MARY spectroscopy in the presence of coordination compound Zn(hfac) 2(PPO)2

Article abstract of DOI:10.1016/j.cplett.2011.01.056

MARY spectroscopy is finding increasing use in the studies of transient organic radical ions and their reactions. Extending this technique to organometallic species will broaden the class of potential target compounds and can help answer important mechanistic questions in organometallic and spin chemistry. We probed this approach using a tailored Zn(hfac)₂(PPO) ₂ complex. The synthesized complex has quantum yield and fluorescence lifetime (n-decane solution) Φ ～0.8 and τ ～1.3 ns, respectively. For this type of complex it is the first observation of MARY spectra different from those of free ligand, thus implying participation of the complex in the development of the observed signal.

Full text of DOI:10.1016/j.cplett.2011.01.056

Chemical Physics Letters 504 (2011) 107–112
Contents lists available at ScienceDirect
Chemical Physics Letters
journal homepage: www.elsevier.com/locate/cplett
MARY spectroscopy in the presence of coordination compound Zn(hfac)2(PPO)2
b
b
N.V. Sergey a, , A.B. Burdukov , N.V. Pervukhina , L.V. Kuibida a,c, I.P. Pozdnyakov a,c, D.V. Stass a,c
⇑
a Institute of Chemical Kinetics and Combustion SB RAS, 3, Institutskaya Str., 630090 Novosibirsk, Russia
b Nikolaev Institute of Inorganic Chemistry, 3, Acad. Lavrentiev Ave., 630090 Novosibirsk, Russia
c Novosibirsk State University, 2, Pirogova Str., 630090 Novosibirsk, Russia
a r t i c l e i n f o
a b s t r a c t
Article history:
MARY spectroscopy is finding increasing use in the studies of transient organic radical ions and their
reactions. Extending this technique to organometallic species will broaden the class of potential target
compounds and can help answer important mechanistic questions in organometallic and spin chemistry.
We probed this approach using a tailored Zn(hfac)2(PPO)2 complex. The synthesized complex has quan-
Received 1 November 2010
In final form 20 January 2011
Available online 23 January 2011
tum yield and fluorescence lifetime (n-decane solution)
u
ꢀ0.8 and
s
ꢀ1.3 ns, respectively. For this type
of complex it is the first observation of MARY spectra different from those of free ligand, thus implying
participation of the complex in the development of the observed signal.
Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction
can be a particular advantage over the more general techniques
like optical absorption spectroscopy.
A clear trend in chemistry today is targeted synthesis of new
compounds, which requires an advanced level of understanding
and control over reaction mechanisms. However, often the critical
stage of a chemical transformation involves a known or postulated
short-lived intermediate that actually determines the chemistry,
but cannot be readily isolated and conveniently studied. Important
clues on the mechanism can then be deduced from indirect
physicochemical experiments that probe the intermediate ‘on the
fly’ imprinting its properties on some observable data such as
polarization pattern in an NMR spectrum, or the magnetic field
dependence of fluorescence intensity [1]. Probing short-lived para-
magnetic intermediates, such as radical and radical ion pairs, with
static and oscillating magnetic fields has provided a wealth of
important data on the reaction mechanisms for organic chemistry
[2]. Short-lived transient species in inorganic chemistry is a less
commonly discussed subject, although they undoubtedly play
roles similar to their organic counterparts, and redox and bond
cleavage processes can also proceed via radical – or radical-ion –
type species derived from inorganic complexes or clusters. As these
transient species are very reactive and thus have rather short life-
times, they cannot be isolated and studied by conventional meth-
ods, such as XRD, suitable for stable compounds. New methods are
needed to study short-lived intermediates in inorganic processes
and reaction mechanisms that they mediate. Furthermore, the sen-
sitivity of such a method to certain specific property of the as-
sumed intermediate, such as the presence of spin and/or charge,
In this work we explore the possibility of adapting one of the re-
cently developed spin-sensitive techniques, MARY (level-crossing)
spectroscopy of radical ion pairs in non-polar solutions, to the
study of transient organometallic species. The method was formu-
lated for organic radical ions, and has proved useful in detecting
several important classes of hitherto unseen radical ions with life-
times of just several nanoseconds, such as primary solvent radical
cations (holes) in irradiated alkanes [3] and radical ions of several
fluorinated [4] and methylated [5] benzenes. The method is also
useful for studies of reactions involving short-lived radical ions,
such as the reactions of ion-molecular charge transfer [6,7], dimer-
ization and higher aggregation [8], and monomolecular chemical
decay [9]. Being useful for short-lived organic radical ions, the
method looks promising for the study of short-lived transient
states of metal-centered complexes, provided these can play the
role of radical ions of the original radical ion pairs. In case of suc-
cess, a new highly specialized but also highly sensitive method
would become available to study the mechanisms and intermedi-
ates, e.g., unusual transient valent states, for inorganic chemistry.
On the other hand, a whole new class of systems will be also
opened for magnetic field effect studies, including but not limited
to MARY spectroscopy. Radical-ionic states of complexes based on
central metal ions are very interesting in this respect since the me-
tal ion can have nuclear spin other than ½ or 1, to which nuclei in
organic radicals are usually limited, and rather high hyperfine cou-
plings to these heavy nuclei. Furthermore, complex compounds
can be very useful for magnetic and spin effect studies as the cen-
tral ion can be readily changed without changing the ligand envi-
ronment, and thus a series of similar systems can be generated
that can have widely varying magnetic properties. In the next
⇑
Corresponding author. Fax: +7 383 3331561.
E-mail address: arabadzhi@kinetics.nsc.ru (N.V. Sergey).
0009-2614/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2011.01.056

108
N.V. Sergey et al. / Chemical Physics Letters 504 (2011) 107–112
sections we briefly summarize the method of MARY spectroscopy
to outline its requirements, discuss the choice of a suitable candi-
date complex, and then describe our results for the first success-
fully created system of this type.
was filtered again, hexane (10 mL) was added, and the mixture
was left at room temperature in a semi-closed flask for slow
crystallization. In four days large clear tabular crystals precipitated,
which were collected on a glass filter, washed with hexane and
dried in air, affording the title complex in 33% yield. The compound
was characterized with X-ray structural analysis, unit cell parame-
1.1. MARY spectroscopy and the choice of complex
ters are: a = 9.5873(4) Å, b = 10.2574(5) Å, c = 11.0795(4) Å,
a =
86.93(0)°, b = 65.48(0)°,
c
= 72.22(0), V = 940.56(58) Å3, Z = 1, space
The implementation of MARY spectroscopy that is used in this
work is a steady-state method based on the continuous cycle of
generating spin-correlated radical ion pairs by X-irradiation of
non-polar solution of suitable charge acceptors, spin evolution of
these pairs in applied static magnetic field, and singlet state
recombination of the pairs producing electronically excited mole-
cules that are observed by fluorescence, provided one of the chosen
acceptors is also a luminophore [5]. Sweeping of the applied mag-
netic field in the region of weak fields produces MARY spectra – the
dependencies of fluorescence intensity on magnetic field, with
characteristic features reflecting the magnetic constitution of the
partners of the recombining radical ion pair.
group P(ꢂ1). Elemental analysis: calculated for C40H32F12N2O6Zn: C
52.1; H 2.0; F 24.8; N 3.0, found: C 52.4; H 2.7; F 24.9; N 3.1.
X-ray quality single crystals were obtained from the synthesis.
Data collection was carried out on a Bruker X8 Apex CCD diffrac-
tometer at 150 K using graphite-monochromated Mo–Ka radiation
(k = 0.71073 Å) [12]. Reflection intensities were integrated using
SAINT software and corrected for absorption by semi-empirical
method with SADABS [13]. The structure was solved by the direct
method and refined by full-matrix least squares against F2 in aniso-
tropic approximation for non-hydrogen atoms using SHELX pro-
gram package [14]. Hydrogen atoms were set in geometrical
positions. Final agreement factors were R1 = 0.0431, wR2 = 0.1121
Here we aim to check whether organometallic complexes can
substitute for the charge acceptors/luminophores in the described
setting, and whether the transient radical-ionic states of such com-
plexes can thus be detected. To this end a suitable model system is
needed that has to meet a number of requirements. A non-polar
solvent is a critical requirement for the method, and to avoid back-
ground processes an alkane solvent, such as n-decane, is normally
chosen, so the complex has to be soluble in alkanes. This limits us
to non-ionic complexes, with the ligands compensating the charge
of the central metal ion. We chose acetylacetonate-type bidentate
charge-compensating ligands, using hexafluoroacetylacetonate
(hfac) for improved solubility in alkanes provided by fluorine
substitution. As the detectable signal in the experiment is fluores-
cence, we further need a luminophore in the radical ionic system,
which is preferably the same complex as the charge acceptor, so a
known luminophore 2,5-diphenyloxazole (PPO) was chosen as the
additional ligand in the complex.
for 5829 reflections (I > 4rI), GOF = 1.058. The structure has been
deposited to the Cambridge Structural Database (CCDC 798462)
and is available free of charge at http://www.ccdc.cam.ac.uk/prod-
ucts/csd/request/.
Solution IR spectra of Zn(hfac)2(PPO)2 and PPO in n-decane
were taken on a Bruker Vector 22 FT-IR spectrometer in a
0.15 mm NaCl cuvette at concentrations 10ꢂ4–10ꢂ3 M.
Optical absorption spectra were taken on a Shimadzu UV-2401
spectrophotometer. Fluorescence spectra and kinetics of fluores-
cence were obtained using an FLS920P system (Edinburgh Instru-
ments). Fluorescence measurements were taken for solutions of
Zn(hfac)2(PPO)2 and PPO in n-decane with excitation at 300 nm
with a Xe900 xenon lamp (450 W), optical density of solutions
0.1, slit width 1 nm for both excitation and detection beams
(Czerny-Terner monochromators, 1.8 nm/mm, F/4.2, 1800 lines
grating). Fluorescence kinetics were registered at 357 nm, with
pump pulse (pulse diode, 100 ps) at 320 nm.
Regarding the central ion, for the first model system we wanted
the magnetically simplest possible case of diamagnetic initial com-
plex with the central ion having zero nuclear spin, and chose zinc
as the central ion in the complex (natural abundance of 67Zn with
spin 5/2 is only 4%). Two hfac ligands compensate the charge of
Zn2+ and use four of the six available coordination sites of the
Zn2+ ion, with the remaining two being occupied by the nitrogen
atoms of the two PPO ligands to produce the complex Zn(hfac)2(P-
PO)2. To take part in the cycle of processes leading to formation of
the observed signal, the candidate complex should be able to do-
nate or accept an electron (thus generating the ‘radical ion’). The
ionization potential (IP) of Zn(hfac)2 is 10.0–10.1 eV [10], IP for
the Zn(hfac)2(PPO)2 complex is not known, but should be lower
due to the presence of two PPO ligands, and n-decane used as
the solvent has IP 9.5–9.7 eV [10], so positive charge transfer yield-
ing ‘radical cation’ of the complex may be possible. The specific
data on the electron affinity of Zn(hfac)2 and Zn(hfac)2(PPO)2 are
not available, but it is known that metals have large electron affin-
ity, so we can expect that our complex will act as electron acceptor.
The experimental setup for MARY spectroscopy was described
in [5]. The samples containing about 1 ml of solution in a quartz
cuvette are degassed by repeated freeze–pump–thaw cycles and
placed in the magnetic field of a Bruker ER-200D CW ESR spec-
trometer equipped with an X-ray tube for sample irradiation
(Mo, 40 kV ꢃ 20 mA), a pair of coils with a separate current source
to provide the constant ‘negative’ shift of the field required to
sweep through the zero of the field, and a PMT (FEU-130) for fluo-
rescence detection. The scanned magnetic field was modulated at a
frequency of 12.5 kHz with an amplitude up to 1 mT. A Stanford
SR-810 Lock-In Amplifier and computer averaging over 20 scans
of 512 points and 200 s sweep time each were used to obtain the
shown spectra, obtained as the first derivatives of the actual field
dependencies. No microwave pumping was applied to the samples.
All experiments were carried out at room temperature. The puri-
fied solvent – n-decane – was treated with sulphuric acid, potas-
sium permanganate, dried over calcium chloride and passed
through a column with a mixture of activated MgO and Al2O3
(courtesy of Mrs. N. Ivanova).
2. Experimental
3. Results and discussions
2.1. Synthesis of Zn(hfac)2(PPO)2
A novel complex of Zn(II) hexafluoroacetylacetonate with two
molecules of 2,5-diphenyloxazole (PPO), Zn(hfac)2(PPO)2, was syn-
thesized and characterized. The solubility of the complex in al-
kanes is about 1 mmol/L and is sufficient for MARY experiments.
The PPO component in the complex acts as the luminescent
fragment and was introduced in a Zn(hfac)2ꢁ2H2O complex via
Zn(hfac)2 H2O was prepared as described in [11]. 2,5-Diphenyl-
oxazole (46 mg, 0.2 mmol) was dissolved in methylene chloride
(1 mL). Zn(hfac)2ꢁ2H2O (50 mg, 0.1 mmol) was dissolved in methy-
lene chloride (10 mL) in a separate flask, the solution was filtered
and added to the solution of 2,5-diphenyloxazole. The mixture

N.V. Sergey et al. / Chemical Physics Letters 504 (2011) 107–112
109
Figure 1. Synthesis of Zn(hfac)2(PPO)2 from zinc(II) hexafluoroacetylacetonate and 2,5-diphenyloxazole.
exchange reaction between H2O and PPO as shown in the scheme
of synthesis in Figure 1. The molecular structure of the complex
was determined by X-ray crystallography. The structural features
of the Zn(hfac)2 and PPO moieties are well known and their discus-
sion is beyond the scope of this study. Details of syntheses and
characterization can be found in the Section 2.
To check in which form the complex Zn(hfac)2(PPO)2 is present
in alkane solutions, IR spectra of Zn(hfac)2(PPO)2 and PPO solutions
in n-decane were taken, with magnified most important region
shown in Figure 2. The spectra were compared, all lines of PPO
vibrations were found in the spectra for Zn(hfac)2(PPO)2 (marked
with asterisk in Figure 2), and the known lines of Zn(hfac)2 vibra-
tions [15] were also found in the spectra of Zn(hfac)2(PPO)2
(marked with ‘+’ in Figure 2). The two yet unidentified bands are
marked with ellipses in Figure 2.
To understand the origin of the new lines for the complex we
compared IR spectra of PPO and Zn(hfac)2(PPO)2 in KBr pellets
(not shown here) and in n-decane solution. The solid state spectra
for both the free ligand and the complex show the split lines
around 950 cmꢂ1 and 770 cmꢂ1, for the free ligand in solution
the split lines merge into single lines, but for the complex the lines
are again split as shown in ellipses in Figure 2 – two resolved lines
are present in the region of 950 cmꢂ1 and a broadened poorly re-
solved line is found at 770 cmꢂ1. The complex in solution thus
shows characteristic features found for its ligand in a pellet, which
is explained by coordination of the ligand to the central Zn ion in
Zn(hfac)2(PPO)2, leading to collective vibrations similar to those
found in the solid state. The conclusion is that in alkane solution
of Zn(hfac)2(PPO)2 at least a substantial fraction of PPO has local
environment similar to its environment in the solid pellet, i.e.,
PPO is indeed part of the complex and is not lost into the bulk.
We thus have a complex that is soluble in alkane and has a fluo-
rescing ligand in it.
Figure 3a shows optical absorption and fluorescence spectra of
the complex and the free ligand in alkane solution. The absorption
spectra of Zn(hfac)2(PPO)2 and PPO are similar, but the spectrum of
Zn(hfac)2(PPO)2 has a more pronounced band at 300 nm, and the
extinction coefficient of Zn(hfac)2(PPO)2 at the absorption maxi-
mum (6.9 ꢃ 104 Mꢂ1cmꢂ1) is noticeably higher then the doubled
extinction coefficient of the free ligand PPO (2.6 ꢃ 104 Mꢂ1cmꢂ1).
Fluorescence spectra of Zn(hfac)2(PPO)2 and PPO are similar for
both compounds. As the quantum yield of PPO is known to be
one (in cyclohexane) [16], we used it as internal reference to esti-
mate the quantum yield of Zn(hfac)2(PPO)2. The quantum yield of
fluorescence with excitation at 300 nm for Zn(hfac)2(PPO)2 is about
0.8 in n-decane relative to PPO with equal amounts of absorbed
excitation photons.
To determine the fluorescence lifetime of the complex, kinetics
of fluorescence were measured with the pump pulse at 320 nm and
detection at 357 nm. The fluorescence kinetics of the complex is
shown in Figure 3b and is practically monoexponential, the fluo-
rescence parameters of PPO are known and were taken from
[16]. Fluorescence lifetimes for Zn(hfac)2(PPO)2 and PPO were
found to be identical within experimental accuracy and equal to
1.3 ns. Although MARY spectroscopy is a steady-state technique
and thus is not critical to fluorescence lifetimes, this time is short
enough to allow more sophisticated time-resolved magnetic field
experiments with this system.Finally, MARY spectra were obtained
for different concentrations of Zn(hfac)2(PPO)2 solution in n-dec-
ane and were compared with the spectra for the free ligand PPO
substituting for the complex. A selection of MARY spectra is shown
in Figure 4, where to aid in comparing similar looking curves the
spectra are grouped into separate panels in pairs having either
equal concentrations of the complex and the ligand (three left
panels), or concentrations of free ligand equal to twice the concen-
tration of the complex, i.e., for equal concentrations per PPO
molecule (three right panels). The spectra are taken and shown
as first derivatives with respect to magnetic field, and their charac-
*
*
*
*
A
+
teristic features are the inflection point B (the maximum of the
r
derivative), which is determined by effective hyperfine couplings
B
in the partners of the recombining radical ion pair A1;2
¼
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P
2þA22
1=3 a2IiðIi þ 1Þ as B ¼ 2 A
[8,17,18], and the narrow inver-
1
r
i
A1þA2
960
920
880
840
800
760
sion in zero field [19].
Wavenumber,cm-1
To understand the expected changes in MARY spectra let us first
discuss what radical ion pairs may form in this system and what
contribution they will give to the observed spectrum. X-irradiation
of alkanes initially produces solvent radical cations (Alk+ꢄ) and
Figure 2. FT-IR spectra for solutions of 10ꢂ3 M PPO (A) and Zn(hfac)2(PPO)2 (B) in
n-decane. Asterisk marks vibrations found both in PPO and in Zn(hfac)2(PPO)2, the
plus sign marks Zn(hfac)2.

110
N.V. Sergey et al. / Chemical Physics Letters 504 (2011) 107–112
4
1x107
a 8x10
b 10
Zn(hfac)2PPO
PPO
2
9x106
7x104
6x104
5x104
4x104
3x104
2x104
8x106
8
6
4
2
0
7x106
6x106
5x106
4x106
3x106
2x106
1x104
1x106
0
0
200
250
300
350
400
450
500
0
2
4
6
8
10 12 14 16 18 20
Time, ns
Wavelength, nm
Figure 3. (a) Optical absorption (shorter wavelengths, left axis) and fluorescence (longer wavelengths, right axis, excitation at 300 nm) spectra for solutions of
Zn(hfac)2(PPO)2 and PPO in n-decane. (b) Kinetics of fluorescence (excitation at 320 nm, detection at 357 nm) for solution of Zn(hfac)2(PPO)2 in n-decane. Open triangles show
the instrument function, solid line shows the measured fluorescence response, and open circles show approximation of the experimental kinetics with a standard fitting
function F = A + B1.exp(ꢂt/T1) + B2.exp(ꢂt/T2), approximation parameters are T1 = 1.3 ꢃ 10ꢂ9 s, B1 = 3.4 ꢃ 10ꢂ2, 99%, T2 = 4.8 ꢃ 10ꢂ9 s, B2 = 1.1 ꢃ 10ꢂ4, 1%.
excess electrons. If a compound with sufficient electron affinity, in
this case Zn(hfac)2(PPO)2 or PPO, is present in solution, it captures
the electrons forming the radical anion Lꢂꢄ, where L stands for both
free ligand and complex. Due to lower ionization potentials as
compared to alkane both the complex and the free PPO can also
capture the positive charge from Alk+ꢄ and thus form the radical
cations L+ꢄ. MARY spectroscopy is sensitive to recombining radical
ion pairs containing at least one luminophore, and in this case two
types of pairs, (Alk+ꢄ/Lꢂꢄ) and (L+ꢄ/Lꢂꢄ), can contribute to the ob-
served signal. To a first approximation their contributions to the
spectrum will be additive and will differ in the widths of the mag-
low concentrations of the complex the contribution of the narrow
signal is higher as compared to the spectrum of free ligand, both
for equal concentrations of the acceptor (left panels) and for equal
concentrations of the luminophore molecules in the system (right
panels). For solutions of PPO the acceptor is clearly PPO itself, while
for solutions of the complex this can be, in principle, either the
complex itself or some its derivative, including PPOꢂ expelled upon
electron capture, and this point requires further clarification. What
is positive, the complex is not completely dissociated in solution
prior to electron capture, and the right curves in Figure 4 support
this conclusion. However, at the highest equal concentrations of
6.6 ꢃ 10ꢂ4 M (bottom left in Figure 4) the MARY spectra for the
complex and for the free ligand completely coincide, while upon
comparison of the 2:1 spectra (bottom right in Figure 4, spectra
of 6.6 ꢃ 10ꢂ4 M PPO and of 3.3 ꢃ 10ꢂ4 M Zn(hfac)2(PPO)2) they al-
netic field effect B , as the effective hyperfine coupling A in the al-
r
kane radical cation Alk+ꢄ is substantially higher than the couplings
for aromatic radical ions of L (both for the complex and for the free
ligand). Therefore, magnetic field effect for the pair (Alk+ꢄ/Lꢂꢄ) is
wider than for the pair (L+ꢄ/Lꢂꢄ). The relative amounts of the pairs
depend on the concentration of L. The radical anions Lꢂꢄ form rap-
idly due to high mobility of excess electrons and are always pres-
ent, even at concentrations of L below 10ꢂ4 M. On the other hand,
the radical cations L+ꢄ form much slower due to lower mobility of
solvent holes, and build up appreciable concentrations only at
higher concentrations of L of about 10ꢂ3 M needed to compete
with recombination. Thus at low concentrations of L only the broad
signal from the pair (Alk+ꢄ/Lꢂꢄ) is present in the observed spectrum,
but as the concentration of L is increased a narrower contribution
from the pair (L+ꢄ/Lꢂꢄ) progressively appears. This produces appar-
ent narrowing of the experimental spectra with increasing concen-
tration of L.
ready differ in the tails beyond the equal B fields. This is due to
r
the increase in the positive charge acceptor concentration, and is
observed both for the complex and for PPO simply upon doubling
their concentrations.
All concentration-dependent transformations of MARY spectra
can be ascribed to more efficient (rapid) charge capture to the com-
plex as compared to free PPO, which can be explained by the fol-
lowing simple model. The processes of charge capture from Alk+ꢄ
to L in the systems under study are diffusion-limited, proceeding
with the diffusion rate constant kD = 4pReffDeff, where Reff is the
effective reaction radius, often taken as the sum of geometrical ra-
dii of the molecules, Deff is the effective diffusion coefficient, the
sum of diffusion coefficients of the reaction participants. The geo-
metrical size of the complex is about three times larger than the
size of the free ligand, but this per se does not lead to more efficient
Figure 5 demonstrates this more clearly. Fig 5a shows three
curves: the spectrum for the maximum possible ligand concentra-
tion containing contributions of broad and narrow signals from
both types of pairs; the smoothed spectrum for the minimum rea-
sonable concentration of L with the broad magnetic field effect
from the pairs (Alk+ꢄ/Lꢂꢄ), and their difference approximating the
spectrum with the narrow magnetic field effect from the pairs
(L+ꢄ/Lꢂꢄ). Thus the spectrum for high concentration of L has been
decomposed into two components. Every experimental spectrum
can now be represented as a weighed sum of these two compo-
nents. As an example, Figure 5b shows a reconstruction of the
upper left panel from Figure 4.
reaction, as according to the Stokes–Einstein relation D = kBT/6pg
R
the diffusion coefficient for the complex should be three times
smaller than for the ligand, and the product DR remains the same.
However, the gain in the diffusion constant for a pair of reagents
can be provided by the domination of diffusion coefficient for the
second reaction partner, i.e., the positive charge carrier should
have high mobility. Indeed in some cases alkane radical cations
are known to have high mobility, with most of the data concerning
cycloalkanes, e.g., cyclohexane [20]. There also was some evidence
of high hole mobility at short times for non-cyclic radical cations,
and highly mobile squalane hole was reported [21,22]. Although
for other alkanes the mobile holes have not been unequivocally
identified, they are periodically implied upon in picosecond and
faster radiolysis studies [23]. Then in the reaction of positive
charge transfer from the solvent hole to charge acceptor, i.e., the
complex or the free ligand, the effective diffusion coefficient is
Returning now to experiment shown in Figure 4, at low concen-
trations of acceptors MARY spectra for equal concentrations of
Zn(hfac)2(PPO)2 and PPO in n-decane are different, the spectrum
for the complex has significantly lower B field (approximately
r
1.5 vs 2.5 mT), and, unexpectedly, much better signal to noise ratio
than the spectrum of free PPO. In the spectra for Zn(hfac)2(PPO)2 at

N.V. Sergey et al. / Chemical Physics Letters 504 (2011) 107–112
111
6.1*10-5M PPO
6.1*10-5M Zn(hfac)2(PPO)2
1.0*10-4M PPO
6.1*10-5M Zn(hfac)2(PPO)2
-4
-4
-4
-2
-2
-2
0
0
0
2
4
6
8
10
12
14
-4
-4
-4
-2
0
2
4
6
8
10
12
14
Magnetic field,mT
Magnetic field,mT
1.7*10-4M PPO
1.7*10-4M Zn(hfac)2(PPO)2
3.3*10-4M PPO
1.7*10-4M Zn(hfac)2(PPO)2
2
4
6
8
10
12
14
-2
0
2
4
6
8
10
12
14
Magnetic field,mT
Magnetic field,mT
6.6*10-4M PPO
3.3*10-4M Zn(hfac)2(PPO)2
6.6*10-4M PPO
6.6*10-4M Zn(hfac)2(PPO)2
2
4
6
8
10
12
14
-2
0
2
4
6
8
10
12
14
Magnetic field,mT
Magnetic field,mT
Figure 4. Pairwise comparison of MARY spectra of Zn(hfac)2(PPO)2 (bold lines) and PPO (lighter lines) at indicated concentrations in n-decane.
a
b
PPO=6.1*10-5M
Zn(hfac)2(PPO)2=6.1*10-5M
(L+/L-) and (Alk+/L-)
(Alk+/L-)
Reconstructed PPO=6.1*10-5M
Reconstructed Zn(hfac)2(PPO)2=6.1*10-5M
(L+/L-)
-4
-2
0
2
4
6
8
10
12
14
-4
-2
0
2
4
6
8
10
12
14
Magnetic field, mT
Magnetic field, mT
Figure 5. (a) Decomposition of experimental spectrum for 6.6 ꢃ 10ꢂ3 M PPO (filled circles) into two components, the wide (6.1 ꢃ 10ꢂ5 M PPO, smoothed, open circles) and
narrow (derived, stars) magnetic field effects. (b) Reconstruction of the upper left panel from Figure 4 as a weighed sum of the two components: the ratio of broad:narrow is
1:0 for the apparently broader spectrum (squares) and 1:0.75 for the apparently narrower spectrum (circles).

112
N.V. Sergey et al. / Chemical Physics Letters 504 (2011) 107–112
determined mostly by the solvent hole, while the larger complex
contributes mostly to the reaction radius. The experiments are
then interpreted as follows: Zn(hfac)2(PPO)2, showing more effi-
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.cplett.2011.01.056.
cient hole capture, forms the signal with the smaller B from the
r
pairs (L+ꢄ/Lꢂꢄ) at lower concentrations than free PPO, but with
increasing concentration of acceptor this effect is gradually satu-
rated, the spectra for the complex and the free ligand coincide
and then behave similarly with further increase in concentration.
References
[1] K.M. Salikhov, R.Z. Sagdeev, A.L. Buchachenko, Yu.N. Molin, Spin Polarization
and Magnetic Field Effects in Radical Reactions, Elsevier, Amsterdam, 1984.
[2] U.E. Steiner, T. Ulrich, Chem. Rev. 89 (1989) 51.
[3] F.B. Sviridenko, D.V. Stass, Yu.N. Molin, Chem. Phys. Lett. 297 (1998) 343.
[4] E.V. Kalneus, D.V. Stass, K.L. Ivanov, Yu.N. Molin, Mol. Phys. 104 (2006) 1751.
[5] E.V. Kalneus, D.V. Stass, Yu.N. Molin, Appl. Magn. Reson. 28 (2005) 213.
[6] D.V. Stass, N.N. Lukzen, B.M. Tadjikov, V.M. Grigoryantz, Yu.N. Molin, Chem.
Phys. Lett. 243 (1995) 533.
[7] G. Grampp, P.J. Hore, M. Justinek, S. Landgraf, N.N. Lukzen, Res. Chem.
Intermed. 31 (2005) 567.
[8] V.N. Verkhovlyuk, D.V. Stass, N.N. Lukzen, Yu.N. Molin, Chem. Phys. Lett. 413
(2005) 71.
4. Conclusions
In this work we implemented MARY spectroscopy in the pres-
ence of a tailored complex Zn(hfac)2(PPO)2 to extend this tech-
nique to organometallic species. The synthesized complex is
moderately soluble in alkanes and does not loose its ligands into
the bulk. In n-decane solution the complex can capture both nega-
tive and positive charges forming ‘radical ionic’ states, and has
[9] F.B. Sviridenko, D.V. Stass, Yu.N. Molin, Mol. Phys. 101 (2003) 1839.
[10] NIST electronic database, http://webbook.nist.gov.
[11] F.A. Cotton, R.H. Holm, J. Am. Chem. Soc. 82 (1960) 2979.
quantum yield and lifetime of fluorescence
u
ꢀ0.8 and ꢀ1.3 ns,
s
respectively, and thus can be used in spin chemical studies under
X-irradiation with fluorescence detection. The first MARY spectra
with the participation of the complex (‘radical ions’ of Zn(hfac)2(P-
PO)2) were obtained bearing characteristic features that can be
attributed to the radical cation and radical anion of the complex.
The concentration-dependent transformations of the spectra can
be explained by a larger positive charge capture radius of the com-
plex as compared to the free ligand.
[12] Bruker AXS Inc.,
Madison, Wisconsin, USA, 2004.
[13] G.M. Sheldrick, ADABS v.2.01, Bruker/Siemens Area Detector Absorption
APEX2 (Version 1.08), Bruker Advanced X-ray Solutions,
S
Correction Program, Bruker AXS, Madison, Wisconsin, USA, 1998.
[14] G.M. Sheldrick, SHELXTL v. 5.10, Structure Determination Software Suite, Bruker
AXS, Madison, Wisconsin, USA, 1998.
[15] M.L. Morris, R.W. Moshier, R.E. Sievers, Inorg. Chem. 2 (1963) 411.
[16] I.B. Berlman, Handbook of Fluorescence Spectra of Aromatic Molecules,
Academic press, New York, 1971.
[17] H. Staerk, R. Treichel, A. Weller, Chem. Phys. Lett. 96 (1983) 28.
[18] M. Justinek, G. Grampp, S. Landgraf, P.J. Hore, N.N. Lukzen, J. Am. Chem. Soc.
126 (2004) 5635.
[19] C.R. Timmel, U. Till, B. Brocklehurst, K.A. McLauchlan, P.J. Hore, Mol. Phys. 95
(1998) 71.
Acknowledgements
[20] J.M. Warman, in: J.H. Baxendale, F. Busi (Eds.), The Study of Fast Processes and
Transient Species by Pulse Radiolysis, Reidel, Boston, 1981, p. 433.
[21] I.A. Shkrob, M.C. Sauer Jr., A.D. Trifunac, J. Phys. Chem. 100 (1996) 5993.
[22] O.M. Usov, D.V. Stass, B.M. Tadjikov, Yu.N. Molin, J. Phys. Chem. A 101 (1997)
7711.
The authors are grateful to Dr. V.V. Korolev, Dr. A.I. Kruppa, V.N.
Verkhovlyuk and A.G. Matveeva for valuable discussions through-
out this work. Financial support of Russian Foundation for BasicRe-
search, Project 08-03-00741-a, and the Program ‘Leading Scientific
Schools’, Project NS-4249.2010.3, is gratefully acknowledged.
[23] A. Saeki, S. Tagawa, Pure Appl. Chem. 81 (2009) 45.