Photobehavior of aqueous uranyl ion and photo-oxygenation of isobutane using light from the visible region

Article abstract of DOI:10.1139/v03-026

The photochemical and photophysical behavior of the aqueous uranyl ion [UO₂(H₂O)₅]²⁺ has been studied under the influence of visible light and with added perchloric acid over the range of 0.01-4 M. In the presence of 2-methylpropane (isobutane), photo-oxygenation of isobutane occurs to yield, as the major product, 2-methyl-2-propanol (tert-butyl alcohol) along with lesser amounts of 2-methyl-2-propene (isobutene) and other C1-C8 products. The quantum yield for formation of tert-butyl alcohol is independent of light intensity at the irradiation wavelength of 415 nm and of uranyl concentration, but it increases from 0.016 ± 0.001 at 0.01 M HClO₄ (pH 2) to 0.13 ± 0.01 at 4 M HClO₄. The emission spectrum from the electronically excited uranyl ion and the associated quantum yields have been measured in the presence and absence of isobutane, as a function of added perchloric acid. While in both cases the shape of the spectrum remains invariant, the quantum yields increase with increasing perchloric acid concentration. The strong dependence on added perchloric acid is interpreted within the context of the presence and interconversion of two electronically excited species, an acid form, *[UO₂(H₂O)₅]²⁺, and a base form, *[UO₂(H₂O)_n(OH)]⁺. It is proposed that both forms react with isobutane to give a tert-butyl radical, and that oxidation of coordinated aqua ligands occur, the latter generating a hydroxyl radical whose reaction with isobutane rapidly leads also to a tert-butyl radical. The reaction of this alkyl radical with ground-state [UO₂(H₂O)₅]²⁺ then gives rise to the stable tert-butyl alcohol product and reduced forms of uranyl ion. Based upon the values of the quantum yields and of excited-state lifetime measurements reported in the literature, a comprehensive mechanism has been developed in a quantitative manner to provide calculated values of the rate constants for the individual mechanistic steps. The calculated rate constants provide a basis to calculate the values of quantum yields for emission and chemical reaction, as well as for lifetimes, that agree very satisfactorily with the experimental values over a 400-fold concentration change in added perchloric acid.

Full text of DOI:10.1139/v03-026

219
Photobehavior of aqueous uranyl ion and photo-
oxygenation of isobutane using light from the
visible region
Trevor M. Bergfeldt, William L. Waltz, Xiangrong Xu, Petr Sedlák, Uwe Dreyer,
Hermann Möckel, Jochen Lilie, and John W. Stephenson
Abstract: The photochemical and photophysical behavior of the aqueous uranyl ion [UO2(H2O)5]2+ has been studied
under the influence of visible light and with added perchloric acid over the range of 0.01–4 M. In the presence of 2-
methylpropane (isobutane), photo-oxygenation of isobutane occurs to yield, as the major product, 2-methyl-2-propanol
(tert-butyl alcohol) along with lesser amounts of 2-methyl-2-propene (isobutene) and other C1–C8 products. The quan-
tum yield for formation of tert-butyl alcohol is independent of light intensity at the irradiation wavelength of 415 nm
and of uranyl concentration, but it increases from 0.016 0.001 at 0.01 M HClO4 (pH 2) to 0.13 0.01 at 4 M
HClO4. The emission spectrum from the electronically excited uranyl ion and the associated quantum yields have been
measured in the presence and absence of isobutane, as a function of added perchloric acid. While in both cases the
shape of the spectrum remains invariant, the quantum yields increase with increasing perchloric acid concentration. The
strong dependence on added perchloric acid is interpreted within the context of the presence and interconversion of two
electronically excited species, an acid form, *[UO2(H2O)5]2+, and a base form, *[UO2(H2O)n(OH)]+. It is proposed that
both forms react with isobutane to give a tert-butyl radical, and that oxidation of coordinated aqua ligands occur, the
latter generating a hydroxyl radical whose reaction with isobutane rapidly leads also to a tert-butyl radical. The reac-
tion of this alkyl radical with ground-state [UO2(H2O)5]2+ then gives rise to the stable tert-butyl alcohol product and
reduced forms of uranyl ion. Based upon the values of the quantum yields and of excited-state lifetime measurements
reported in the literature, a comprehensive mechanism has been developed in a quantitative manner to provide calcu-
lated values of the rate constants for the individual mechanistic steps. The calculated rate constants provide a basis to
calculate the values of quantum yields for emission and chemical reaction, as well as for lifetimes, that agree very sat-
isfactorily with the experimental values over a 400-fold concentration change in added perchloric acid.
Key words: photo-oxidation, photo-oxygenation, uranyl ion, isobutane, tert-butyl alcohol, lifetime, quantum yield, acid–
base dissociation. 229
Résumé : On a étudié les comportements photochimique et photophysique de l’ion uranyle aqueux [UO2(H2O)5]2+ sous
l’influence de la lumière visible, en présence d’acide perchlorique ajouté à des concentrations allant de 0,01 M à 4 M.
En présence de 2-méthylpropane (isobutane), il se produit une photooxygénation de l’isobutane qui conduit à la forma-
tion de 2-méthylpropan-2-ol (tert-butanol) comme produit principal aux côtés de quantités plus faibles de 2-méthylprop-
2-ène (isobutène) ainsi que d’autres produits comportant de un à huit atomes de carbone. À une longueur d’onde
d’irradiation de 415 nm, le rendement quantique pour la formation du tert-butanol est indépendant de l’intensité de la
lumière et de la concentration en uranyle, mais il augmente de 0,016 0,001 à une concentration de HClO4 de 0,01 M
(pH = 2) jusqu’à 0,13 0,01 à une concentration de HClO4 de 4 M. On a mesuré le spectre d’émission de l’ion ura-
nyle électroniquement excité et les rendements quantiques associés, en présence et en l’absence d’isobutane et en fonc-
tion de la quantité d’acide perchlorique ajouté. Même si dans les deux cas la forme des spectres ne varie pas, les
rendements quantiques augmentent avec une augmentation de la concentration d’acide perchlorique. Cette forte dépen-
dance sur l’acide perchlorique ajouté est interprétée dans le contexte de la présence et de l’interconversion de deux es-
pèces électroniquement excitées, une forme acide *[UO2(H2O)5]2+ et une forme basique *[UO2(H2O)n(OH)]+. Il est
Received 10 October 2002. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 19 March 2003.
Contribution from the Department of Chemistry and Department of Mathematics and Statistics, University of Saskatchewan,
110 Science Place, Saskatoon, SK S7N 5C9, Canada and the Hahn-Meitner-Institut Berlin GmbH, Glienicker Str. 100, D-14109
Berlin, Germany.
Trevor M. Bergfeldt, William L. Waltz,1 Xiangrong Xu, Petr Sedlák. Department of Chemistry, University of Saskatchewan,
110 Science Place, Saskatoon, SK S7N 5C9, Canada.
Uwe Dreyer, Hermann Möckel, Jochen Lilie. Hahn-Meitner-Institut Berlin GmbH, Glienicker Str. 100, D-14109 Berlin, Germany.
John W. Stephenson. Department of Mathematics and Statistics, University of Saskatchewan, 106 Wiggins, Saskatoon,
SK S7N 5E6, Canada.
1Corresponding author (e-mail: bill.waltz@usask.ca).
Can. J. Chem. 81: 219–229 (2003)
doi: 10.1139/V03-026
© 2003 NRC Canada

220
Can. J. Chem. Vol. 81, 2003
suggéré que les deux formes réagissent avec l’isobutane pour produire le radical tert-butyle et qu’il se produit une oxy-
dation des ligands aqua coordinés qui génèrent la formation du radical hydroxyle qui, par réaction avec l’isobutane,
conduit aussi rapidement à la formation du radical tert-butyle. La réaction de ce radical avec du [UO2(H2O)5]2+ à l’état
fondamental conduit alors à la formation du tert-butanol, le produit stable, ainsi qu’aux formes réduites de l’ion ura-
nyle. Sur la base des valeurs des rendements quantiques et des mesures de temps de vie rapportées dans la littérature,
on a développé un mécanisme global de manière quantitative qui permet d’obtenir des valeurs calculées pour les cons-
tantes de vitesse de chacune des étapes mécanistiques individuelles. Les constantes de vitesse calculées ont permis de
calculer des valeurs des rendements quantiques de l’émission et de la réaction chimique ainsi que des temps de vie qui
sont en bon accord avec les valeurs expérimentales pour des changements de concentration en acide perchlorique
ajouté qui varient par un facteur allant jusqu’à quatre cents.
Mots clés : photooxydation, photooxygénation, ion uranyle, isobutane, tert-butanol, temps de vie, rendement quantique,
dissociation acide-base.
[Traduit par la Rédaction] Bergfeldt et al.
However, two additional features need to be considered in
Introduction
evaluating our results. There is now substantial evidence that
because of the powerful oxidizing nature of *[UO2(H2O)5]2+,
water can be oxidized to form a hydroxyl radical that will
readily react with isobutane to form further amounts of tert-
butyl radical (1, 10, 19–21).
This investigation was undertaken with two complemen-
tary objectives in mind: namely, to investigate in a quantita-
tive manner the photochemical and photophysical features of
the electronically excited uranyl ion, *[UO2(H2O)5]2+, and at
the same time to study the use of this species as a quantita-
tive means to convert 2-methylpropane (isobutane) into 2-
methyl-2-propanol (tert-butyl alcohol). Baird and Kemp
have recently provided an in-depth review of the many facets
of the intriguing yet complicated photobehavior of the ura-
nyl ion and other related forms found in aqueous media (1).
Uranyl ion absorbs in the 350–500 nm region to yield a
long-lived excited state (µs timescale); however, its behavior
is sensitive to pH, temperature, and the nature and concen-
tration of the counter anion. Under our conditions, where we
have used the perchlorate salt and added amounts of
perchloric acid (0.01–4 M) to vary the pH, the ground state
species is the monomeric form [UO2(H2O)5]2+. At the higher
concentrations of added HClO4, ion pairing between
[2]
[3]
*[UO2(H2O)5]2+ → [UO2(H2O)4]+ + H+ + ·OH
·OH + (CH3)3CH → H2O + (CH3)3C·
Above, we have portrayed that the oxidation of water is
of coordinated water because Xu (21) finds that when the
coordinated waters are replaced by a fluoride ion, the oxi-
dation of isobutane is not detectable. The observations by
many researchers that the excited-state reactivity and emis-
sion lifetime are strongly dependent on the acidity of the
media has given rise to different mechanistic interpretations.
Marcantontos proposed, at relatively high concentrations of
uranyl ion, the additional formation of an exciplex complex,
*[U2O4H]2+, to account for bi-exponential lifetimes, multiple
emission spectra, and dependence of such on concentrations
of acid and uranyl ion (1, 12). Under our conditions we do
not observe bi-exponential behavior or multiple emission,
and thus we conclude that exciplex formation is not a signif-
icant factor in our study. Burrows and co-workers initially
offered a mechanism focusing on the interconversion be-
tween two closely spaced excited states of the uranyl ion (1,
22). More recently they have adopted the proposal made by
Darmanyan, Khudyakov, and co-workers of an electronically
excited acid–base mechanism in competition with other re-
activity modes (22, 23); the number of waters of hydration
for the base form, *[UO2(H2O)n(OH)]+, are unknown.
–
*[UO2(H2O)5]2+ and ClO4 is likely to occur to some extent
in the excited-state species (2); however, our results can be
satisfactorily interpreted without invoking this facet.
Electronically excited uranyl ion is a powerful oxidant
(E° ~ 2.6 V) and is well known to engage in electron-
transfer processes and to participate in hydrogen-abstraction
reactions with many organic compounds (1, 3–16). Thus one
anticipates that in the presence of isobutane, the excited state
will abstract a hydrogen atom to generate an alkyl radical
and the uranoyl ion, [UO2]+, leading to subsequent free radi-
cal reactions to form the organic products. Mao and Bakac
have found with related alkanes and alcohols that the loga-
rithm of the quenching constant for excited uranyl ion in-
creases with decreasing C—H bond energy, which is
supportive of hydrogen-abstraction mechanisms (17). The
extension of this observation to isobutane implies that hy-
drogen abstraction will occur preferentially at the tertiary
hydrogen rather than at the primary positions, and our find-
ings that the tertiary alcohol is the major product in
deaerated media — whereas acetone occurs in the presence
of dioxygen — is commensurate with the major occurrence
of the tert-butyl radical (18).
[4]
[5]
*[UO2(H2O)5]2+ → *[UO2(H2O)n(OH)]+ + H+ k12
*[UO2(H2O)n(OH)]+ + H+ → *[UO2(H2O)5]2+ k21
Our findings are consistent with these acid–base steps, as
shown in more detail in the Discussion section (see Fig. 1).
In this report, we present a comprehensive mechanism incor-
porating acid–base behavior, radiative and nonradiative pro-
cesses, and quenching by water and isobutane; specific
values are provided for the rate constants associated with the
individual steps of the mechanism. A comprehensive and
self-consistent set of values has not been available, to our
knowledge, until now. The estimates of the rate constant val-
[1]
*[UO2(H2O)5]2+ + (CH3)3CH → [UO2(H2O)5]+
+ H+ + (CH3)3C·
© 2003 NRC Canada

Bergfeldt et al.
221
Fig. 1. Schematic diagram of the proposed acid–base model. Note: Processes involving the acid and base forms are distinguished by
subscripts 1 and 2, respectively; Ia is the intensity of absorbed light; and BH and B. are the isobutane and tert-butyl radical, respec-
tively.
ues have been achieved by using an integrated package of
Fig. 2. Variable pathlength cell.
information obtained from both time-resolved and steady-
state experiments. Specifically we have measured the inte-
grated emission intensities that lead to emission quantum
yields and chemical quantum yields for tert-butyl alcohol as
a function of acidity. We have used and analyzed in detail
the lifetime measurements reported by Moriyasu et al. after
we first confirmed that their findings are appropriate to our
conditions (13).
The choice of isobutane used here in the quenching and
photo-oxygenation processes warrants comment because it
has associated with it several useful and advantageous fea-
tures. Nizova and Shul’pin appear to have been the first to
report on the photo-oxygenation of alkanes (cyclohexane
and methane) with uranyl ion, although such was performed
in acetic acid and acetonitrile solvents (24). Subsequently
Mao and Bakac extended such studies to aqueous 0.1 M
H3PO4 for alkyl aromatic compounds (17). Of concern in
such studies is that the products of the reactions are known
or expected to be highly reactive in their own right. To mini-
mize such secondary reactions is difficult; however, our ob-
served major product tert-butyl alcohol is considerably less
reactive as a quenching agent and in its reaction with
hydroxyl radical than is isobutane. Even to a much higher
conversion (ca. 65%) than normally done here (ca. 10%), we
find no evidence for complications arising from the reactiv-
ity of the alcohol product. As a further precaution, we have
in general carried out the experiments with continuous bub-
bling of the isobutane gas through the liquid phase before
and during photolysis.
CWL415, λmax = 415 nm, fwhm = 12 nm) was placed in the
light beam prior to the reaction cell. Two types of reaction
cells (water-cooled, 25.0 0.15°C) were used: one of fixed
path length (8.7 cm) and the second of variable path length
(ca. 5 cm to 10 cm). The latter was constructed from a
100 mL syringe (J. Young Scientific, GTS/100/GS) with the
end of the barrel having — affixed to it by optical adhesive
(Norland 61) — a borosilicate window. A schematic dia-
gram of this cell is shown in Fig. 2. A major advantage of
this arrangement is that periodic sampling of the liquid solu-
tion can be done during the course of irradiation without the
development of a separate gas phase. For ease of carrying
out the actinometry, both the actinometric solution and the
uranyl solutions have been irradiated under optically dense
conditions (absorbance = 2–2.5). Because of this procedure
and because of the longer light path, a more dilute concen-
tration of the ferrioxalate actinometry solution (3.9 mM) was
used than is generally employed (25); however, all such
measurements have been carried out using linear calibration
plots. Spectroscopic data were obtained using a Cary 2315
or HP 8451A PhotoDiode Array spectrophotometer for UV–
vis spectra; Spex Fluorolog-2 for emission spectra; Bio-Rad
FTS-40 FT-IR and FT-NMR, Bruker AM-300 MHz (1H).
Mass spectrometry was done with a Fisions GC 8000 (in
splitless mode) interfaced with a VG70-VSE mass spectrom-
eter with a Dec Vax 4000 model 60 workstation; authentic
Experimental
Apparatus
For steady-state measurements of quantum yields and
product analysis, a 1000 W Hg–Xe arc lamp (Oriel 6293)
was used with the collimated light beam being passed
through an 11-cm water cell and a 9.6-cm cell of 0.1 M
CuSO4(aq) (310 nm < band pass < 640 nm) to filter out IR
and UV light. In addition, either a cut-off filter (Schott Glass
Technologies Inc., GG400) or, for quantum-yield determina-
tions, an interference filter (Edmund Scientific Co.,
© 2003 NRC Canada

222
Can. J. Chem. Vol. 81, 2003
samples and the NIST library were used to identify products.
Gamma radiolysis experiments of saturated isobutane with-
out the presence of uranyl ion were performed with a
Gammcell 60Co source (Serial #260, Atomic Energy of Can-
ada Limited (AECL)) using either Millipore or triply dis-
tilled water, and acidified with HClO4 (Ultrapure, BDH).
These radiolysis results allow us to compare the nature of
the organic products arising from the reaction of the
hydroxyl radical with isobutane to those products formed in
the photolysis of the uranyl–isobutane solutions.
Materials and solutions
Aqueous solutions were prepared from water purified by a
Millipore Super-Q system, and where compared with those
of triply distilled water, the results were found to be the
same. The chemicals employed were: isobutane (Linde, in-
struments grade; Aldrich, 99.9%); isobutene (Matheson,
99.0% min); 2,2,3,3-tetramethylbutane (Aldrich, 99%); tert-
butyl and iso-butyl alcohols (BDH, reagent); and
tetrahydrofuran (BDH, AnalaR grade). Identifications by GC
retention times of C1–C4 alkanes and alkenes were made us-
ing a Matheson standard gas mixture. Uranium solutions
were prepared from UO2(ClO4)2·6H2O (ROC/RIC, reagent;
reclaimed uranium) or more generally from UO3 (Cameco,
99.4%, natural abundance). As these two sources gave the
same results, the background radioactivity of the uranium
was not a detectable factor. Uranyl stock solutions were typi-
cally prepared by dissolving 178 g UO3 in 102 mL of 70%
perchloric acid (BDH, analytical grade) and 150 mL water,
heating at 80°C until dissolution followed by filtering, and
allowing the solution to stand for 1 day before use. The con-
centration of uranyl ion was determined spectrophotometri-
cally using a calibration curve based on dissolved UO3 in
perchloric acid, and this was verified by using U3O8 as an al-
ternate source. The latter was prepared from heating, at 800°C
for 4 h, uranyl acetate (Fisher, ACS grade), and the uranium
content was measured by laser phosphorescence and delayed
neutron counting (Saskatchewan Research Council). The mo-
lar absorption coefficients, at 414 nm, of [UO2(aq)]2+ in aque-
ous perchloric acid media were 7.8 0.2 M–1 cm–1 (pH 1.1),
Chromatography
A Hewlett-Packard 5890 GC was employed in conjunction
with a PC286 computer with Baseline 810 software. The in-
jection port and FI-detector temperatures were 200°C. For
aqueous solutions, injections were either 1 mL samples taken
directly from the photolysis cell, 1 mL tetrahydrofuran (THF)
extractions, or an injection of solid phase micro-extraction
fibres. THF extractions were prepared as follows: a sample
was withdrawn from the photolysis cell with a 5-mL syringe
and passed through a Sep-Pak cartridge (Millipore, Sep-Pak
Plus, t-C18) using a syringe pump (rate 1 mL min–1, Sage
Instruments 341B). The retained organic material was then
eluted at the same rate with 1.2 mL of THF solution contain-
ing isopropanol as the GC internal standard followed by
1.5 mL of THF. Extraction using solid phase micro-
extraction (SPME) involved the following procedure: a sam-
ple was withdrawn from the photolysis cell with a 5 mL sy-
ringe and transferred to a 2 dram vial containing 2 g NaCl
and a magnetic stirring bar; 0.5 mL 1,4-dioxane was added
as the GC internal standard. The SPME fibre (100 mm
polydimethyl-siloxane, Supelco) was introduced into the
headspace of the sealed vial and the solution magnetically
stirred for a period of 10 min in a thermostated water jacket
set at 50°C. The fibre was then introduced into the injection
port of the GC and left to desorb for 5 min. There was found
to be no difference between results obtained by Sep-Pak and
SPME methods ( 10%). The use of these two different pro-
cedures gives considerable confidence as to the validity of
the results in the context of our desire to keep the percent
product formation as low as possible (normally <10%). The
use of the SPME technique also permitted us to carry out re-
actions at much higher concentrations of HClO4 (>0.1 M)
than was possible with the Sep-Pak cartridge approach. To
circumvent the degradation of the SPME fibres in the
headspace at extreme acid conditions, appropriate amounts
of aqueous NaOH were added to bring the pH of the solu-
tions up to approximately one. For gaseous samples, 5–
100 mL gas-tight syringes were used to draw samples from
the reaction vessel. For hydrocarbon substances of a carbon
number less than 5, a GSQ column (J&W Scientific, 30 m ×
0.546 mm) was used, while a DB-210 column (J&W Scien-
tific, 30 m × 0.53 mm) was employed for all other analyses.
Calibration of the GC for isobutane was done by injection of
weighted amounts of isobutane dissolved in isopropanol. For
our typical irradiated solution (ca. 35 mM aqueous uranyl
solution at pH = 1 (HClO4)), the solubility of isobutane was
found to be 0.91 0.02 mM at 22°C and 95.33 kPa partial
pressure of isobutane; this compares very favorably with that
of 0.94 mM isobutane in pure water at 25°C and
101.325 kPa (26).
8.0
0.2 M–1 cm–1 (pH 2.0), and 8.3
0.2 M–1 cm–1
(pH 3.0); Bell et al. (27) reported 7.7 M–1 cm–1 at pH 0.95
(HClO4).
Uranium(V) was measured spectrophotometrically, in the
presence of [UO2(aq)]2+, where a complex [U2O4]3+ occurs
with an absorption peak at 740 nm (ε = 24.3 M–1 cm–1) (28,
29). Uranium(IV) in aqueous perchloric acid media (pH ~ 1)
was prepared by photoreduction of uranyl ion in the pres-
ence of ethanol, and its concentration was determined by us-
ing potassium dichromate oxidation – titration (30). From
this procedure, the molar absorption coefficient at 652 nm
was determined to be 43.2 0.9 M–1 cm–1 (pH 1.0, 22°C),
which agrees with that reported by Betts (31).
Results
Our most frequently used condition was a pH of 1 and
35 mM [UO2(aq)]2+ with the solutions being deaerated and
saturated with isobutane by continuously bubbling with iso-
butane gas during and prior to irradiation (30 min or more at
room pressure before irradiation to achieve saturation). For
specific cases, the concentration of uranyl ion was varied
from 25–148 mM, the light was changed from 2.1 × 10–7
–
1.2 × 10–6 ein L–1 s–1 (1 ein = 1 mol of light photons), and
the concentration of HClO4 ranged from 0.01–4 M. At pHs
of about 2 and higher, polyuranium species occur, leading in
part to bi-exponential decay of the electronically excited ura-
nyl ion. To confirm the mono-exponential decay behavior re-
ported by Moriyasu and co-workers (13, 14) for conditions
very similar to our perchloric acid media, lifetime measure-
ments were performed using laser excitation at 347 nm with
a pH of 1 for both argon-saturated and aerated solutions. We
© 2003 NRC Canada

Bergfeldt et al.
223
Fig. 3. Plot of the experimental ( ) and calculated (᭝) values of
the emission quantum yield of the electronically excited uranyl
ion [UO2(aq)]2+ in the absence of isobutane, as a function of the
concentration of HClO4 at 34.8 mM [UO2(aq)]2+. Average root-
mean-square difference between experimental and calculated val-
ues is 8%.
Fig. 4. Plot of the experimental ( ) and calculated (᭝) values of
the emission quantum yield of the electronically excited uranyl
ion in the presence of isobutane, as a function of the concentra-
tion of HClO4 at 34.8 mM [UO2(aq)]2+. Average percent differ-
ence between experimental and calculated values is 21%.
10 mM, lower than for 35 mM uranyl ion, the calculated
quantum yields for emission were the same within experi-
mental error. Of particular note is that while the emission in-
tensities increased with perchloric acid concentration, the
spectral features for the peak maxima and minima and their
relative intensities were invariant. Thus we conclude that un-
der our conditions, the emission is predominantly from one
species, and because the emission intensities are greatest at
high concentrations of HClO4, the emission can be ascribed
to *[UO2(H2O)5]2+, the acid form. Baird and Kemp in their
review (1) have proposed that the absolute emission quan-
tum yield value of 0.0055 for 1.0 M HClO4, found by
Katsumura et al. (33), was the most reliable measurement,
and we have used this value to convert our integrated emis-
sion intensities into the quantum yields shown in Figs. 3 and 4.
Aqueous uranyl solutions saturated with isobutane and
kept in the dark are found to be thermally unreactive. Upon
exposure to steady-state irradiation at 415 nm radiation, the
major products are tert-butyl alcohol (t-BuOH) and the inor-
ganic products [UO2(aq)]+ and [U(aq)]4+, accompanied by
minor amounts of isobutene and other organic products. The
U(IV) observed is likely due to the disproportionation of
uranoyl [UO2(aq)]+ (28, 34):
find the emission decay (510 nm), as well as the excited-
state absorption decay (590 nm), to exhibit a single expo-
nential decay with a lifetime of 2.2 0.1 µs. This single-
decay behavior and the value of the associated lifetime are
in agreement with the findings of Moriyasu and co-workers,
as well as those of other reports (1, 7, 13–16). Results pre-
sented earlier by us indicate that quenching by isobutane is
about 3 × 107 M–1 s–1 (20). In our modeling studies, as put
forward in the Discussion section, this value has proven use-
ful as a guide; however, the value is, from our current find-
ings, indicative of the quenching by isobutane of two
excited-state species coupled through acid–base processes
(see below).
As one would anticipate from the presence of acid–base
processes, the emission quantum yields are strongly depend-
ent on the acidity of the medium, as shown in Figs. 3 and 4
(solid circles). These quantum yields and emission spectra
were determined as follows. The emission spectrum, cen-
tered at the 506 nm peak, is broad and overlaid with a vibra-
tional structure (1, 32); maxima or estimated maxima for
shoulders occurred at 470, 484, 506, 530, and 546 nm. The
positions of the maxima and their relative intensities did not
change whether the solutions were aerated or saturated with
either oxygen or argon, and thus, in general the data was ob-
tained for aerated media. The emission spectrum was re-
corded two to four times at each concentration of perchloric
acid (0.01–4 M), and then for each recording the emission
intensities from 430–600 nm were integrated with reference
to a Rhodamine B quantum counter (Spex Fluorolog-2
Spectrofluorometer) and averaged for each set of perchloric
acid concentrations. The reproducibility of the integrated
emission values was good, being within several percent. To
minimize the possibility of systematic error, the measure-
ments for different acid concentrations were done in a ran-
dom manner. The emission spectra were recorded for 35 and
10 mM uranyl ion; the latter concentration was used by
Moriyasu et al. in their lifetime studies (0.01–10 M HClO4)
(13). While the integrated emission intensities were, for
[6]
2[UO2(aq)]+ + 4H+ → [UO2(aq)]2+
+ [U(aq)]4+ + 2H2O
At pH 3, where the disproportionation of uranoyl to U(IV)
is slow, the visible spectrum clearly shows the characteristic
peaks of [U(aq)]4+ at 652 nm and of the U(V)–U(VI) com-
plex [U2O4]3+ at 400 and 740 nm, whereas only [U(aq)]4+ is
seen at pH 1 (28, 34). The results of our mass-balance ex-
periments at pH 1 show the ratios of d[U(VI)]/d[U(IV)] and
d[isobutane]/d[U(IV)] to be (1.0 0.1)/1 and (0.9 0.1)/1,
respectively.
The organic products are indicative of C—H bond activa-
tion. The tertiary alcohol accounts for 85 5% of the loss of
isobutane, along with isobutene (ca. 5%) and, to a lesser
amount, other C1–C8 products. Although isobutene can un-
© 2003 NRC Canada

224
Can. J. Chem. Vol. 81, 2003
Fig. 5. Plot of the experimental ( ) and calculated (᭝) quantum
yields of t-BuOH as a function of the concentration of HClO4 at
34.8 mM [UO2(aq)]2+. Average root-mean-square difference be-
tween experimental and calculated values is 8%.
dergo hydrolysis to yield t-BuOH, we find that the thermal
rate is approximately two to three times slower than the
photoproduction of t-BuOH. This is in accord with the re-
sults of a previous study (35), and therefore, it seems un-
likely that the observed t-BuOH is the result of the
hydrolysis of isobutene. The possible bimolecular self-
recombination products arising from tert-butyl and (or) iso-
butyl radicals (i.e., 2,2,3,3-tetramethylbutane, 2,2,4-
trimethylpentane, and 2,5-dimethylhexane) are found only in
trace amounts. Gamma radiolysis of aqueous solutions at
pH 2, saturated with isobutane but without uranyl ion, re-
sults in the major products (>80%) being the aforementioned
bimolecular self-recombination products, while t-BuOH is a
minor product (~20%). Additionally, when an aqueous
acidic uranyl ion solution containing both oxygen and isobu-
tane is irradiated, the major products are acetone and tert-
butyl alcohol (18). These results all point to the tert-butyl
radical as the common reactive intermediate resulting from
hydrogen abstraction. The overall major chemical reaction at
pH 1 can therefore best be described as
[7]
[UO2(aq)]2+ + (CH3)3CH + 2H+ → [U(aq)]4+
tion products. Our earlier report on the photo-oxidation of
water in aqueous uranyl solutions using conductivity detec-
tion shows the formation of a hydroxyl radical (eq. [2]).
Thus, one recognizes that both hydrogen abstraction
(eq. [1]) and hydroxyl radical reaction with isobutane
(eq. [3]) lead to generation of the tert-butyl radical, desig-
nated as B·, and its subsequent conversion to tert-butyl alco-
hol (t-BuOH).
The emission and lifetimes of the uranyl ion in aqueous
acidic solutions are well documented to be dependent on pH
(1, 2, 13, 14, 16, 22, 23, 36–39). Our results for the emission
quantum yield and quantum yields for formation of t-BuOH
show that the yields increase with increasing HClO4
(Figs. 3–5). The yields of t-BuOH range from 0.016 at
0.01 M HClO4 to 0.13 at 4 M HClO4 ( 10%).
We have interpreted our results in terms of the presence of
acid–base processes for two electronically excited species, to
account for this strong dependence on acid concentration.
The electronically excited “acid” form is denoted as
*[UO2(H2O)5]2+, and it is the nascent species formed upon
absorption of visible light by the ground-state species
[UO2(H2O)5]2+, according to the Franck–Condon Principle.
Subsequent and rapid loss of a proton from one of the coor-
dinated aqua ligands (eq. [4]) can then lead to the formation
of the “base” form *[UO2(H2O)n(OH)]+; however, reproto-
nation of this conjugate base must occur to some extent, to
account for the observed dependence on acid concentration.
The important roles played by coordinated water both in
these acid–base steps and in the nonradiative events is un-
derscored by the fact that the replacement of aqua ligands by
fluoride and phosphate anions, which are not easily oxi-
dized, leads to much longer lifetimes (1, 14).
The excited-state portion of our mechanism is given in
Fig. 1. We have assumed that both the acid and base forms
undergo reactions with isobutane to form the tert-butyl radi-
cal (CH3)3C·, and that both excited-state species can oxidize
coordinated water to give rise to a hydroxyl radical; the re-
sults of our modeling studies, given below, support this in-
terpretation. The second-order rate constants kq1 and kq2
+ (CH3)3COH + H2O
The chemical quantum yield measurements were con-
strained to t-BuOH, owing to the low concentrations of the
other products. The quantum yield (Φ) for formation of t-
BuOH was determined from the slope of plots of the quo-
tient (alcohol concentration) × (irradiation volume)/(ab-
sorbed light intensity) vs. time. Where the conversion of
isobutane was high (upwards of 65%, as in the mass-balance
experiments), curvature in the plots occurred. Linear plots
were obtained, however, when isobutane was bubbled
through the reaction solution during the entire irradiation pe-
riod. At pH 1, the average observed quantum yield of t-
BuOH was 0.022 0.002, as determined from the slopes. In
the mass-balance experiments, the curvature was well de-
scribed by a second order polynomial. The average coeffi-
cient for the linear term gave the same average Φ value as
that previously cited for the continuous bubbling experi-
ments. This comparison indicates that continuous bubbling
of isobutane did not give rise to detectable loss of t-BuOH
from the liquid phase. The chemical quantum yield is inde-
pendent of light intensity (2.1 × 10–7 – 1.2 × 10–6 ein L–1 s–1)
and of uranyl concentration (25–148 mM) but is dependent
on the concentration of added HClO4, as shown in Fig. 5
(solid circles).
Discussion
Results from numerous studies have established that the
electronically excited uranyl ion in aqueous media can ab-
stract a hydrogen atom from the C—H bonds of hydrocar-
bon materials, and, in some instances, products also arise
from C—C bond cleavage (1, 4, 6, 17). Our results are in
agreement with these features, as we observe tert-butyl alco-
hol as the major product arising from activation of the
weaker tertiary C—H bond (85 5%). Activation of the pri-
mary C—H bonds and C—C bonds also occurs, as evi-
denced by formation of isobutene (ca. 5%) along with much
lesser amounts of methane, propane, and radical recombina-
© 2003 NRC Canada

Bergfeldt et al.
225
Fig. 6. Comparison of the data of Moriyasu et al. ( ) and the
fitted values based on our model (᭝) for the inverse lifetime of
the electronically excited uranyl ion as a function of the concen-
tration of HClO4. Average root-mean-square difference between
experimental (Moriyasu et al. (13)) and calculated values is 6%.
pertain to the hydrogen-abstraction processes for the
quenching by isobutane of the acid and base entities, respec-
tively. Correspondingly, the first-order constants kOH1 and
kOH2 pertain to the formation of hydroxyl radical, which re-
acts rapidly under our conditions to yield additional amounts
of tert-butyl radical, eq. [3]. Consequently, two competitive
modes are envisioned in the formation of the tert-butyl radi-
cal. Nonradiative decays of the excited states are designated
by knr1 (acid form) and knr2 (base form) and those for emis-
sion by kem1 and kem2. The invariant shape of the emission
spectrum with changes in acid concentration, however, im-
plies that kem2 is of little consequence under our conditions,
and the outcomes of our modeling investigation are support-
ive of this.
The tert-butyl radical formed via the foregoing excited-
state events and via eq. [3] is proposed to react primarily —
and nearly quantitatively — with ground state [UO2(H2O)5]2+
to yield t-BuOH and, initially, uranoyl ion, which can then
undergo disproportionation (eq. [6]).
[8]
(CH3)3C· + [UO2(H2O)5]2+ + H2O
→ [UO2(aq)]+ + H+ + (CH3)3OH
The basis for this proposal rests, in part, on the fact that t-
BuOH is the major product (85 5%) and on the mass-
constants (lifetimes) occur where the entry into the excited-
state scheme (Fig. 1) takes place by a short pulse of light ab-
sorbed by [UO2(H2O)5]2+ to give the electronically excited
*[UO2(H2O)5]2+. The values of the rate constants given be-
low indicate that the shorter of the two time constants is less
than 100 ns, so it is not observed on the µs time scale either
by us or by Moriyasu et al. (13). The longer time constant
gives rise, on the µs scale, to mono-exponential decay in
emission and excited-state absorption under both our and
Moriyasu’s conditions, and the reciprocal lifetime (τ–1) that
equals the observed first-order rate constant in the absence
of isobutane is given by
balance ratios given in the Results section, although the ac-
tual source of oxygen in t-BuOH is not known. Furthermore,
the products of bimolecular self-reactions of tert-butyl and
iso-butyl radicals such as isobutene (disproportionation),
2,2,3,3-tetramethylbutane, 2,2,4-trimethylpentane, and 2,5-
dimethylhexane (recombination), which are major products
in steady-state gamma radiolysis in the absence of uranyl
ion, are found to be minor products of photolysis in the pres-
ence of uranyl ion.
Both the tert-butyl radical and the α-ethyl alcohol radical
react readily with the mild oxidant [Fe(CN)6]3– (E° =
0.36 V), and the latter also reacts with [UO2(aq)]2+ (E° =
0.16 V) (7, 40–42). The electrochemistry of the tert-butyl
radical in acetonitrile also indicates it is a mild reducing
agent (43). These comparisons point towards a reaction of
[UO2(aq)]2+ with tert-butyl radical (eq. [8]). Attempts to
measure the rate constant for the reaction of [UO2(aq)]2+
with tert-butyl radical were undertaken using our pulse-
radiolysis conductivity and optical techniques at low dose
(0.5 µM radical concentration, pH 2.7), to try to minimize
the self-reaction of the tert-butyl radical. This pulse
radiolysis approach was only partially successful because
even at low dose, the bimolecular self-reactions of the tert-
butyl radical dominated. An estimate to the upper limit for
the rate constant in eq. [8] is 1 × 106 M–1 s–1. In the context
of our steady-state photolysis experiments where the radical
concentration will be very low (<< 10–8 M), the situation is
likely to be quite different. Even if the rate constant was
many powers of ten below 106 M–1 s–1, the rate of eq. [8]
would prevail over other processes, particularly second-order
events such as the self-reaction of tert-butyl radical, which
has, therefore, not been included in our treatment.
–1
[9]
τ
= 1/2{(k1 + k(2) + k21[H+])
– ((k1 – k(2) – k21[H+])2 + 4k12k21[H+])}1/2
where:
[10] k1 = knr1 + kem1 + k12 + kOH1
[11] k(2) = knr2 + kem2 + kOH2
[12] k2 = k(2) + k21[H+]
We have used the lifetimes measured by Moriyasu et al.
over the range of 0.01–5 M HClO4 (13) for 10 mM uranyl
perchlorate at 25°C. The software program SigmaPlot for
Windows 4.00, SPSS Inc. was then employed to fit the life-
time data, to obtain values of the parameters k1, k(2), k12, and
k21. The SigmaPlot program employs a convergence routine,
and as a confirmation of the SigmaPlot results, an Excel
spreadsheet (Microsoft Office 2000) was used to allow for
the input of a range of values for the individual rate con-
stants. The two approaches were in agreement. This proce-
dure has been used in the other cases of the fitting of data,
as described below. For the lifetimes, the resulting best fit of
the experimental points and the calculated curve is shown in
Fig. 6. The fit is very “rigid” in the sense that changes as
small as 10% to the individual rate constants give rise to
clearly discernable differences in lifetimes between the cal-
To fit our model and thus obtain values of the rate con-
stants, both time-resolved measurements of lifetimes and
quantum yields are required. The mathematical development
for the lifetimes is given elsewhere (32, 38), and two time
© 2003 NRC Canada

226
Can. J. Chem. Vol. 81, 2003
Table 1. Summary of values of the rate constants and efficiencies.
Reaction
ki
k value
900 s–1
Efficiencya (ηi)
0.00083
0.017
*[UO2(H2O)5]2+ → [UO2(H2O)5]2+ + hν′
*[UO2(H2O)5]2+ → [UO2(H2O)5]2+ + ∆
kem1
knr1
k12
1.8 × 104 s–1
1.1 × 106 s–1
2.2 × 107 M–1 s–1
0 s–1
3.5 × 106 s–1
1.0 × 107 M–1 s–1
1500 s–1
*[UO2(H2O)5]2+ → *[UO2(H2O)n(OH)]+ + H+
*[UO2(H2O)n(OH)]+ + H+ → *[UO2(H2O)5]2+
*[UO2(H2O)n(OH)]+ → [UO2(H2O)n(OH)]+ + hν′′
*[UO2(H2O)n(OH)]+ → [UO2(H2O)n(OH)]+ + ∆
*[UO2(H2O)5]2+ + BH → [UO2(H2O)5]+ + B· + H+
*[UO2(H2O)5]2+ + H2O → [UO2(H2O)5]+ + ·OH + H+
*[UO2(H2O)n(OH)]+ + BH → [UO2(H2O)5]+ + B·
*[UO2(H2O)n(OH)]+ + H2O → [UO2(H2O)5]+ + ·OH
0.97
0.39b
k21
kem2
knr2
kq1
0
0.61b
0.0088c
0.0014
0.0033b,c
0.00044c
kOH1
kq2
2.0 × 107 M–1 s–1
2500 s–1
kOH2
aEfficiency is defined as η1i = k1i /∑ k1i and η2i = k2i /∑ k2i, where subscripts 1 and 2 refer to acid and base forms, respectively (e.g., the efficiency of
the dissociation of the acid form to the base form is given by η12; the efficiency of the reverse reaction is given by η21).
b[HClO4] = 0.1 M.
c[isobutane] = [BH] = 0.9 mM.
culated and experimental values. The “best” fit has been de-
termined from the minimization of the value of the root-
mean-square difference between the calculated points and
experimental ones; the average difference is small ( 6%).
Furthermore, the differences show no readily apparent sys-
tematic trend over the 500-fold change in perchloric acid
concentration. Thus the fit is considered to be satisfactory,
and the values of the calculated parameters are: k1 = 1.1 ×
cohol via the hydrogen abstraction and the formation of ·OH
from the acid excited-state species. Correspondingly, kRBx
equals kq2[BH] + kOH2 for the base where [BH] is the con-
centration of isobutane (0.91 mM). Thus, the ratios kRAx/k1
and kRBx/k2 pertain to the total efficiencies for the chemical
reactions from *[UO2(H2O)5]2+ and *[UO2(H2O)n(OH)]+, re-
spectively.
Values for kq1, kq2, kOH1, and kOH2 were obtained by “best
fitting” the calculated quantum yields to the experimental
ones. The fits were judged on the basis of the minimization
of root-mean-square differences, and they are very sensitive
106 s–1, k(2) = 3.5 × 106 s–1, k12 = 1.1 × 106 s–1, and k21
=
2.2 × 107 M–1 s–1. The ratio of k12/k21 leads to a pKa of 1.3.
A range of values from 1.7 to >3.6 have been reported by
others (1, 32). However, some caution should be exercised
in comparing the different pKa values. Different methods
have been used to obtain the values and, frequently, for sys-
tems having different counter anions, so ion pairing may
play a subtle but notable role. From the viewpoint of the de-
pendence of the behaviors of the excited states on the acid
concentration, the pKa value is not the critical facet; rather,
it is the values of k12 and k21 in relation to the rate constants
of the competing processes that determines the course of
events. The rate constants given above, in conjunction with
the quantum yields for emission and for formation of the
tert-butyl alcohol, now provide the basis to evaluate the
chemical and nonchemical rate constants for the reactions in
competition with the acid and base processes.
to changes in the inputted values of the kqs and kOH
s
( 10%). The comparison between the calculated and experi-
mental values for the best fit is shown in Fig. 5 with the two
sets agreeing very well within experimental error ( 8% aver-
age root-mean-square difference), and the deviations appear
to be random. The error bars shown for the experimental
points (filled circles) are the estimated standard deviations (1
σ), and the calculated points are shown with a deviation of
10%. (The same procedure has been used for the emission
results presented in Figs. 3 and 4). The rate constants ob-
tained from the fit in Fig. 5 are kq1 = 1.0 × 107 s–1, kq2
=
2.0 × 107 s–1, kOH1 = 1500 s–1, and kOH2 = 2500 s–1. Gaziev
et al. (10) provided an estimate of kOH from the photo-
stimulated 18O exchange of 4.4 × 104 s–1, and that obtained
from time-resolved lifetime measurements by Waltz et al.
(20) is 5.4 × 104 s–1; the latter studies also yielded a value of
3.3 × 107 M–1 s–1 for quenching by isobutane. Our values of
kOH1 and kOH2 are lower by a factor of about ten, whereas the
values of kq1 and kq2 are of the same magnitude as that re-
port previously (20). The comparison of our present results
and those reported previously is perhaps somewhat tenuous
because the previous studies were modeled on the presence
of only one excited-state species (i.e., the acid form).
In the absence of isobutane, our model gives rise (Appen-
dix 2) to the following expression for the quantum yield of
emission, where the efficiencies are,
The expression for the total quantum yield for the produc-
tion of t-BuOH alcohol arising from the hydrogen-
abstraction reactions (kq1 and kq2) and from generation of the
·OH radical (kOH1 and kOH2), followed by eqs. [3] and [8], is
ΒΗ ΒΗ
[13] ΦTotal = {1/(1 – η η21 )}(kRAx/k1 + kRBx/k2)
12
The derivation is given in Appendix 1, and it is based on
the
assumption
that
multiple
passes
between
*[UO2(H2O)5]2+ and *[UO2(H2O)n(OH)]+ via eqs. [4] and
[5] occur in the decay of these excited-state species. The
ΒΗ
ΒΗ
symbols η and η designate the efficiencies of loss of a
12
21
proton from the acid *[UO2(H2O)5]2+ to form the base
*[UO2(H2O)n(OH)]+ (eq. [4]) and the re-protonation of the
latter (eq. [5]) in the presence of saturated isobutane solu-
tions, respectively. The value, kRAx, represents the sum,
kq1[BH] + kOH1, that leads to the formation of tert-butyl al-
Total
°
° °
[14]
φ
= (ηem1 + ηem2η12){1/(1 – η12η21)}
em
as previously, symbolized by ηs. The comparison between
the experimental results and the calculated ones is shown in
© 2003 NRC Canada

Bergfeldt et al.
227
Fig. 3. The average root-mean-square difference for all
points is 8% with no apparent systematic deviations. Al-
though we initially assumed that both excited-state species
could emit, the best fit as shown here was obtained for a
For a saturated isobutane solution, the percent quenching is,
in total, about 7%. Because isobutane is a condensable gas
at room temperature, it was not feasible to increase the pres-
sure and thus its concentration. However, with less
condensable, low molar mass alkanes such as methane,
which is known to quench excited uranyl (24), the solubility
factor should prove less restrictive through the use of in-
creased pressure and lower temperatures. As a further ap-
proach to increasing the apparent quantum yields, the free
radicals generated can be used to initiate chain processes. In
this context, we report elsewhere (21, 32) that the addition
of millimolar amounts of potassium peroxydisulphate gives
a 75-fold increase in the apparent quantum yield for tert-
butyl alcohol formation (Φobserved ~ 1.5) with the additional
advantage that there is no net consumption of [UO2(aq)]2+
(i.e., the [U(aq)]4+ product is oxidized back to [UO2]2+ dur-
ing photolysis). The use of visible light coupled with the
broad absorption spectrum of uranyl ion provides a strategy
to explore further the conversion of alkanes, in particular
ones with low molecular mass where product selectivity is
of lesser concern, into useful oxygenated products.
kem1 of 900 s–1 and for kem2 being zero, i.e., kem2 << kem1
.
This situation implies that the emission arises predominantly
from *[UO2(H2O)5]2+, and this is consistent with the emis-
sion spectral features remaining stationary as the perchloric
acid concentration was varied, as described in the Results
section. Using the aforementioned rate constants derived
from the model, we extended the fitting procedure to the sit-
uation where isobutane was now present. The expression for
Total
Φ
is given in Appendix 2, and it is analogous to that of
em
eq. [14]. The results are shown in Fig. 4. The average root-
mean-square difference between the calculated and experi-
mental yields is somewhat larger, at 21%, than that in the
absence of isobutane ( 8%) with deviations being more ap-
parent at low concentrations of perchloric acid. This less sat-
isfactory situation seems to derive, in part, from the low
signal levels for integrated emission encountered at low acid
concentrations (coupled perhaps with scattered light) and
from the further lowering of emission in the presence of iso-
butane relative to that in its absence.
Acknowledgments
The information given in Table 1 summarizes the values
of the rate constants calculated from the excited-state model.
To provide perspective on the relative influence of the differ-
We thank the Natural Sciences and Engineering Research
Council of Canada (NSERC) for financial support, Dr. D.
Garratt of the Cameco Corporation for uranium trioxide, and
Mr. Ken Thoms for GC–MS and helpful suggestions.
ent excited-state processes, the efficiencies (η) of the indi-
i
vidual steps are presented based upon calculations for the
condition of 0.1 M HClO4 and 0.9 mM isobutane. For the
acid form *[UO2(H2O)5]2+, the dominant event for all of our
conditions is the deprotonation reaction (eq. [4]) with a η12
of ca. 0.97. Furthermore, the efficiency of hydrogen-
abstraction reactions is greater for the acid form than for the
base species, whereas the reverse prevails in the generation
of hydroxyl radical. The efficiency of protonation of
*[UO2(H2O)n(OH)]+ is 0.39 at 0.1 M HClO4; however, this
value rapidly increases on going to higher acid concentra-
tions. Consequently, the calculated percentage of tert-butyl
alcohol, which has its origins in the hydrogen-abstraction
and ·OH-formation reactions of the acid form
*[UO2(H2O)5]2+, accounts for greater than 90% (0.8–4 M
HClO4) of the product formation. Even at our lowest
perchoric acid concentration, the acid form is the dominant
reacting species (64%).
The mechanism that is provided here, involving the acid–
base processes of the electronically excited uranyl ion, pro-
vides a comprehensive framework in which to interpret the
experimental results. For over a 400-fold change in
perchloric acid concentration (0.01–4 M HClO4), this model
predicts well the lifetimes of the excited uranyl ion, the
quantum yields of emission in both the absence and pres-
ence of a quencher (i.e., isobutane), as well as the quantum
yields for the production of tert-butyl alcohol.
References
1. C.P. Baird and T.J. Kemp. Prog. React. Kinet. 22, 87 (1997).
2. M.D. Marcantonatos, M. Deschaux, F. Celardin, and M.
Levental. Chem. Phys. Lett. 65, 316 (1979).
3. V. Balzani, F. Bolletta, M.T. Gandolfi, and M. Maestri. Top.
Curr. Chem. 75, 1 (1978).
4. H.D. Burrows and T.J. Kemp. J. Chem. Soc. Rev. 3, 139
(1974).
5. E. Rabinowitch and R.L. Belford. Spectroscopy and
photochemistry of uranyl compounds. Pergamon Press, Ox-
ford. 1964.
6. D.M. Roundhill. Photochemistry and photophysics of metal
complexes. Plenum Press, New York. 1994. Chap 8.
7. M.E.D.G. Azenha, H.D. Burrows, S.J. Formosinho, and M. da
G.M. Miguel. J. Chem. Soc. Faraday Trans. 85, 2625 (1989).
8. H.D. Burrows and S.J. Formosinho. J. Chem. Soc. Faraday
Trans. 2, 73, 201 (1977).
9. M. Deschaux and M.D. Marcantonatos. Chem. Phys. Lett. 63,
283 (1979).
10. S.A. Gaziev, N.G. Gorshkov, L.G. Mashirov, and D.N.
Suglobov. Inorg. Chim. Acta, 139, 345 (1987).
11. R.J. Hill, T.J. Kemp, D.M. Allen, and A.J. Cox. J. Chem. Soc.
Faraday Trans 1, 70, 847 (1974).
12. M.D. Marcantonatos. J. Chem. Soc. Faraday Trans. 1, 76,
1093 (1980).
This study clearly brings into focus the fact that organic
products can arise from both hydrogen-abstraction processes
and the competitive occurrence of the oxidation of coordi-
nated water to generate hydroxyl radical, which is, in gen-
eral, highly reactive towards a large variety of organic
substances (19). The apparent quantum yields for organic
products derived from alkanes reported here and elsewhere
(17) tend to be low, partly owing to solubility limitations.
13. M. Moriyasu, Y. Yokoyama, and S. Ikeda. J. Inorg. Nucl.
Chem. 39, 2211 (1977).
14. M. Moriyasu, Y. Yokoyama, and S. Ikeda. J. Inorg. Nucl.
Chem. 39, 2199 (1977).
15. Y.-Y. Park and H. Tomiyasu. J. Photochem. Photobiol. A, 64,
25 (1992).
16. Y.-Y. Park, Y. Sakai, R. Abe, T. Ishii, M. Harada, T. Kojima,
and H. Tomiyasu. J. Chem. Soc. Faraday Trans. 86, 55 (1990).
© 2003 NRC Canada

228
Can. J. Chem. Vol. 81, 2003
17. Y. Mao and A. Bakac. J. Phys. Chem. 100, 4219 (1996).
18. C. Sonntag and H.-P. Schuchmann. Angew. Chem. Int. Ed.
Engl. 30, 1229 (1991).
η
′
q1
[A2] φARx
=
=
BH BH
1 − η
η
12 21
19. G.V. Buxton, C.L. Greenstock, W.P. Helman, and A.B. Ross. J.
where
Phys. Chem. Ref. Data, 17, 513 (1988).
20. W.L. Waltz, J. Lilie, X. Xu, P. Sedlák, and H. Möckel. Inorg.
Chim. Acta, 285, 322 (1999).
21. X. Xu. Ph.D. Dissertation. Department of Chemistry, Univer-
sity of Saskatchewan, Saskatoon, Sask., Canada. 1997.
22. M.E.D.G. Azenha, H.D. Burrows, S.J. Formosinho, M. da
G.M. Miguel, A.P. Darmanyan, and I.V. Khudyakov. J. Lumin.
48/49, 522 (1991).
kq1[BH]
[A3]
[A4]
[A5]
η
η
′
q1
k1 + kq1[BH]
k12
BH
=
12
k1 + kq1[BH]
k21
23. A.P. Darmanyan and I.V. Khudyakov. Photochem. Photobiol.
BH
η
=
21
52, 293 (1990).
k2 + kq2[BH]
24. G.V. Nizova and G.B. Shul’pin. Zh. Obshch. Khim. 60, 2124
(1990). English Translation: J. Gen. Chem. 60, 1895 (1991).
25. J.G. Calvert and J.N. Pitts. Photochemistry. Wiley, New York.
1966. pp. 783–786.
26. D.B. Wetaufr. J. Am. Chem. Soc. 86, 508 (1964).
27. J.T. Bell and R.E. Biggers. J. Mol. Spectrosc. 22, 262 (1967).
28. A. Ekstrom. Inorg. Chem. 13, 2237 (1974).
29. J.T. Bell, H.A. Friedman, and M.R. Billings. J. Inorg. Nucl.
Chem. 36, 2563 (1974).
Accounting for hydrogen abstraction of isobutane by
hydroxyl radical produced by species A
ηOH1
1 − η
[A6] φOH1
=
BH BH
η
12 21
where
kOH1
30. Z.F. Zhao and D.A. Liu (Editors). Analytical chemistry. Peo-
[A7] ηOH1
=
k1 − kq1[BH]
ple’s Education, Beijing. 1978. p. 295.
31. J. H. Betts. Can. J. Chem. 33, 1775 (1955).
32. T.M. Bergfeldt. Ph.D. Dissertation. Department of Chemistry,
University of Saskatchewan, Saskatoon, Sask., Canada, 2001.
33. Y. Katsumura, H. Abe, T. Yotsuyanagi, and K. Ishingure. J.
Photochem. Photobiol. A, 50, 183 (1989).
34. T.W. Newton and F.B. Baker. Inorg. Chem. 4, 1166 (1965).
35. H.J. Lucas and W.F. Eberz. J. Am. Chem. Soc. 56, 460 (1934).
36. M.D. Marcantonatos. J. Chem. Soc. Faraday Trans. 1, 75,
2273 (1979).
Similarly, the quantum yield for the reaction from species
B, along with hydrogen abstraction from isobutane by
hydroxyl radical produced by species B are given by
BH
η
η′
η
12 q2
[A8] φBRx
=
BH BH
1 − η
12 21
BH
η
ηOH2
η
12
[A9] φOH2
=
37. M.D. Marcantonatos and M.M. Pawlowska. J. Chem. Soc. Far-
BH BH
1 − η
12 21
aday Trans. 1, 85, 2481 (1989).
38. S.J. Formosinho and M. da G.M. Miguel. J. Chem. Soc. Fara-
day Trans. 1, 80, 1717 (1984).
39. M. da G. M. Miguel, S.J. Formosinho, and A.C. Cardoso. J.
The total quantum yield is
[A10] φTotal = φARx + φOH1 + φRBx + φOH2
Chem. Soc. Faraday Trans. 1, 80, 1735 (1984).
or
40. S. Steenken and P. Neta. J. Am. Chem. Soc. 104, 1244 (1982).
41. E. Hensler and W.J. Lorenz. In Standard potentials in aqueous
solutions. Edited by A.J. Bard, R. Parsons, and J. Jordon. Mar-
cel Dekker, New York. 1985. Chap 14.
42. L. Martinot and J. Fuger. In Standard Potentials in Aqueous
Solutions. Edited by A.J. Bard, R. Parsons, and J. Jordan. Mar-
cel Dekker, New York. 1985. Chap 21.
BH
η′ + ηOH1 + η (η′ + ηOH2
)
q1
12
q2
[A11] φTotal
=
BH BH
1 − η
η
12 21
Using the rate constants from the fitting of Moriyasu et
BH
al.’s data (A1), η = 0.973. Let ηQ represent the total effi-
12
43. D.D.M. Wayner, D.J. McPhee, and D. Griller. J. Am. Chem.
Soc. 110, 132 (1988).
ciency of quenching of both species by isobutane; then
BH
[A12] ηQ = ηq′1 + η ηq′2 ≅ ηq′1 + ηq′2
12
and
Appendix 1: Derivation of the expression
for the total chemical quantum yield by
species A (acid form) and B (base form)
BH
[A13] ηOH = ηOH1 + η ηOH2 ≅ ηOH1 + ηOH2
12
Equation [A1.11] can be written as
The quantum yield for the reaction of the A species is
given by
A



 
 

1
kRx kB
Rx 
[A14] φTotal
=
+
1 − ηBHη21
k1
k2
BH  

12
[A1] φRAx = η + η ηBHηBH + ηq1 (ηBHηB21H)2 +...
′
q1
′
′
q1 12 21
12
where
For multiple passes through the excited states via
protonation and deprotonation and for ηs less than one, this
series converges to the limit of
kRAx
′
kq1 + kOH1
[A15]
=
= η + ηOH1
′
q1
k1
k1
© 2003 NRC Canada

Bergfeldt et al.
229
k12
k1
and
°
[B6] η12
=
=
kRBx
′
kq2 + kOH2
[A16]
=
= η + ηOH2
′
q2
k21[H+ ]
k2
k2
°
[B7] η21
k2
where
°
°
For η12η21 < 1, eq. [A2.5] converges to
[A17] k1 = knr1 + kem1 + k12 + kOH1
ηem1
[B8] φem1
=
[A18] k2 = knr2 + kem2 + kOH2 + k21[H+]
1 − η1°2η2°1
Similarly, the quantum yield of emission of the B species
in the absence of isobutane can be shown to be
References
A1. M. Moriyasu, Y. Yokoyama, and S. Ikeda. J. Inorg. Nucl.
°
ηem2η12
Chem. 39, 2211 (1977).
[B9] φem2
=
°
°
1 − η12η21
Appendix 2. Derivation of expression for
emission quantum yields
The total emission quantum yield from both species in the
absence of isobutane is
The efficiencies of emission by the acid (A) species (ηem1
and the base (B) species (ηem2) can be expressed as:
)




1
Total



°
= (ηem1 + ηem2η12)
[B10] φ
em
°
°
21 
1 − η η
12
kem1
[B1] ηem1
=
k1
In the presence of isobutane, the total emission quantum
yield from both species is
kem2
k2
[B2] ηem2
=





1
Total
BH
[B11] φ
= (ηem1 + ηem2
η
)
12
em
1 − ηBHη21
BH 
12
where
where
[B3] k1 = knr1 + kem1 + k12 + kOH1
k12
BH
[B12] η
=
[B4] k2 = knr2 + kem2 + kOH2 + k21[H+]
12
k1 + kq1[BH]
The quantum yield of emission by the A species in the ab-
sence of isobutane can be expressed as
k21[H+]
BH
[B13] η
=
21
k2 + kq2[BH]
2
°
°
° °
[B5] φem1 = ηem1 + η12η21ηem1 + (η12η21) ηem1 +...
and kq1 and kq2 are the quenching constants of the A species
where
and the B species by isobutane, respectively.
© 2003 NRC Canada