The yield of phosgene was approximately propor-
tional to the O2 percentage, as was the HCl yield, although very
small. Because of this, most of the tests conducted were under an
oxygen atmosphere. No asymptotic behavior was observed.
In the presence of 3% acetonitrile, the photodecomposition of
chloroform catalyzed by CuCl24ꢀ/Dowex slowed, but was not
quenched entirely, as can be seen in Table 4. In addition, there
was a discrepancy in the phosgene determined by the methanol/
TPP test and by the iodide test. Since in the iodide test, triiodide
is also formed by reaction with peroxides, it is possible that
CCl3OOH or related species accumulated in solution during
irradiation.
Photodecomposition in the presence of added chloride ion
The addition of Bu4NCl to chloroform drastically lowered the
rate of photodecomposition, as can be seen in Table 5. This sug-
gests that there may be an equilibrium step in the photodecom-
position in which Clꢀ is a product, although there are other
explanations for the retardation of the photodecomposition rate
that may be considered (vide infra).
Photodecomposition in the presence of cyclohexane and
ethanol
Catalyst surface area
When photodecomposition was carried out as in Table 3, irradi-
ating for 60 min, but with 30 lL of cyclohexane added to the
It was expected that increasing the surface area of the
CuCl24ꢀ/Dowex catalyst would increase the rate of photodecom-

1238 Brent M. Harvey and Patrick E. Hoggard
Table 5. Product yields (in lmol) following 60-min irradiation (k
>
Table 6. Product yields (in lmol) following 60-min irradiation
(k > 345 nm, 350-W Hg lamp, O2 balloon) of a suspension in 1 mL
CHCl3 of 3.5 mg of CuCl24ꢀ/Dowex, normal (16-50 mesh) or powdered.
345 nm, 350-W Hg lamp, O2 balloon) of a suspension of 31 mg CuCl42ꢀ
Dowex in 1 mL CHCl3 with and without the addition of chloride ion.
/
no Bu4NCl
5 mg Bu4NCl
HCl
COCl2*
HCl
COCl2*
6.3
322
1.0
6
Normal
Powdered
0.4
107
8
1100
*Methanol/TPP test.
*Iodide test.
Table 7. Product yields (in lmol) following 60-min irradiation
(k > 345 nm, 350-W Hg lamp, O2 balloon) of 2 9 10ꢀ3 mol of CuCl42ꢀ
in 1 mL of CHCl3 in a homogeneous solution and in a heterogeneous
suspension.
position and, as Table 6 shows, this was decidedly true in the
comparison of 3.5 g of the normal resin (16–50 mesh) with the
same amount of powdered resin. The dramatic increase in yield
can be ascribed to the much lower transmittance of the suspen-
sion, and thus higher probability of absorption, with the smaller
particles (27).
Catalyst
HCl
COCl2
(Et4N)2CuCl4 in solution
CuCl24ꢀ on Dowex 2-X8
14
35
300
490
Heterogeneous vs homogeneous catalysis
Heterogenization of
a homogeneous photocatalyst generally
more direct evidence that the photodissociation of chlorine atoms
from CuCl24ꢀ/Dowex does not occur under the conditions of this
study. What HCl was found can potentially be ascribed to phos-
gene hydrolysis, water being formed when peroxides or hydro-
peroxides react by breaking the O–O bond (43), vide infra,
Eqs. (26) and (27).
results in a lower catalytic activity for an equivalent amount of
catalyst, due to less efficient light absorption in the solid state
(32). There are, however, examples where only a small loss of
activity is observed (33,34), or even an increase in activity (35),
due predominantly to the stabilization of intermediate species by
the solid matrix (36).
A comparison was made between the homogeneous photocat-
alytic activity of (Et4N)2CuCl4 and the heterogeneous activity of
CuCl24ꢀ on Dowex 2-X8, the results of which are shown in
Table 7. Unexpectedly, the heterogenized form exhibited some-
what higher yields of both HCl and COCl2.
Is dichlorocarbene produced in the primary photochemical
step?
Given the lack of evidence for the photodissociation of chlorine
atoms from either CuCl24ꢀ or from CHCl3 directly, one possibil-
ity is that under visible irradiation in the presence of CuCl42ꢀ
chloroform decomposes to HCl and CCl2, as it does in uncata-
lyzed thermal degradation (3). This can be ruled out from the
low yield of HCl and the absence of the typical CCl2 secondary
products in chloroform, C2Cl4 and C2HCl5 (3).
Mechanism 1—Photodissociation of a chlorine atom
Our initial hypothesis was that the photocatalytic activity of
CuCl24ꢀ on Dowex was due to the photodissociation of a chlorine
atom through irradiation into the Cu(II) LMCT bands in the visi-
ble and near-UV, as in Eq. (10), which would be consonant with
the common understanding of the photochemistry of chlorocop-
per(II) complexes in organic solvents (37–40), specifically
including CuCl24ꢀ (41).
For several reasons, this mechanism can be ruled out. One
consideration is that as a homogeneous photocatalyst for dichlo-
romethane decomposition, the tetrachlorocuprate ion is not itself
active, but is in equilibrium with the hexachlorodicuprate ion, as
in Eq. (1), which is photoactive (1). The photocatalytic activity
of the Cu2Cl26ꢀ ion may be due, in part, to stabilization of photo-
dissociated chlorine atoms by the chloride ions in solution as a
result of the equilibrium of Eq. (1), whereby the resulting Cl2ꢀ
ion prevents recombination sufficiently to allow diffusion out of
the solvent cage (1). From the reflectance spectrum of CuCl42ꢀ
on Dowex 2-X8, there was no detectable Cu2Cl26ꢀ peak; thus,
the formation of the dimer as in Eq. (1) is greatly suppressed on
the ion exchange resin, if it takes place at all.
Mechanism 2—Photocatalysis by Clꢀ/Amberlite in
equilibrium with CuCl42ꢀ
As shown in Table 2, the chloride form of the anion exchange
resin is an efficient catalyst for the decomposition of chloroform,
possibly more efficient than the CuCl24ꢀ form of the resin. Because
of the potential lability of chloride ions in copper(II) complexes,
this raises the question of whether the photocatalytic activity of
CuCl24ꢀ on Dowex 2-X8 is actually due to dissociated chloride
ions. Hypothetically, a ligand dissociation equilibrium populates a
certain fraction of resin sites with chloride ions, that is
ðRþÞCuCl42ꢀðRþÞ ꢀ ðRþÞCuClꢀ3 þ ðRþÞClꢀ
ð17Þ
where R+ represents the cationic resin sites, which have the
structure R-CH2N+(CH3)2(C2H4OH) in Dowex 2-X8 (20). If
this hypothesis were correct, the observed photocatalytic activ-
ity would actually result from light absorption by the resin in
the vicinity of chloride ions rather than by CuCl24ꢀ complexes.
This mechanism is consistent with the high ratio of CCl4 to
C2Cl6 observed. The mechanism for the photocatalysis of chloro-
form decomposition by the chloride form of the resin appears to
comprise two sequential photochemical processes, the first of
which leads to CCl4 (19):
Experimentally, this mechanism is directly contradicted by the
finding that only a very small quantity of C2Cl6 is produced.
When chlorine atoms abstract hydrogen from chloroform to
make CCl3 radicals, as in Eq. (4), the trichloromethyl radicals
self-terminate to make C2Cl6 the major side product when chlo-
rine atoms initiate the photodecomposition of chloroform (42).
The small amount of HCl produced in our experiments is even

Photochemistry and Photobiology, 2014, 90 1239
hv
Chloride ions in solution would suppress this equilibrium by
displacing coordinated chloroform, thereby hindering further
reaction. When copper with chloroform in the first coordination
sphere absorbs light, the excited state would then be able to react
with oxygen, leading to the peroxide.
ðRþÞClꢀ þ CHCl3 þ O2 !ðRþÞHO2ꢀ þ CCl4
ð18Þ
The second process, representing the major decomposition
pathway for chloroform, would then be the chain reaction Eqs.
(4)–(7), initiated by the photodissociation of CCl4.
½ðRþÞCuCl3ðCl3CHÞꢀðRþÞClꢀꢄꢃ þ O2 !
The reflectance spectrum, on the other hand, does not sup-
port this hypothesis, because no trace of the trigonal CuCl3ꢀ
ion spectrum (44) can be found. This is not conclusive in
itself, however, since the Cu2+ ion may remain four-coordinate
even after dissociation of a chloride ion, through coordination
of chloroform molecules (vide infra). A weightier argument
against the mechanism is that even if the dissociation equilib-
rium, Eq. (17), lay completely to the right, still only half of
the resin cation sites would have chloride ions. Furthermore,
for wavelengths above 345 nm the tetrachlorocuprate(II) ion
absorbs much more strongly than the polystyrene resin, greatly
reducing net light absorption by the resin. It seems unlikely
that there would be significant energy transfer from the copper
(II) to the resin, given the low-lying CuCl24ꢀ ligand field
excited states (45). It is difficult to see how under these cir-
cumstances a photodecomposition efficiency could result that
was two-thirds as great as that achieved when all the resin
sites are occupied by Clꢀ.
ð21Þ
ðRþÞHO2ꢀ þ ðRþÞCuClꢀ3 þ CCl4
Chloroform is not normally considered as a potential ligand
(46), but in fact there is NMR evidence of chloroform coordina-
tion with first-row transition metal ions (47).
The tetrachlorocuprate(II) ion would be regenerated following
the protonation of the hydrogen peroxide anion by hydrogen
chloride.
HCl þ ðRþÞHO2ꢀ ! ðRþÞClꢀ þ H2O2
ð22Þ
ðRþÞClꢀ þ ðRþÞCuCl3ꢀ ! ðRþÞCuCl42ꢀðRþÞ
ð23Þ
As with the photodecomposition of chloroform in the presence
of Clꢀ on Amberlite IRA-900 (19), phosgene would be formed
through the subsequent catalyzed photodissociation of CCl4, with
CCl3 radicals as chain carriers. This could conceivably occur in
exactly the same way through absorption of light by the polysty-
rene resin at sites with Clꢀ ions, achieved through the equilibrium
of Eq. (20). However, the presumably small value for the equilib-
rium constant of Eq. (20), together with the much higher fraction
of light absorbed by CuCl24ꢀ, leads to the supposition of an analo-
gous process involving CuCl24ꢀ and a CuCl52ꢀ intermediate.
Experimentally, the Clꢀ and CuCl24ꢀ forms of the resin also
exhibit different relationships between photocatalytic activity and
the partial pressure of oxygen. The resin excited state is quenched
by O2, which with the Clꢀ/Amberlite catalyst causes an asymp-
totic increase in the photochemical yield with the partial pressure
of oxygen (19). In contrast, the CuCl24ꢀ/Dowex system exhibits a
direct proportionality between O2 partial pressure and photochem-
ical yield up to 1 atm, thus no quenching of the excited state
within this range of partial pressures.
A further incompatibility with this hypothesis is furnished
by the behavior with added chloride ion. Additional chloride
ion would not be expected to alter the equilibrium of
Eq. (17), unless one or the other of the two copper complexes
could be displaced by Clꢀ, for which there was no spectro-
photometric evidence. Chloride could, however, displace some
of the small fraction of perchlorate ions (~13% of the resin
sites) remaining after preparation of the CuCl24ꢀ form of the
resin. Thus, if there were any effect, it would be expected to
be that of enhancement of the reaction rate through the addi-
tional Clꢀ sites on the resin, whereas a decrease in yield was
observed experimentally.
ðRþÞCuCl24ꢀꢃðRþÞ þ CCl4 ! ðRþÞCuCl2ꢀðRþÞ þ CCl ð24Þ
ꢁ
3
5
ðRþÞCuCl25ꢀðRþÞ þ CHCl3 ! ðRþÞCuCl2ꢀðRþÞ þ HCl þ CCl
ꢁ
3
4
ð25Þ
Trichloromethyl radicals can react with oxygen to generate
phosgene as in Eqs. (4)–(7), and in addition, hydrogen peroxide
can react with CCl3 radicals, breaking the O–O bond (48) and
propagating the chain:
CCl !ꢁOH þ CCl OHð! HCl þ COCl Þ
H O þ
ð26Þ
ꢁ
2
2
3
3
2
ꢁOH þ HCl ! H O þ
Cl
ð27Þ
ꢁ
2
Mechanism 3—Chlorine atom transfer with photoreduction
of O2
As water builds up, it can be expected to hydrolyze phosgene
to yield carbon dioxide and water. If the rate of Eq. (26) is high
enough, it prevents the self-termination of CCl3 radicals, which
would explain the low quantities of C2Cl6 found.
Consider a process similar to that of Eq. (18) in which immobi-
lized CuCl24ꢀ is the photoactive species:
DHo for the reaction represented by Eq. (26) is estimated to be
ꢀ160 kJ molꢀ1, using the gas phase data for H2O2 in preference
to the highly hydrogen-bonded liquid phase, and using the value
of -261 kJ molꢀ1 inferred from ab initio calculations for CCl3OH
(15). Trichloromethanol decomposes rapidly to hydrogen chloride
and phosgene (15,49), with DHo of approximately -50 kJ molꢀ1
(15). As water builds up, it can be expected to hydrolyze phos-
gene to yield carbon dioxide and hydrogen chloride.
O2
ðRþÞCuCl24ꢀꢃðRþÞ þ CHCl3 ! CCl4 þ ðRþÞHO2ꢀ þ ðRþÞCuClꢀ3
ð19Þ
It is doubtful that this could take place as a single step, and it
is not clear what the mechanistic steps would be, but we suggest
that chloroform displaces some fraction of the coordinated chlo-
ride ions in an equilibrium process, possibly assisted by the
geometry of the resin cation sites.
In the process described above, the photocatalytic activity of
CuCl24ꢀ appears to be exerted with no involvement of Cu(I), at
odds with generally accepted models for copper complex redox
ðRþÞCuCl24ꢀðRþÞ þ CHCl3¡ðRþÞCuCl3ðCl3CHÞꢀðRþÞClꢀ ð20Þ

ƒƒƒƒƒƒƒ
ƒ!
1240 Brent M. Harvey and Patrick E. Hoggard
photochemistry (50). However, the reaction shown in Eq. (21)
could be a multistep process involving a copper(I) intermediate.
In any case, it seems likely that the reaction would proceed
through a charge transfer to solvent (CTTS) excited state
(51–53). CTTS states have specifically been implicated in the
photooxidation of various substrates by Cu(II) complexes in
chloroform and other halocarbon solvents (54).
The time sequence data in Fig. 1 can be fit reasonably well to
a simple model for two sequential photochemical processes, the
first of which produces CCl4, both induced by excited state tetra-
chlorocuprate(II), here represented as Cu*.
cyclohexane, its quenching function must also result from the dis-
placement of chloroform from the photoactive sites, presumably
through the strong London forces shared with the polystyrene
matrix.
The discrepancy between the TPP test and the iodide test for
phosgene was ascribed to the presence of oxidizing species other
than phosgene in irradiated solutions. While hydroperoxides,
formed by the reaction of CCl3 radicals with O2 followed by
hydrogen abstraction, may account for some of these oxidizing
equivalents, hydrogen peroxide, if formed as in Eqs. (21) and
(22), would presumably be the predominant species of this type.
The direct proportion observed between the photodecomposi-
tion rate and the partial pressure of oxygen is also explained
better by Mechanism 3. In the other mechanisms considered, O2
reacts with CCl3 to produce CCl3OO radicals, leading to an
expected asymptotic variation of the decomposition rate with the
O2 partial pressure, a relationship that has been observed in other
systems involving trichloromethylperoxy radicals (56). In Mecha-
nism 3, O2 reacts with a chlorocopper(II) excited state complex
to which chloroform is coordinated. Given the expected high rate
of deexcitation of the Cu(II) excited state, saturation of the pho-
tochemical reaction channel with O2 would not be expected.
Finally, consider how it can be that equivalent molar quanti-
ties of CuCl24ꢀ catalyze the photodecomposition of chloroform
better when immobilized on an anion exchange resin than in
homogeneous solution. This can be ascribed to a matrix effect in
which the chloroform molecules are brought into the vicinity of
CuCl24ꢀ for a longer duration, promoting coordination as in
Eq. (20), which we suggest is a prerequisite for the excited
Cu(II) species to react with O2. CHCl3 molecules are attracted
by London forces to the polystyrene, but also by ion-dipole
forces to the tetraalkylammonium cation sites, which is consis-
tent with an enforced proximity to the CuCl24ꢀ anions and a shift
of the equilibrium in Eq. (20) toward the chloroform complex.
In water, CuCl24ꢀ is rapidly aquated, which rules this complex
out as a means to catalyze the photodegradation of halocarbons
in water by means of sunlight, the goal of a number of ongoing
research efforts (57–59). However, the ability of CuCl24ꢀ to cata-
lyze chloroform photodegradation with near-UV irradiation plus
the relatively low cost and low toxicity of copper, compared to
the heavy metals most commonly under investigation (57), sug-
gest that related copper complexes that are not susceptible to
ligand displacement may prove useful.
Cuꢃ þ CHCl3 ! CCl4
ð28Þ
Cuꢃ þ CCl4 ! n COCl2
ð29Þ
Differential equations for these two steps were modeled numeri-
cally using small time intervals and the rate constants were
adjusted to yield the fit in Fig. 1. This should not be considered
evidence offered in proof of this interpretation, because there are
other models that are also consonant with the data of Fig. 1, e.g.
one in which CCl4 undergoes photodissociation directly.
The net reaction predicted by Eqs. (20) and (21) is
hv;ðRþÞ2CuCl4
CHCl þ O þ HCl
CCl þ H O
ð30Þ
3
2
4
2
2
The overall reaction is exothermic by approximately
40 kJ molꢀ1 (again using gas phase data for H2O2), although
one or more of the steps implied by Eq. (21) must be endother-
mic, requiring light. Eqs. (26) and (27), together with Eq. (4),
constitute a radical chain with net stoichiometry
CHCl3 þ H2O2 ! HCl þ H2O þ COCl2
ð31Þ
The net reaction is exothermic by approximately
300 kJ molꢀ1 (using H2O and H2O2 gas phase data), and each
of the three steps is also exothermic.
CONCLUSIONS
Virtually all of the experimental observations find an explanation
with the steps outlined above for Mechanism 3. The low ratio of
hydrogen chloride to phosgene can be attributed to the protonation
of hydrogen peroxide anions, Eq. (22), and hydroxyl radicals,
Eq. (27), by HCl. The suppression of photodecomposition by eth-
anol and acetonitrile, without either acting as a radical scavenger,
can be ascribed to the same process by which chloride ions retard
the reaction, that is, by suppressing the complexation of chloro-
form, Eq. (20), preferentially occupying copper(II) coordination
sites in place of the weaker ligand chloroform. Both acetonitrile
and ethanol may also displace chloroform from the photoactive
sites indirectly because of strong intermolecular attractions to the
polystyrene chain or to the cation resin sites, or both. The attrac-
tion of the resin to ethanol may be particularly enhanced by
hydrogen bonding with the 2-hydroxyethyltrialkylammonium
moieties that constitute the resin cation sites (55), which might
explain why ethanol suppressed photodecomposition more effec-
tively than acetonitrile.
Acknowledgements—The authors thank the donors of the American
Chemical Society Petroleum Research Fund for support of this research
through grant 49334-UR3.
REFERENCES
1. Doyle, K. J., H. Tran, M. Baldoni-Olivencia, M. Karabulut and P. E.
Hoggard (2008) Photocatalytic degradation of dichloromethane by
chlorocuprate(II) ions. Inorg. Chem. 47, 7029–7034.
2. Chan, A. M., B. M. Harvey and P. E. Hoggard (2013) Photodecom-
position of dichloromethane catalyzed by tetrachloroferrate(III) sup-
ported on a Dowex anion exchange resin. Photochem. Photobiol.
Sci. 12, 1680–1687.
3. Won, Y. S. and J. W. Bozzelli (1992) Chloroform pyrolysis:
Experiment and detailed reaction model. Combust. Sci. Technol. 85,
345–373.
The importance of matrix effects is underscored by the quench-
ing of the photodecomposition in the presence of cyclohexane.
Since there is no evidence of any hydrogen abstraction from
4. Taketani, F., K. Takahashi and Y. Matsumi (2005) Quantum yields
for Cl(2Pj) atom formation from the photolysis of chlorofluorocar-

Photochemistry and Photobiology, 2014, 90 1241
bons and chlorinated hydrocarbons at 193.3 nm. J. Phys. Chem. A
109, 2855–2860.
5. Yang, X., P. Felder and J. R. Huber (1994) Photodissociation of the
CHFCl2 and CHCl3 molecules and the CHCl2 radical in a beam at
193 nm. Chem. Phys. 189, 127–136.
6. Unger, I. and G. P. Semeluk (1966) The photolytic and photosensi-
tized decomposition of chloroform. Can. J. Chem. 44, 1427–1436.
7. Reid, M., V. Green and S. P. K. Koehler (2014) Near-threshold pho-
todissociation dynamics of CHCl3. Phys. Chem. Chem. Phys. 16,
6068–6074.
and thermochemistry of trichloromethyl radical and cation. J. Phys.
Chem. 95, 4400–4405.
29. McMillen, D. F. and D. M. Golden (1982) Hydrocarbon bond disso-
ciation energies. Annu. Rev. Phys. Chem. 33, 493–532.
30. Hoggard, P. E. and A. Maldotti (2010) Catalysis of the photodecom-
position of carbon tetrachloride in ethanol by an amberlite anion
exchange resin. J. Catal. 275, 243–249.
31. Zhang, X., J. Wu and Y. Zhou (1994) The reaction of a-hydroxyeth-
yl radical in aqueous solution and ethyl alcohol. Radiat. Phys. Chem.
43, 335–338.
8. Nachbor, M. D., C. F. Giese and W. R. Gentry (1995) UV Photodis-
sociation dynamics of single-carbon freons and related molecules.
J. Phys. Chem. 99, 15400–15409.
9. Cooper, R., J. B. Cumming, S. Gordon and W. A. Mulac (1980)
The reactions of the halomethyl radicals CCl3 and CF3 with oxygen.
Radiat. Phys. Chem. 16, 169–174.
10. Hautecloque, S. (1980) On the photooxidation of gaseous trichlorom-
ethane and chlorosyl radical formation. J. Photochem. 14, 157–165.
11. Mosseri, S., Z. B. Alfassi and P. Neta (1987) Absolute rate constants
for hydrogen abstraction from hydrocarbons by the trichloromethyl-
peroxy radical. Int. J. Chem. Kinet. 19, 309–317.
32. Shul’pin, G. B., G. V. Nizova and Y. N. Kozlov (1996) Photochemi-
cal aerobic oxidation of alkanes promoted by iron complexes. New
J. Chem. 20, 1243–1256.
33. Maldotti, A., A. Molinari, G. Varani, M. Lenarda, L. Storaro, F.
Bigi, R. Maggi, A. Mazzacani and G. Sartori (2002) Immobilization
of (n-Bu4N)4W10O32 on mesoporous MCM-41 and amorphous silicas
for photocatalytic oxidation of cycloalkanes with molecular oxygen.
J. Catal. 209, 210–216.
34. Maldotti, A., G. Varani and A. Molinari (2006) Photo-assisted chlo-
rination of cycloalkanes with iron chloride heterogenized with
Amberlite. Photochem. Photobiol. Sci. 5, 993–995.
12. Catoire, V., R. Lesclaux, W. F. Schneider and T. J. Wallington
(1996) Kinetics and mechanisms of the self-reactions of CCl3O2 and
CHCl2O2 radicals and their reactions with HO2. J. Phys. Chem. 100,
14356–14371.
13. Lesclaux, R., A. M. Dognon and F. Caralp (1987) Photooxidation of
halomethanes at low temperature: the decomposition rate of CCl3O
and CFCl2O radicals. J. Photochem. Photobiol. A 41, 1–11.
14. Sun, H. and J. W. Bozzelli (2001) Structures, intramolecular rotation
barriers, and thermochemical properties of radicals derived from H
atom loss in mono-, di-, and trichloromethanol and parent chloro-
methanols. J. Phys. Chem. A 105, 4504–4516.
15. Wallington, T. J., W. F. Schneider, I. Barnes, K. H. Becker, J. Sehest-
ed and O. J. Nielson (2000) Stability and infrared spectra of mono-,
di-, and trichloromethanol. Chem. Phys. Lett. 322, 97–102.
16. Desjardins, S. R., K. W. Penfield, S. L. Cohen, R. L. Musselman
and E. I. Solomon (1983) Detailed absorption, reflectance, and UV
photoelectron spectroscopic and theoretical studies of the charge-
transfer transitions of tetrachlorocuprate(2-) ion: correlation of the
square-planar and the tetrahedral limits. J. Am. Chem. Soc. 105,
4590–4603.
35. Molinari, A., A. Bratovcic, G. Magnacca and A. Maldotti (2010)
Matrix effects on the photocatalytic oxidation of alcohols by
[nBu4N]4W10O32 incorporated into sol-gel silica. Dalton Trans. 39,
7826–7833.
36. Maldotti, A., A. Molinari and R. Amadelli (2002) Photocatalysis
with organized systems for the oxofunctionalization of hydrocarbons
by O2. Chem. Rev. 102, 3811–3836.
37. Bergamini, P., A. Maldotti, S. Sostero, O. Traverso and J. Sykora
(1984) Photochemical redox reactivity in chlorocopper(II) complexes.
Inorg. Chim. Acta 85, L15–L17.
38. Horvath, E., J. Sykora and J. Gazo (1978) Thermal and photochemi-
cal properties of chlorocopper(II) complexes in acetonitrile. I. Zeit-
schrift fuer Anorganische und Allgemeine Chemie 442, 235–244.
39. Horvath, O. (1989) Geminate-pair scavenging mechanism in
photochemical reactions of chlorocuprate(II) complexes in acetoni-
trile solutions containing alcohols. J. Photochem. Photobiol. A 48,
243–248.
40. Kochi, J. K. (1962) Photolysis of metal compounds: cupric chloride
in organic media. J. Am. Chem. Soc. 84, 2121–2127.
41. Cervone, E., F. D. Camassei, I. Giannini and J. Sykora (1979) Pho-
toredox behavior of chlorocopper(II) complexes in acetonitrile:
mechanism and quantum yields. J. Photochem. 11, 321–332.
42. Hoggard, P. E., M. Gruber and A. Vogler (2003) The photolysis of
iron(III) chloride in chloroform. Inorg. Chim. Acta 346, 137–142.
43. Kochi, J. K. (1973) Oxidation-reduction reactions of free radicals
and metal complexes. Free Rad. 1, 591–683.
17. Sharnoff, M. and C. W. Reimann (1967) Charge-transfer spectrum
of the tetrachlorocuprate ion. J. Chem. Phys. 46, 2634–2640.
€
18. Gab, S. and W. V. Turner (1985) Photooxidation of chloroform: Iso-
lation and characterization of trichloromethyl hydroperoxide. Angew.
Chem. 97, 48.
~
19. Pena, L. A., A. M. Chan, A. S. Cohen, K. Hou, B. M. Harvey and
P. E. Hoggard (2013) Catalysis of the photodecomposition of chloro-
form by a polystyrene anion exchange resin. Curr. Cat. 2, 79–87.
20. Bio-Rad Laboratories. AG 1, AG MP-1 and AG 2 Strong Anion
Exchange Resin Instruction Manual (LIT212 Rev C).
44. Yi, H.-B., F.-F. Xia, Q. Zhou and D. Zeng (2011) [CuCl3]ꢀ and
[CuCl4]2ꢀ hydrates in concentrated aqueous solution: A density
functional theory and ab Initio study. J. Phys. Chem. A 115, 4416–
4426.
21. Du, H., R.-C. A. Fuh, J. Li, L. A. Corkan and J. S. Lindsey (1998)
Photochem CAD: A computer-aided design and research tool in pho-
tochemistry. Photochem. Photobiol. 68, 141–142.
22. Lindsey, J. S. PhotochemCAD, spectra recorded by Junzhong Li and
Richard W. Wagner. http://omlc.ogi.edu/spectra/PhotochemCAD/
html/index.html, Accessed on 1 April 2013.
45. Hitchman, M. A. and P. J. Cassidy (1979) Polarized crystal spectrum
of bis(methylphenethylammonium) tetrachlorocuprate(II): Analysis of
the energies, vibrational fine structure, and temperature dependence
of the “d-d” transitions of the planar tetrachlorocuprate(2–) ion.
Inorg. Chem. 18, 1745–1754.
~
46. Pena, L. A. and P. E. Hoggard (2010) Hexachlororhodate(III) and
23. Stone, A. and E. B. Fleischer (1968) The molecular and crystal
structure of porphyrin diacids. J. Am. Chem. Soc. 90, 2735–2748.
24. Harvey, B. M. and P. E. Hoggard (2012) Determination of micromo-
lar concentrations of strong acids in chloroform by protonation of
tetraphenylporphyrin. Monatsh. Chem. 143, 1101–1105.
25. Solis Montiel, E. and J. A. Solano (1986) Spectrophotometric analy-
sis for chlorine by the extraction of triiodide formed in chloroform
solution of tetrabutylammonium perchlorate. Ingenieria y Ciencia
Quimica 10, 45–48.
the photocatalytic decomposition of chloroform. J. Mol. Catal. A:
Chem. 327, 20–24.
47. Kitaigorodskii, A. N., A. V. Kessenikh and A. V. Bulatov (1981)
Inner-sphere coordination of chloroform with triphenylphosphine
complexes of cobalt(2+) and nickel(2+). Zh. Fiz. Khim. 55, 1200–
1203.
48. Gonzalez, M. C., G. C. Le Roux, J. A. Rosso and A. M. Braun
(2007) Mineralization of CCl4 by the UVC-photolysis of
hydrogen peroxide in the presence of methanol. Chemosphere 69,
1238–1244.
49. Hasson, A. S. and I. W. M. Smith (1999) Chlorine atom initiated
oxidation of chlorinated ethenes: Results for 1,1-dichloroethene
(H2C=CCl2), 1,2-dichloroethene (HClC=CClH), trichloroethene
(HClC=CCl2) and tetrachloroethene (Cl2=CCl2). J. Phys. Chem. A
103, 2031–2043.
26. Olsen, J. C., G. E. Ferguson, V. J. Sabetta and L. Scheflan (1931)
Quantitative determination of phosgene. Ind. Eng. Chem. 3,
189–191.
~
27. Hoggard, P. E., L. R. Cohen, L. A. Pena, B. M. Harvey and A. M.
Chan (2013) The dependence of photocatalytic reaction yields on
catalyst mass in solid-liquid suspensions. Curr. Cat. 2, 2–6.
28. Hudgens, J. W., R. D. III Johnson, R. S. Timonen, J. A. Seetula and
D. Gutman (1991) Kinetics of the reaction trichloromethyl + bromine
50. Horvath, O. and K. L. Stevenson (1993) Charge Transfer Photo-
chemistry of Coordination Compounds. VCH, Weinham, Germany.

1242 Brent M. Harvey and Patrick E. Hoggard
~
51. Osman, A. H. and A. A. M. Aly (1992) Photooxidation of bis(eth-
ylxanthato)nickel(II) containing aromatic nitrogen heterocyclic
ligands in chloroform. Monatsh. Chem. 123, 309–314.
52. Sandrini, D., M. Maestri, V. Balzani, L. Chassot and Z. A. Von
(1987) Photochemistry of the orthometalated cis-bis[2-(2-thienyl)pyr-
idine]platinum(II) complex in halocarbon solvents. J. Am. Chem.
Soc. 109, 7720–7724.
53. Vogler, A. and H. Kunkely (1982) Photooxidation of 1,2-dithiolene
complexes of nickel, palladium, and platinum in chloroform. Inorg.
Chem. 21, 1172–1175.
56. Chan, A. M., L. A. Pena, R. E. Segura, R. Auroprem, B. M. Harvey,
C. M. Brooke and P. E. Hoggard (2013) Photocatalysis of chloro-
form decomposition by the hexachlororuthenate(IV) ion. Photochem.
Photobiol. 89, 274–279.
57. Bae, E. and W. Choi (2003) Highly enhanced photoreductive degra-
dation of perchlorinated compounds on dye-sensitized metal/TiO2
under visible light. Environ. Sci. Technol. 37, 147–152.
58. Chatterjee, D. and S. Dasgupta (2005) Visible light induced photo-
catalytic degradation of organic pollutants. J. Photochem. Photobiol.
C 6, 186–206.
54. Osman, A. H., A. A. M. Aly, E.-M. M. Abd and G. A. H. Gouda
(2004) Photoreactivity and thermogravimetry of copper(II) com-
plexes of N-salicylideneaniline and its derivatives. Bull. Korean
Chem. Soc. 25, 45–50.
55. Sigma-Aldrich, Ion Exchange Resins: Classification and Properties.
http://www.sigmaaldrich.com/etc/medialib/docs/Aldrich/Instructions/
ion_exchange_resins.Par.0001.File.tmp/ion_exchange_resins.pdf,
accessed Aug 14, 2014.
59. Chen, B., W. Lee, P. K. Westerhoff, S. Krasner and P. Herckes
(2010) Solar photolysis kinetics of disinfection byproducts. Water
Res. 44, 3401–3409.
60. Kubelka, P. (1948) New contributions to the optics of intensely
light-scattering materials. Part I. J. Opt. Soc. Amer. 38, 448–457.
61. Kubelka, P. and F. Munk (1931) Ein Beitrag zur Optik der Farban-
striche. Zeitschrift f€ur Technische Physik 12, 593–601.