Estimation of the Fraction of Spin-Correlated Radical Ion Pairs in Irradiated Alkanes using Magnetosensitive Recombination Luminescence from Exciplexes Generated upon Recombination of a Probe Pair

Z. Phys. Chem. 2016; aop

Anatoly R. Melnikov*, Vladimir N. Verkhovlyuk, Evgeny V. Kalneus,

Valeri V. Korolev, Vsevolod I. Borovkov, Peter S. Sherin,

Maria P. Davydova, Sergei F. Vasilevsky and Dmitri V. Stass

Estimation of the Fraction of Spin-Correlated

Radical Ion Pairs in Irradiated Alkanes

using Magnetosensitive Recombination

Luminescence from Exciplexes Generated

upon Recombination of a Probe Pair

DOI 10.1515/zpch-2016-0819

Received June 14, 2016; accepted August 26, 2016

Abstract: We suggest a convenient probe exciplex system for studies in radia-

tion spin chemistry based on a novel acceptor-substituted diphenylacetylene,

1-(phenylethynyl)-4-(trifluoromethyl)benzene that has a very short fluorescence

lifetime (<200 ps) and low quantum yield (0.01) of intrinsic emission, provides

efficient electron capture in alkanes and efficient exciplex formation upon

recombination in pair with DMA radical cation, while exhibiting a shifted to red

exciplex emission band as compared to the parent system DMA – diphenylacety-

lene. After chemical, luminescent, radiation and spin-chemical characterization

of the new system we used the magnitude of magnetic field effect in its exciplex

emission band for experimental estimation of the fraction of spin-correlated

radical ion pairs under X-irradiation with upper energy cutoff 40 keV in a set of

11 alkanes. For linear and branched alkanes magnetic field effects and the corre-

*Corresponding author: Anatoly R. Melnikov, Voevodsky Institute of Chemical Kinetics and

Combustion SB RAS, 3, Institutskaya Str., 630090 Novosibirsk, Russian Federation; and

Novosibirsk State University, 2, Pirogova Str., 630090 Novosibirsk, Russian Federation,

e-mail: melnikov@kinetics.nsc.ru

Vladimir N. Verkhovlyuk, Evgeny V. Kalneus, Valeri V. Korolev: Voevodsky Institute of Chemical

Kinetics and Combustion SB RAS, 3, Institutskaya Str., 630090 Novosibirsk, Russian Federation

Vsevolod I. Borovkov, Maria P. Davydova, Sergei F. Vasilevsky and Dmitri V. Stass: Voevodsky

Institute of Chemical Kinetics and Combustion SB RAS, 3, Institutskaya Str., 630090,

Russian Federation; and Novosibirsk State University, 2, Pirogova Str., 630090 Novosibirsk,

Russian Federation

Peter S. Sherin: International Tomography Center SB RAS, 3a, Institutskaya Str., 630090

Novosibirsk, Russian Federation; and Novosibirsk State University, 2, Pirogova Str., 630090

Novosibirsk, Russian Federation

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

2ꢀ

sponding fractions are approximately 19–20% and 0.28, while for cyclic alkanes

they are lower at 16–17% and 0.22, respectively.

Keywords: diphenylacetylene; exciplex luminescence; magnetic field effect;

radical ion pair; spin correlation; X-ray irradiation.

Dedicated to: Kev Salikhov on the occasion of his 80^thbirthday.

1 Introduction

Paramagnetic intermediates are key participants in diverse chemical processes

and determine the direction and depth that they proceed [1]. Often they are the

“witnesses” of the occurring events, such as radiation of photo-induced damage

to a material, or are introduced deliberately as labels or probes that provide impor-

tant information on their local environment in a solid matrix [2, 3], on the type

and magnitude of molecular mobility in a condensed matter [4–7], on the degree

of oxygenation and acidity in biological systems [8, 9], on specific for electron

spins interactions that determine the suitability of a material for application in

the devices of modern molecular electronics [10–13], etc. The most direct experi-

mental methods to study the radicals are certainly the various implementations

of ESR, which has recently enjoyed an explosive experimental and methodologi-

cal development similar to the earlier proliferation of pulsed NMR.

In those cases when radical concentrations are too low and do not reach the

sensitivity threshold of conventional ESR with detection of microwave absorp-

tion the methods of indirect ESR detection can be called for [14–17]. Here spins

are also manipulated with static and microwave magnetic fields, but detected

is not directly MW absorption, but some product of a spin-dependent reaction

in a correlated radical pair or more complex system of electron spins, such as

luminescence or the yield of cage recombination products. By accumulating the

result of reaction in a stable product or converting the act of recombination in the

emission of an optical photon, such methods attain exceptional sensitivities for

concentrations of transient radicals. On the other hand, they had posed a very

important question on the role of spin correlation in chemical reactivity, which

resulted in the establishment of a new field of chemical physics, the so-called

spin chemistry [18]. The methods of spin chemistry have significantly broadened

our knowledge on the mechanisms of important reactions including electron

transfer and bond cleavage, as well as on the structure and properties of short-

lived radicals and radical ions, and have, e.g. much helped in establishing the

mechanism of photosynthesis and the structure of photosynthetic reaction center

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

ꢂ3

Fraction of Spin-Correlated Pairsꢂ

[19–22]. The systems of spin chemistry are today used to study the problems of tel-

eportation of a quantum state [23] and of quantum measurement [24–27], the well

established ideas and results of spin chemistry, although slowly, but are finding

their way into the field of modern functional materials, molecular electronics and

spintronics [10–13].

While the methods of spin chemistry provide a tremendous boost in sensitiv-

ity to low concentrations of the studied radicals, they also bring their specific

complications related to signal formation by not just one, but at least a pair of

correlated spins. Thus, for example, for contact radical pairs forming upon pho-

tolysis of systems like aromatic ketones and observed by polarized NMR spectra

of their recombination products (CIDNP) [28], and for contact radical ion pairs

forming upon photoinduced electron transfer to produce equilibrated exciplex

and solvent-separated pair and detected by exciplex luminescence [29], it is nec-

essary to correctly take into account a substantial exchange interaction between

pair partners and its spatial dependence.

The effects of inter-radical interaction in contact pairs are nearly eliminated

in radiation spin chemistry, when radical ion pairs are generated by ionizing irra-

diation of a nonpolar solution of suitable electron acceptors and donors, with

detecting the intensity of recombination luminescence as a function of applied

static and/or resonant MW magnetic field [30, 31]. In this case the initial partner

separation in the pair is tens of angstroms, while back electron transfer – the

recombination – occurs at a distance of about 10 angstroms, and thus the two

radical ions of the pair, apart from the obvious coulomb attraction, can be con-

sidered independent and linked only by the initial spin correlation imposed upon

pair generation at the moment of ionization of a solvent molecule. For this reason

radiation chemical systems have served as a proving ground for many theoreti-

cal and experimental approaches of spin chemistry, including time-resolved

magnetic field effects and quantum beats [32, 33], optically detected ESR [16, 17],

level-crossing (MARY) spectroscopy [34], etc. These techniques certainly have

their photoinduced counterparts, but these are too abundant to be discussed

here.

However, in radiation spin chemistry as well the gain in the simplicity of

describing the spin evolution of radical ion pair has to be paid for by certain

experimental complications. Radiation generation produces the pairs not as iso-

lated entities, but rather as the so-called spurs, which for the X-irradiation with

upper energy cutoff of 40 keV used in this work are about 100 angstroms in diam-

eter and contain about 5–6 overlapping pairs [35–37]. The rapid cross-recombi-

nation of the primary pairs electron/solvent radical cations in the spurs within

picoseconds leaves the last remaining pair that is not necessarily geminate, i.e.

from partners coming from the same parent molecule, and thus not necessarily

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

4ꢀ

spin-correlated. Since all methods of spin chemistry usually rely on the initial

spin correlation in the pair, this leads to a substantial drop in the magnitudes of

the experimentally observed effects and to introduction of an empirical parameter

in the theoretical description, the fraction of spin-correlated pairs θ, that in the

conditions of X-irradiation normally falls in the range 0.1–0.4. Several experimen-

tal approaches have been proposed to determine θ and are discussed later, but

all of them are rather complex and can be applied to a limited range of systems,

therefore an extension in the set of tools to determine θ would be most welcome.

Furthermore, radiation generation of the pairs with detection of recombi-

nation luminescence is accompanied by a host of background processes that

also produce emission of the luminophore present in the solution, which is

not related to spin correlation in the pair. An example would be direct energy

transfer from an excited solvent molecule to a luminophore molecule, which

produces the same emission as recombination of the spin-correlated pair com-

prising radical ion of the luminophore. Therefore the fraction of magnetosen-

sitive emission in the total intensity of sample luminescence is relatively low

and usually is much lower than the maximum possible value determined by the

fraction of spin-correlated pairs, being at best several percent in a CW experi-

ment. Moreover, as all emission usually occurs in the same spectral band, it

is practically impossible to separate magnetosensitive luminescence related

to spin evolution of spin-correlated pairs from background emission by spec-

tral methods in a CW experiment. This is in principle possible in time-resolved

experiments, since temporal behavior of the kinetics of recombination lumines-

cence for different channels of its generation may be different, but this would

require a complete first-principles modeling of the experimental kinetics, which

is at present not feasible due to the complexity of the reaction scheme of radia-

tion chemical processes and the necessity in a large number of poorly known

parameters for its specification [38, 39].

In this work we suggest a simple experimental approach to substantially

increase the sensitivity of CW methods of radiation spin chemistry with detec-

tion of recombination luminescence due to spectral shift of the recombina-

tion luminescence in a separate emission band and its selective detection,

and apply it to experimentally estimate the fraction of spin-correlated radical

ion pairs in X-irradiated alkane solutions. The approach relies on the recently

found efficient exciplex formation upon recombination of radiation-induced

radical ion pairs [40, 41]: since at the moment of recombination the two radical

ions are relatively close, after the back electron transfer to yield an electroni-

cally excited molecule the second partner of the pair happens to be next to

it, which creates optimal conditions for exciplex generation. As discussed

in the previous works [40, 41], the recombination exciplexes can arise from

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

ꢂ5

Fraction of Spin-Correlated Pairsꢂ

practically arbitrary donor-acceptor pairs that energetically and sterically

allow this, and produce a new longer wavelength band in the emission spec-

trum. The difference of this radiation generation of exciplexes in comparison

with well-studied photochemical and electrochemical methods is in the arbi-

trary short lifetime of the excited state generated upon recombination, as the

precursor pair for the exciplex is formed nearly at contact and the diffusion-

controlled bulk quenching reaction needed for optical generation [42–50] is

not required, and quite relaxed requirements to radical ion stability germane

for electrochemical generation of exciplexes from radical ions injected from

electrodes or chemically [51–53], since radiation-induced pairs recombine in

tens of nanoseconds.

Experimentally convenient pairs to study the recombination exciplexes

turned out to be pairs N,N-dimethylaniline (DMA) – aromatic electron acceptor,

which produce emission spectra containing the bands of intrinsic emission of

DMA and the acceptor, as well as a red-shifted emission band of exciplex (DMA-

acceptor)*. Especially convenient is the pair DMA – diphenylacetylene (tolane, T)

[40], since the very short lifetime of intrinsic emission of T (8 ps [54, 55]) for sure

eliminates exciplex generation via bulk reaction of its excited molecule with a

DMA molecule, so the exciplexes form exclusively via the recombination channel,

while its very low quantum yield of intrinsic luminescence (0.00336 [56]) sim-

plifies the spectrum or recombination luminescence, leaving only the bands of

intrinsic DMA emission and exciplex emission.

Furthermore, as the exciplex emission in this pair is generated only via the

pair recombination, its intensity is magnetosensitive with the magnitude of mag-

netic field effect of up to 20%, reaching the upper limit set by the fraction of spin-

correlated pairs. Since recombination of the radical ion pair in these conditions is

irreversible, all magnetosensitivity of the exciplex emission is completely inher-

ited from spin evolution in the radiation-induced radical ion pair, while the sole

effect of exciplex formation is simply a transfer of the recombination lumines-

cence in the exciplex emission band, which separates it from other, background

processes of excited state generation. Therefore the magnetosensitivity of emis-

sion of radiation-induced exciplexes in alkanes is principally different from the

magnetosensitivity of emission of photoinduced exciplexes in moderately polar

solvents that is due to reversible transitions between exciplexes and solvent-sep-

arated radical ion pairs [29, 57–63].

The only inconvenience of the DMA – diphenylacetylene pair in this context

is a rather substantial overlap of the emission bands of DMA and exciplex, which

requires spectrum decomposition into individual components before analyzing

the magnetic field effect. However, electronic (and steric) properties of donor-

acceptor systems can be adjusted by introducing donor or acceptor substituents

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

6ꢀ

in the parent molecules [57, 64, 65]. While for most classic donor-acceptor systems

(naphthalene – DMA, anthracene – DMA, pyrene – DMA) synthesizing a broad

range of substituted electron acceptors is a rather daunting task, as for them there

is no single and straightforward reaction of aromatic substitution and each target

substituted molecule would require a dedicated synthetic procedure, almost any

substituted diphenylacetylenes can be readily prepared by Sonogashira or Castro

cross-coupling [66, 67] from the corresponding arylhalogenide and either termi-

nal acetylene or acetylenide.

In this work we suggest a more convenient probe exciplex system for studies

in radiation spin chemistry. To this end using the Sonogashira reaction we syn-

thesized a novel acceptor-substituted diphenylacetylene, 1-(phenylethynyl)-

4-(trifluoromethyl)benzene (p-trifluorotolane, CF₃-tolane, FT), which retains

all useful in this context properties of diphenylacetylene (a very short lifetime

and rather low quantum yield of intrinsic emission, efficient electron capture in

alkanes to produce radical anion and efficient exciplex formation upon recombi-

nation in pair with DMA radical cation), while exhibiting a shifted to red exciplex

emission band as compared to the parent system DMA – diphenylacetylene. After

chemical, luminescent, radiation and spin-chemical characterization of the new

system we used the magnitude of magnetic field effect in its exciplex emission

band for experimental estimation of the fraction of spin-correlated pairs in a rep-

resentative set of linear, cyclic and branched alkanes.

2 Materials and methods

2.1 Experimental

Purified n-hexane, n-heptane, n-octane, n-decane, n-dodecane, n-hexadecane,

isooctane, 2,3-dimethylbutane, squalane, cyclohexane, methylcyclohexane were

used as the solvents. DMA (99%, Aldrich) prior to experiments was distilled over

zinc powder collecting the 193–195ꢀ°C fraction. Chromatographic analysis combined

with mass-spectrometry showed DMA purity more than 99%, with N-methylaniline

and N,N-dimethyl-m-aminoaniline as the major impurities. All experiments were

carried out at room temperature without explicit temperature control.

The CW spectra of magnetosensitive recombination luminescence were

obtained with a Magnetically Affected Reaction Yield (MARY) spectrometer

with spectral resolution of luminescence (an MDR-206 grating monochroma-

tor, LOMO Photonics, St. Petersburg, Russia, objective focus length 180 mm,

grating 1200 lines·mm⁻¹, inverse linear dispersion 4.3 nm·mm⁻¹, and a FEU-100

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

ꢂ7

Fraction of Spin-Correlated Pairsꢂ

PMT for detection) described elsewhere [68]. The samples were irradiated with

a BSV-27-Mo X-ray tube (Svetlana, St. Petersburg, Russia), 40 kV, 20 mA, detec-

tion monochromator slits 2.2 mm/2.2 mm (spectral resolution about 10 nm). The

presented spectra are averages over 8–16 independent wavelength scans of 256

points. Luminescence intensities are given as arbitrary units corresponding

to the output signal of detector and are consistent for all spectra. The pairs of

spectra in magnetic field (20 mT) and in zero field were recorded simultaneously

for each scan by registering at each wavelength point the intensity of emission

with applied field and without field before advancing to the next point. The

samples for magnetic field effect measurements were 300 μL of degassed solu-

tion sealed in a thin-walled round ampoule with O.D. 5 mm made from molyb-

denum glass, which produces no intrinsic luminescence under X-radiation. The

concentration dependence and the oxygen quenching spectra in low-volatile

n-dodecane were taken in a reusable cuvette with vacuum stopcock instead of

individual samples to improve repeatability of sample geometry. Further details

on the experimental procedure and the choice of ampoule material can be

found elsewhere [68].

Optically detected ESR spectra and modulation optically detected magnetic

field effects (MARY spectra) under X-irradiation were recorded in stationary con-

ditions as described in refs. [69, 70], respectively. The sample containing about

1 mL of degassed solution in a quartz cuvette was placed into magnetic field of

a Bruker ER-200D CW X-Band ESR spectrometer equipped with an X-ray tube for

sample irradiation (Mo, 40 kV, 20 mA), a pair of coils with a separate current

source to provide constant “negative” shift of the field (for MARY experiments),

and a PMT for fluorescence detection. All integral emission without spectral

shaping was detected. The scanned magnetic field was modulated at a frequency

of 12.5 kHz with an amplitude of up to 0.5 mT. A Stanford SR-810 Lock-In Ampli-

fier and computer averaging over four scans were used to get the spectra as the

first derivatives of the actual field dependencies. Microwave power for OD ESR

experiments was 0.8 W. MARY experiments were carried out without microwave

pumping.

The discussed radiation-induced luminescence kinetics were obtained on a

home-built nanosecond pulse fluorometer [71]. The luminescence is generated

by short – about 1.5–3 ns – X-ray pulses generated by hitting a molybdenum

target by a bunch of electrons with the energy of 30–40 keV. The full intensity of

recombination luminescence without spectral selection is detected in this experi-

ment using time-correlated single photon counting. The magnetic field at the

sample can be controlled in the range of 0–1.9 T, 0.1 T was used in this work. A

more detailed description of the experimental procedure for X-ray induced time-

resolved experiment can be found elsewhere [72].

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

8ꢀ

The luminescent properties of the synthesized 1-(phenylethynyl)-4-

(trifluoromethyl)benzene under optical excitation were characterized on a com-

mercial Edinburgh Instruments FLSP-920 spectrofluorometer (Great Britain).

CW emission and excitation spectra were recorded using a wide-band Xenon

lamp (OSRAM XBO, 450W, Germany), slit widths for excitation and detection

monochromators 2.5 mm. Time-resolved fluorescence decay profiles were taken

using an EPLED-270 diode, excitation wavelength 275 nm, full width at the half

maximum 0.8 ns, with spectral resolution of the detection monochromator 20

nm. Kinetics of fluorescence decay were analyzed using a numerical convolu-

tion of the instrument function with an exponential function. Optical absorp-

tion spectra were recorded using a Shimadzu UV-2401PC (Japan) and an Agilent

HP8453 (USA) spectrometers.

2.2 Synthesis of 1-(phenylethynyl)-4-(trifluoromethyl)benzene

The target 1-(phenylethynyl)-4-(trifluoromethyl)benzene was prepared by Sonoga-

shira cross-coupling [66] of phenylacetylene and 1-bromo-4-(trifluoromethyl)

benzene as shown in Scheme 1.

Reagent weights were calculated for 2·10⁻³mol of the starting bromide with

10% molar excess of the terminal acetylene in 10 mL of toluene. The reaction was

catalyzed by 5 mg [Pd(PPh₃)₂Cl₂], with 5 mg CuI and 1 mL triethylamine (3× molar

excess) added to the mixture, which was stirred in the argon atmosphere for 1 h

at 70°C. After completion of the reaction (control by TLC, Merck Silica gel 60 F₂₅₄

plates) the produced toluene solution was passed through a column with SiO₂

using toluene as the eluent and evaporated in vacuo.

The obtained residue was dissolved in n-hexane and passed through a column

with SiO₂, using n-hexane as the eluent and monitoring product separation with

TLC. The eluate of 1-(phenylethynyl)-4-(trifluoromethyl)benzene was evaporated

in vacuo and recrystallized twice from n-hexane. The obtained compound is a

white crystalline powder with mp 102–104ꢀ°C (lit. 100ꢀ°C), IR: 1107; 1323; 1608;

2220 cm⁻¹; ¹H NMR (400 MHz, CDCl₃): 7.33–7.37 (m, 3H); 7.51–7.54 (m, 2H); 7.60 (dd,

1

4H). The obtained analytical results agree with published data [73], the raw H

NMR and FTIR spectra are given in Supporting Information.

[Pd(PPh₃)₂CI₂]

F₃C

Br +

Ar, CuI, NEt₃

Scheme 1:ꢀSynthesis of 1-(phenylethynyl)-4-(trifluoromethyl)benzene by Sonogashira cross-

coupling [66].

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

ꢂ9

Fraction of Spin-Correlated Pairsꢂ

3 Results and discussion

3.1 Optical and luminescent properties of 1-(phenylethynyl)-4-

(trifluoromethyl)benzene

Figure 1 shows optical absorption spectra of the synthesized compound FT in

alkane in comparison with unsubstituted diphenylacetylene T. As can be seen,

introduction of the -CF₃substituent in the para-position of diphenylacetylene

does not change the shape of the absorption spectrum and only leads to its red

shift of about 5 nm. This is rather typical of para-substituted diphenylacetylenes

with compact substituents that do not extend π-system and significantly lower its

symmetry [74, 75].

As demonstrated elsewhere [76], the luminescent properties of para-substituted

diphenylacetylenes are also rather close to unsubstitited tolane. Thus, 1-metoxy-4-

(phenylethynyl)benzene, similar to T, emits from its second singlet excited state

having lifetime of 7 ps [74, 75, 77]. Transition S₁ꢀ→ꢀS₀in this molecule is also forbid-

den, and still S₁lifetime is very short at 270 ps, as determined in ref. [76].

Figure 2 shows spectra of luminescence and excitation and luminescence

decay profile for solution of FT in acetonitrile. Luminescence quantum yield

determined by procedure described in ref. [78], taking solution of T in ethyl

alcohol as the reference with a known and similar quantum yield (0.00336 at

room temperature [56]), was found to be 0.010ꢀ ꢀ0.001 at room temperature. The

estimated from fitting upper bound for lifetime of FT excited state is about 200

ps, which is attainable time resolution for the used excitation diode, and most

probably corresponds to forbidden S₁ꢀ→ꢀS₀transition, as in unsubstituted T [74,

40000

30000

1

20000

2

10000

0

200

225

250

275

300

325

350

λ, nm

Fig. 1:ꢀOptical absorption spectra for solutions of T (1) and FT (2) in n-dodecane.

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

10ꢀ

ꢀA.R. Melnikov et al.

10⁴

10³

10²

10¹

1

2

90

2

1

A

B

3

4

60

30

0

250

300

350

400

450

500

550

0

2

4

6

8

10

Time, ns

λ, nm

Fig. 2:ꢀ(A) Normalized for equal emission intensity for excitation at 287 nm excitation spectra of

FT acetonitrile with detection at 310 nm (1), 330 nm (2) and 450 nm (3). Normalized luminescence

spectrum of FT in acetonitrile (4), excitation at 275 nm, spectral resolution 2.5 nm, FT concen-

tration 6·10⁻⁶M. (B) Luminescence decay profile of FT in acetonitrile, excitation at 275 nm and

detection at 330 nm (1), and instrument function of the excitation source (2). Spectral resolution

20 nm, FT concentration 6·10⁻⁶M. Curve 3 shows fitting of the luminescence decay profile by an

exponential function with characteristic time 211 ps. Similar to unsubstituted diphenylacetylene,

the time provided by such fitting is most probably attributable to the slower forbidden transition

S₁ꢁ→ꢁS₀, with no observable traces of the faster allowed transition S₂ꢁ→ꢁS₀, while bimodal shape

of the spectrum corresponds to both transitions S₁ꢁ→ꢁS₀and S₂ꢁ→ꢁS₀[74, 76].

75, 77]. The shorter-living S₂state was not resolved in the luminescence decay pro-

files. Thus, the synthesized para-CF₃-substituted diphenylacetylene has lumines-

cent properties similar to unsubstituted tolane, exhibiting very short lifetimes of

excited states (both S₂and S₁) and low but finite luminescence quantum yield. As

demonstrated below, in CW experiments under X-irradiation FT produces prac-

tically no intrinsic luminescence in comparison with emission in the bands of

the second partner (DMA) and exciplex, while in time-resolved experiments with

single photon counting its quantum yield, being a factor of three higher than for

T, turns out to be sufficient to produce kinetics of recombination luminescence at

nanosecond time scale for comfortably low concentrations of about 10⁻³M.

3.2 Exciplex formation in the pair 1-(phenylethynyl)-4-

(trifluoromethyl)benzene – DMA via radical ion pair

recombination

Figure 3A shows the spectra of radiation-generated luminescence for mixtures of

1-(phenylethynyl)-4-(trifluoromethyl)benzene and DMA in n-dodecane for varied

concentrations of the electron acceptor. The major components of the spectra are

exciplex emission (long-wavelength band) that dominates as FT concentration is

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

Fraction of Spin-Correlated Pairsꢂ

ꢂ11

20

16

12

8

B

A

16

12

8

4

0

250

300

350

400

450

500

550

600

250

300

350

400

450

500

550

600

λ, nm

Fig. 3:ꢀ(A) Averaged spectra of X-ray generated luminescence for mixtures of 1-(phenylethynyl)-

4-(trifluoromethyl)benzene and DMA in n-dodecane. Concentration of DMA 1·10⁻²M in all

spectra; FT: ● – 1·10⁻²M, ○ – 6·10⁻³M,  – 4·10⁻³M,  – 2·10⁻³M, ■ – 1·10⁻³M, □ – 6·10⁻⁴M,

 – 2·10⁻⁴M. (B) Spectra of X-ray generated luminescence for mixtures of: ● – 6·10⁻³M T,

1·10⁻²M DMA; ○ – 6·10⁻³M FT, 1·10⁻²M DMA in n-dodecane.

increased and intrinsic emission of DMA (short-wavelength band), which is itself

a satisfactory luminophore (quantum yield 0.11, fluorescence lifetime 2.4 ns [79])

and whose excited states are mostly generated by direct energy transfer from gen-

erously forming excited states of solvent molecules (n-dodecane quantum yield

0.0055 [80], fluorescence lifetime 4.2 ns [81]). Figure 3B shows taken in identical

conditions spectra of radiation-generated luminescence for mixtures of T or FT

and DMA in n-dodecane. It can be seen that the introduced acceptor substituent

did not hinder exciplex formation and shifted exciplex emission band to the red

by about 35 nm, significantly improving band separation in the spectrum as com-

pared to unsubstituted diphenylacetylene.

Using the earlier suggested approach for estimation of lifetimes of radiation-

induced exciplexes [82], the lifetime of 1-(phenylethynyl)-4-(trifluoromethyl)

benzene – DMA exciplex in n-dodecane at room temperature was estimated as

61ꢀ ꢀ3 ns, which is somewhat higher than 49ꢀ ꢀ2 ns for tolane – DMA exciplex

[82]. The lifetime was estimated from the degree of relative quenching of exci-

plex emission with dissolved oxygen in comparison to quenching of a reference

luminophore with similar excited state lifetime in identical conditions. Typical

spectra of radiation-generated luminescence for a degassed sample and the same

sample equilibrated with normal atmosphere are shown in Figure 4A. Figure 4B

shows the reconstructed exciplex emission spectrum for this system obtained by

subtracting the spectra from Figure 4A normalized to the maxima of the short-

wavelength band as suggested by the Reviewer to paper [82].

The area under a spectrum band is a standard measure of the number of

species contributing to the band, and thus the area under the exciplex emission

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

12ꢀ

12

10

8

6

4

2

0

A

B

6

4

2

0

250

300

350

400

450

500

550

600

300

350

400

450

500

550

600

λ, nm

Fig. 4:ꢀ(A) X-ray generated luminescence spectra of degassed (○) and equilibrated with normal

atmosphere (●) mixtures of 6·10⁻³M FT, 1·10⁻²M DMA in n-dodecane. The integral quenching

of the exciplex emission band I₀/Iꢁ−ꢁ1 in the indicated wavelength range is 3.7ꢁ ꢁ0.2. (B) Emis-

sion spectrum of 1-(phenylethynyl)-4-(trifluoromethyl)benzene – DMA exciplex obtained from

emission spectra of the same sample solution of 6·10⁻³M FT and 1·10⁻²M DMA recorded in the

presence and absence of dissolved oxygen (see ref. [82] for details).

band can be used to study the amount of generated exciplexes as a function of

experimental parameters, such as sample composition. For the 1-(phenylethynyl)-

4-(trifluoromethyl)benzene – DMA exciplex in n-dodecane integration can be

performed in the range 415–475 nm, in which emission comes exclusively from

the exciplex and contains no contribution from intrinsic DMA luminescence.

Baseline correction is performed by averaging over the initial part of the spec-

trum containing no emission and subtracting the average value from the spec-

trum. Figure 5A shows thus obtained areas vs. FT concentration for the spectra

0.20

0.15

0.10

0.05

0.00

100

80

60

40

20

0

B

A

0

100

200

300

400

500

600

700

0

2

4

6

8

10

12

[

FT]

, mM

[FT] ^-0,7, M^-0,7

Fig. 5:ꢀ(A) Area under the exciplex emission band in the range 415–475 nm vs. FT concen-

tration in the system 1-(phenylethynyl)-4-(trifluoromethyl)benzene – DMA in n-dodecane.

(B) Linearization of the dependence of Figure 5A in indicated coordinates, shown linear fit is

I⁻¹ꢁ=ꢁ0.0031ꢁ+ꢁ2.6·10⁻⁴/[FT]^0.7, αꢁ=ꢁ34.5, see expression (3).

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

Fraction of Spin-Correlated Pairsꢂ

ꢂ13

of radiation-generated luminescence from Figure 3A, which can be used to verify

the origin of the detected exciplex emission.

The characteristic shape of the obtained dependence is similar to concentra-

tion dependence of the intensity of luminescence for hexafluorobenzene (HFB)

solutions irradiated with β-particles reported in ref [83]. In this case the intensity

of HFB luminescence is determined by the efficiency of electron capture in the

primary radical ion pair by HFB according to reaction (1) below. The subsequent

recombination of the forming secondary radical ion pair via reaction (2), where

+

S ˙ is solvent radical cation, produces an excited HFB molecule that emits the

detected luminescence.

•

e⁻+HFB → HFB⁻

(1)

(2)

•

S⁺+HFB⁻→ HFB^∗+ S

To detect only the recombination luminescence the authors of ref. [83] used

alkanes with very low intrinsic quantum yield of fluorescence ϕ_f, such as isooc-

tane (ϕ_fꢀ<ꢀ10⁻⁵[80]). As was earlier experimentally demonstrated in refs. [84, 85],

the probability of electron capture by an acceptor molecule in the case of detect-

ing luminescence of the acceptor is given by the following expression:

(αC)^0.7

p^†=

(3)

1+(αC)^0.7

+

where p^†is the probability of quenching the primary radical ion pair S ˙/e⁻, which

is equal to the probability of electron capture by HFB in reaction (1), α is a quench-

ing parameter that depends on the particular solvent – acceptor system and does

not depend on acceptor concentration, C is HFB concentration in the mixture.

A similar expression was also obtained for the dependence of radiation yield of

alkyl radicals on the concentrations of alkylhalide electron acceptors/quenchers,

such as CH₃Br, in cyclohexane [86].

Within the context of the cited works, if all the observed excited species in

solution are generated only as a result of radical ion pair recombination via reac-

tion (2), the experimentally registered luminescence intensity should be pro-

portional to p^†. As a result, the reciprocal of the intensity of HFB luminescence

should be a linear function of C^−0.7where C is HFB concentration in the mixture,

while the ratio of the y-axis cutoff and the slope for the obtained linear function

produces the unknown parameter α that can be independently estimated theo-

retically [83, 84].

As Figure 5B demonstrates, in the described coordinates the concentration

dependence of integral exciplex emission from Figure 5a is also linear in the range

of FT concentration 10⁻⁴–10⁻²M. This confirms that in the system under study the

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

14ꢀ

exciplexes are indeed formed only via recombination of the secondary radical

+

ion pair D ˙/A⁻˙, where D is DMA and A is 1-(phenylethynyl)-4-(trifluoromethyl)

benzene. It should also be stressed that, as opposed to the cited concentration

dependence for HFB, in the system 1-(phenylethynyl)-4-(trifluoromethyl)benzene

– DMA there is no need in non-luminescing solvents, in which the excited states

are produced only upon recombination of radical ion pairs, as recombination

formation of exciplexes with emission shifted to the red relative to the intrinsic

emission of the mixture components automatically separates the recombina-

tion luminescence and helps get rid of non-recombination contributions to the

observed emission.

3.3 OD ESR and MARY spectra of the probe pair

Figure 6 shows typical results of CW radiation spin chemistry experiments,

i.e. Optically Detected ESR (OD ESR) and Magnetically Affected Reaction Yield

(MARY) spectra, for the mixtures of FT and DMA in n-dodecane. OD ESR provides

combined spectra of the short lived radical anion and radical cation that recom-

bine to produce detectable luminescence as a single trace that is approximately

the superposition of the independent spectra for the two partners [16, 17], apart

from the finer spectral shape details due to interlinked spin relaxation of the two

radical ions [87, 88] and, on very rare occasions, certain coherent features specific

A

B

2

1

2

1

-10

0

10

B, mT

20

30

40

315

320

325

330

335

340

B, mT

Fig. 6:ꢀ(A) OD ESR spectrum for a mixture of 1·10⁻²M FT and 1·10⁻²M DMA in n-dodecane:

1 – experimental spectrum taken with modulation amplitude 0.2 mT at 12.5 kHz, MW power

800 mW, 2 – its simulation (see text for details). (B) MARY spectrum for the same mixture of

1·10⁻²M FT and 1·10⁻²M DMA in n-dodecane: 1 – experimental spectrum taken with modulation

amplitude 0.5 mT at 12.5 kHz, 2 – its simulation (see text for details).

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

Fraction of Spin-Correlated Pairsꢂ

ꢂ15

for spin-correlated pairs [89–91] or partial spectral inversion at high MW power

due to spin-locking [92, 93], as well as consecutive pairs [94, 95]. When applied to

exciplex-forming systems as discussed in this work, the registered OD ESR spec-

trum can be used to identify the recombining pair from the characteristic hyper-

fine couplings of the two radical ions, with the possibility to detect all integral

luminescence from the sample without spectral limiting in the case of dominat-

ing exciplex emission.

Figure 6A shows an OD ESR spectrum for a mixture of 1-(phenylethynyl)-4-

(trifluoromethyl)benzene and DMA in n-dodecane. The presence of a -CF₃sub-

stituent in diphenylacetylene significantly improves the identification capability

of this approach, as fluorines usually have rather substantial hyperfine coupling

(HFC) constants [96], while several equivalent nuclei with the dominant HFC

produce a very characteristic ESR spectrum. Thus, a quartet from three equivalent

spin ½ nuclei is immediately recognizable in the spectrum shown in Figure 6a.

The accompanying simulation was obtained using WinSim package [97]. DMA

radical cation was modeled with a fixed set of HFC constants that were estimated

in DFT calculations in ref. [98]: 1.30 mT (6H), −0.72 mT (1N), 0.83 mT (1H), −0.43

mT (2H), 0.09 mT (2H) with a homogeneous width at half magnitude of 0.41 mT.

Radical anion of 1-(phenylethynyl)-4-(trifluoromethyl)benzene for this spectrum

was modeled as three equivalent fluorines with (hitherto unknown) HFC con-

stant of 1.67 mT and inhomogeneous peak-to-peak width of 0.58 mT. An addi-

tional species with an unresolved spectrum having a homogeneous width at half

magnitude of 0.57 mT was also introduced in the simulation, which corresponds

to exchange-narrowed in its parent matrix radical cation of n-dodecane, the pre-

+

cursor to DMA ˙.

We note that the shown spectrum taken at FT concentration of 10⁻²M already

corresponds to the region of intermediate exchange [99] (mean residence time

at the diffusion-controlled rate constant of about 10¹⁰M⁻¹s⁻¹is about 10 ns and is

comparable to the inverse ESR spectrum width of the radical anion in frequency

units), and upon reducing the concentration of FT down to 10⁻³M to suppress the

−

degenerate electron exchange between FT ˙ and neutral FT molecules the spec-

trum is modeled with slightly higher HFC for the three fluorines of about 1.87 mT,

which are closer to the intrinsic HFC values. However, even for such inconvenient

for quantitative interpretation concentrations the spectrum has a very character-

istic shape and unequivocally identifies the radical anion of the recombining pair.

The recombining radical ion pair can also be identified by its MARY spectrum

that is the first derivative of the magnetic field dependence of the intensity of

recombination luminescence in the vicinity of zero field without MW pumping

applied [34]. In this case the shape of the recorded spectrum is determined by

the second moments of the ESR spectra of the pair partners and their relative

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

16ꢀ

magnitudes (close values or dominating hyperfine interactions in one of the pair

partners), as well as lifetimes and relaxation times of the recombining spin-cor-

related radical ion pair [100].

Figure 6B shows an experimental MARY spectrum for the same sample used

for Figure 6A and its modeling that was performed in the model of equivalent

nuclei [101] which account for the required second moment of the ESR spec-

trum, with power-law recombination kinetics α/2(1ꢀ+ꢀαt)^3/2approximated as a set

of five exponentials [102]. HFC in the radical anion of FT was modeled as three

equivalent spin ½ nuclei with HFC constant 1.67 mT, HFC of DMA radical cation

was approximated as a set of 10 equivalent spin ½ nuclei with HFC constant 1.1

+

mT, which provides the required second moment of the ESR spectrum of DMA ˙.

Relaxation times T₂for both partners were set at 11 ns to account for the process

of degenerate charge exchange in both radical ions having concentration of their

parent molecules of 1·10⁻²M, which in MARY spectra reveals itself as phase relax-

ation [100, 103, 104].

The general look of the MARY spectrum with a broad featureless background

of magnetic field effect and not very prominent zero field line is rather typical for

a pair with substantial and evenly distributed between partners hyperfine cou-

plings. Although for systems with simple hyperfine structure the spectrum could

have had additional lines in magnetic fields that are multiples of the HFC con-

stant [101, 105–114], and three equivalent nuclei could have produced a specific

line at 15/4 of the coupling constant, as demonstrated earlier for radical anion of

1,3,5-trifluorobenzene [113], the substantial couplings in the other partner, rather

low value of the HFC constant for the three fluorines, and short relaxation times

together wipe away any resolved structure. It can also be seen that the model

curve at the wings is decaying somewhat faster than the experimental spectrum.

+

This is due to modeling the HFC of DMA ˙ as a set of equivalent spin ½ nuclei

giving a Gaussian contour, a situation for which a closed analytical solution is

+

available [101, 106], while the actual ESR spectrum of DMA ˙ noticeably departs

from Gaussian due to the presence of a spin 1 nucleus (N). The importance of

explicitly accounting for higher-spin nuclei to accurately reproduce the wings of

MARY spectra was earlier demonstrated for pairs comprising radical anions of

quinolinate complexes of metal cations with nuclear spins 3/2–5/2 that bear the

dominant HFC [115]. Another available model to describe spin evolution in MARY

spectra, the semiclassical description [70, 116–118], also represents the HFC as a

Gaussian contour and thus does not mend the situation with DMA radical cation,

while an explicit description of the hyperfine structure in low magnetic fields,

although undertaken several times for selected systems [103, 116, 119], drastically

complicates description without providing any new insight for the purpose of this

work, and therefore was not attempted.

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

Fraction of Spin-Correlated Pairsꢂ

ꢂ17

We also note that the presented OD ESR and MARY spectra have about an order

of magnitude better signal-to-noise ratio in comparison to spectra taken in identical

conditions for systems with similar intrinsic luminescence properties that do not

form exciplexes, which is a direct consequence of generating a new high-intensity

exciplex emission band that comprises all the magnetosensitive recombination

luminescence that for the discussed probe pair is the dominant emission band.

3.4 Time-resolved magnetic field effect for a pair comprising

FT radical anion

Luminescent properties of FT allow directly observing temporal spin state evolu-

tion of spin-correlated radical ion pairs comprising its radical anion to character-

ize the radical anion itself and to determine the fraction of spin-correlated pairs

among all pairs that recombine at the nanosecond time scale using the method of

Time-Resolved Magnetic Field Effect (TR MFE) [120, 121]. To this end we measured

the kinetics of recombination fluorescence decay for solutions of 2·10⁻³M FT in

n-dodecane and cyclohexane in zero field and in the applied field of Bꢀ=ꢀ0.1 T.

The second partner of the probe pair, DMA, in these experiments was not intro-

duced into the system, since then unavoidable formation of the recombination

exciplex with a long emission lifetime would render impossible observation of

any resolved quantum beats. On the other hand, the very short intrinsic emis-

sion lifetime of FT provides for excellent quantum beats in the pair with solvent

+

−

radical cation S ˙/FT ˙. This is demonstrated in Figure 7 that shows the result of

dividing the kinetics obtained in applied field 0.1 T by the corresponding kinetics

in zero field, i.e. the TR MFE curves.

Acceptor concentration, 2·10⁻³M, was chosen to rapidly, in a fraction of nano-

second, capture the excess electrons in n-dodecane by FT (typical rate constant

is about 2·10¹¹M⁻¹·s⁻¹[122]), while having the characteristic time of encounters

50–100 ns for species with molecular mobility, thus ensuring that the counter ion

−

+

for FT ˙ is solvent radical cation S ˙. ESR spectrum of the latter is narrowed due to

degenerate electron exchange with parent solvent molecules [123], therefore its

contribution to the observed beats is not so pronounced. In cyclohexane solu-

tion electron capture proceeds even faster, but due to faster solvent hole mobility

in a cyclic alkane [124–126] radical cation capture by FT is also possible, which

explains the larger width of the counter-ion that was introduced in modeling the

cyclohexane TR MFE curve in Figure 7B as compared to the n-dodecane curve of

Figure 7A.

The most revealing feature of the presented TR MFE curves is a pronounced

dip at approximately 18 ns, which is very typical for quantum beats in radical ion

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

18ꢀ

1.6

1.5

1.4

1.3

1.2

1.1

1.0

0.9

A

B

1.5

1.4

1.3

1.2

1.1

1.0

0.9

2

1

2

1

0

20

40

60

80

100 120 140

0

20

40

60

80

100 120 140

Time, ns

Fig. 7:ꢀ(A) Experimental (1) and calculated (2, shifted upwards by 0.2) TR MFE curves for

−

solution of 2·10⁻³M FT in n-dodecane. Modeling parameters: A(3F)ꢁ=ꢁ1.93 mT (FT ˙), Δꢁ=ꢁ0.15 mT

+

(S ˙), T₁ꢁ=ꢁ650 ns, T₂ꢁ=ꢁ16 ns, θꢁ=ꢁ0.28. Temperature 298 K, B₀ꢁ=ꢁ0.1 T. (B) Experimental (1) and

calculated (2, shifted upwards by 0.2) TR MFE curves for solution of 2·10⁻³M FT in cyclohexane.

−

+

Modeling parameters: A(3F)ꢁ=ꢁ1.93 mT (FT ˙), Δꢁ=ꢁ0.35 mT (S ˙), T₁ꢁ=ꢁ1200 ns, T₂ꢁ=ꢁ19 ns, θꢁ=ꢁ0.22.

Temperature 298 K, B₀ꢁ=ꢁ0.1 T.

pairs comprising partners, in which the unpaired electron interacts with an odd

number of equivalent nuclei in one partner and the ESR spectrum width of the

second partner is much less than of the first one [127]. Modeling demonstrates

that the shown case indeed requires three equivalent spin ½ nuclei with HFC

about 1.93 mT that have to be attributed to the radical anion of FT. A certain

discrepancy between peak amplitudes in the experimental and simulated curves

at early times is explained by a contribution of direct energy transfer from

excited solvent molecules to luminophore, as the corresponding luminescence

at times of units of nanoseconds is only weakly sensitive to magnetic field and

smears the beats at slightly later times. The coupling of about 2 mT provides

an optimal match with the time resolution of the experimental setup, allowing

to detect several periods of high-quality well-resolved beats, and radical anion

of 1-(phenylethynyl)-4-(trifluoromethyl)benzene will be a partner of choice in

further quantum beats studies. Earlier available examples of fluorinated radical

anions with equivalent nuclei, like hexafluorobenzene radical anion [96], had

too high a coupling and thus too fast spin evolution to directly resolve it in a TR

MFE experiment.

We finally note that the amplitudes of MFE in n-dodecane is somewhat higher

than in cyclohexane, and that in both solvents the curves decay with a character-

istic time of 600–700 ns, which can be considered as spin-lattice relaxation time

for the observed radical ion pair in the conditions of this experiment. The frac-

tions of spin-correlated pairs extracted from modeling of these curves and quali-

tatively given by the initial heights of the decaying tails of the TR MFE curves,

were found to be 0.28 for n-dodecane and 0.22 for cyclohexane, these values will

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

Fraction of Spin-Correlated Pairsꢂ

ꢂ19

be used in the next section to calibrate the values for the representative set of

solvents obtained from the magnetic field effect in the exciplex emission band

for the probe pair.

3.5 Magnetic field effect in the exciplex emission band in a set

of alkanes

Donor-acceptor pair 1-(phenylethynyl)-4-(trifluoromethyl)benzene – DMA

turned out to be a very convenient system to study the magnitude of magnetic

field effect in the exciplex emission band, and the fraction of spin-correlated

radical ion pairs that is directly linked to it, in various nonpolar solvents. First,

as has been demonstrated above, all exciplexes in this system form only as a

result of magnetosensitive radical ion pair recombination and do not appear

as a result of a sequence of not magnetically sensitive reactions, such as direct

energy transfer from an excited solvent molecule to any of the mixture compo-

nents with subsequent bulk reaction between the formed excited molecule and

its partner. Second, exciplex emission band is substantially shifted to the red

with respect to intrinsic luminescence of the individual mixture components,

with a very low intrinsic quantum yield of the “luminophore” itself, which sim-

plifies a reliable determination of the magnitude of magnetic field effect due

to pair recombination without emission spectrum decomposition into separate

components.

As an example of the probe application of the FT – DMA pair we determined

the magnitudes of magnetic field effect in the exciplex emission band (χ_E) in

11 alkanes with different properties that can be divided into three groups. The

first group includes n-alkanes of varied chain length: C₆, C₇, C₈, C₁₀, C₁₂, C₁₆, for

which viscosity, luminescence quantum yield and lifetime of the excited state

all increase from C₆to C₁₆[80, 81, 128]. Of relevance to radiation spin chemis-

try is also the increasing lifetime of solvent radical cation in neat liquids with

increasing chain length [129]. The second group comprises cyclohexane and

methylcyclohexane, which feature the so-called “fast solvent hole”, i.e. the rate

+

of reaction of the solvent radical cation S ˙ with an electron donor D to produce

+

D ˙ is an order of magnitude higher than the diffusion-controlled limit [124–126].

The third group consists of various branched alkanes, including non-luminesc-

ing isooctane (2,2,4-trimethylpentane) [80, 128], having high viscosity but fast

solvent hole squalane (2,6,10,15,19,23-hexamethyltetracosane) [130, 131] and a

typical vicinal alkane 2,3-dimethylbutane with moderate luminescence [80, 128].

Some physico-chemical and luminescent properties of the studied alkanes are

listed in Table 1.

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

20ꢀ

For each solvent five independent pairs of spectra of radiation-generated

recombination luminescence in applied field 20 mT, sufficient to saturate the

hyperfine-induced spin evolution in the probe radical ion pair (cf. its MARY spec-

trum in Figure 6B) and in zero field were experimentally obtained, from which

χ_Evalues were determined and statistically evaluated. Figure 8 shows one pair

of typical spectra for four representative solvents across all groups, the spectra

for all 11 solvents are given in the Supporting Information. The deviations to χ_E

values listed in the figure caption are the confidence interval (Pꢀ=ꢀ0.95) for a given

set of experimental spectra.

As above, to improve the signal-to-noise ratio when determining χ_Ethe exci-

plex emission bands in the spectra in applied field and in zero field were inte-

grated in wavelength range 415–475 nm that contains only the exciplex emission.

The magnitude of the magnetic field effect in the band of intrinsic DMA lumi-

nescence χ_DMAwas similarly determined by integrating in the range 315–360 nm.

Thus obtained χ_Eand χ_DMAvalues for all 11 solvents are listed in Table 2, the devia-

tions to the listed χ_Eand χ_DMAvalues are the confidence interval (Pꢀ=ꢀ0.95) for a

given set of experimental spectra.

A comparison of Tables 1 and 2 demonstrates that for linear alkanes an

increase in integral DMA emission S_DMAcorrelates with increasing luminescence

quantum yield (ϕ_f) and excited state lifetime (τ_f) of the solvent. This clearly shows

that generation of DMA excited states in this system proceeds mostly via direct

energy transfer from an excited solvent molecule, which grows more efficient

Tab. 1:ꢀPhysico-chemical and luminescent properties of the studied alkanes.

3

e

Solvent

ꢁ

η, cP^aꢁ μ·10⁴, cm²/(V·s)^bꢁ

τ_f, ns^cꢁ

ϕ_f¹⁴⁷⋅10 ꢁ

ϕ_f¹⁶⁵⋅10 ꢁ

λ_m, nm^f

d

n-hexane

ꢃ

0.31ꢃ

0.42ꢃ

0.55ꢃ

0.91ꢃ

1.50ꢃ

3.45ꢃ

0.98ꢃ

0.73ꢃ

0.51ꢃ

0.34ꢃ

35.4ꢃ

5.7ꢃ

–ꢃ

<0.7ꢃ

1.1ꢃ

1.5ꢃ

3.0ꢃ

4.2ꢃ

5.2ꢃ

1.1ꢃ

1.3ꢃ

<0.2ꢃ

1.4ꢃ

0.2ꢃ

0.7ꢃ

1.1ꢃ

2.0ꢃ

3.2ꢃ

4.4ꢃ

3.5ꢃ

5.5ꢃ

<0.01ꢃ

0.9ꢃ

–ꢃ

0.6ꢃ

1.6ꢃ

206

207

201

213

–

n-heptane

n-octane

3.8ꢃ

2.4ꢃ

1.7ꢃ

0.8ꢃ

1.9ꢃ

3.0ꢃ

3.6ꢃ

0.14ꢃ

2.5ꢃ

n-decane

4.2ꢃ

n-dodecane

n-hexadecane

cyclohexane

5.5ꢃ

7.6ꢃ

8.8ꢃ

methylcyclohexane ꢃ

isooctane

2,3-dimethylbutaneꢃ

squalane

11.2ꢃ

<0.01ꢃ

6.1ꢃ

ꢃ

242

–

ꢃ

1.6ꢃ

The table lists: ^aviscosity at 20ꢁ°C [132]; ^bmobility of tolane radical ions at 20ꢁ°C [133]; ^cexcited

state lifetime [81, 128] (squalane [130, 131]); ^dquantum yield of luminescence for excitation at

147 nm at 20ꢁ°C [80, 128]; ^equantum yield of luminescence for excitation at 165 nm at 20ꢁ°C [80,

128]; ^fwavelength of maximum emission [80].

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

Fraction of Spin-Correlated Pairsꢂ

ꢂ21

15

12

9

12

10

8

A

B

6

4

3

2

0

250

300

350

400

450

500

550

600

250

300

350

400

450

500

550

600

λ, nm

7

6

5

4

3

2

1

0

12

10

8

C

D

6

4

2

0

250

300

350

400

450

500

550

600

250

300

350

400

450

500

550

600

λ, nm

Fig. 8:ꢀSpectra of X-ray generated luminescence for a mixture of 6·10⁻³M 1-(phenylethynyl)-

4-(trifluoromethyl)benzene and 1·10⁻²M DMA in different alkanes in applied magnetic field

20 mT (●) and in zero field (○): (A) n-decane, χ_Eꢁ=ꢁ19.3ꢁ ꢁ0.4%; (B) cyclohexane, χ_Eꢁ=ꢁ17.3ꢁ ꢁ0.5%;

(C) 2,3-dimethylbutane, χ_Eꢁ=ꢁ20.3ꢁ ꢁ0.5%; (D) isooctane, χ_Eꢁ=ꢁ19.1ꢁ ꢁ0.5%. Magnetic field effect

was determined in the marked wavelength range.

with increasing the emission lifetime and quantum yield of the solvent. On the

other hand, as the viscosity of the linear alkane increases (by about an order of

magnitude upon transition from n-hexane to n-hexadecane) the integral exciplex

emission S_Eremains constant, which confirms exciplex formation via diffusion-

indifferent radical ion pair recombination. Furthermore, the magnitude of the

magnetic field effect in the exciplex emission band is also practically identical for

the entire set of the studied n-alkanes. The small χ_DMAin all alkanes under study

+

is due to partial recombination of secondary radical ion pairs D ˙/A⁻˙ into DMA

excited molecules. As was mentioned above generation of DMA excited states

is also possible via nonmagnetosensitive direct energy transfer from an excited

solvent molecule, that’s why the magnitude of χ_DMAis rather small compared to

χ_E, where all exciplexes are formed via recombination.

When cycloalkanes or squalane are used as the solvent the values χ_Edrop by

about 2–3% as compared to other alkanes which is a statistically reliable decrease.

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

22ꢀ

Tab. 2:ꢀMagnetic field effects for the 1-(phenylethynyl)-4-(trifluoromethyl)benzene – DMA pair

in 11 solvents.

Solvent

ꢁ

χ_E, %^aꢁ

χ_DMA, %^bꢁ

S_E, rel. u.^cꢁ

S_DMA, rel. u.^d

n-hexane

ꢃ

19.6ꢁ ꢁ0.7ꢃ

19.2ꢁ ꢁ0.2ꢃ

19.4ꢁ ꢁ0.7ꢃ

19.3ꢁ ꢁ0.4ꢃ

19.8ꢁ ꢁ0.6ꢃ

19.4ꢁ ꢁ0.6ꢃ

17.3ꢁ ꢁ0.5ꢃ

16.2ꢁ ꢁ0.6ꢃ

19.1ꢁ ꢁ0.5ꢃ

20.3ꢁ ꢁ0.5ꢃ

18.1ꢁ ꢁ0.6ꢃ

4.4ꢁ ꢁ0.7ꢃ

3.3ꢁ ꢁ0.5ꢃ

4.4ꢁ ꢁ0.9ꢃ

3.8ꢁ ꢁ0.6ꢃ

4.4ꢁ ꢁ0.8ꢃ

4.2ꢁ ꢁ0.6ꢃ

3.7ꢁ ꢁ0.4ꢃ

3.9ꢁ ꢁ0.9ꢃ

4ꢁ ꢁ3ꢃ

47ꢁ ꢁ3ꢃ

45ꢁ ꢁ6ꢃ

52ꢁ ꢁ5ꢃ

53ꢁ ꢁ4ꢃ

51ꢁ ꢁ5ꢃ

54ꢁ ꢁ5ꢃ

45ꢁ ꢁ5ꢃ

27ꢁ ꢁ3ꢃ

39ꢁ ꢁ3ꢃ

25ꢁ ꢁ3ꢃ

20ꢁ ꢁ3ꢃ

5.1ꢁ ꢁ0.5

7.6ꢁ ꢁ0.4

11ꢁ ꢁ2

n-heptane

n-octane

n-decane

16ꢁ ꢁ2

n-dodecane

n-hexadecane

cyclohexane

16ꢁ ꢁ2

19ꢁ ꢁ2

11ꢁ ꢁ1

methylcyclohexane ꢃ

isooctane

2,3-dimethylbutaneꢃ

squalane

9.9ꢁ ꢁ0.5

2.7ꢁ ꢁ0.3

3.3ꢁ ꢁ0.4

13ꢁ ꢁ1

ꢃ

3ꢁ ꢁ2ꢃ

ꢃ

5.8ꢁ ꢁ0.8ꢃ

The table lists: ^amagnetic field effect in exciplex emission band (415–475 nm); ^bmagnetic field

effect in intrinsic DMA emission band (315–360 nm); ^cabsolute area under the exciplex emission

band (415–475 nm); ^dabsolute area under the intrinsic DMA emission band (315–360 nm).

We also note that for branched alkanes, i.e. isooctane, 2,3-dimethylbutane,

squalane, and methylcyclohexane, the absolute area under the exciplex emission

band S_Eis substantially lower than for linear alkanes, i.e. the number of forming

exciplexes is smaller. A possible explanation is the known more efficient in com-

parison to n-alkanes formation of alkenes and products of C–C cleavage next to

the branching site in the solvent radical cation, which reduces the number of

solvent radical cations available for capturing, and thus the number of secondary

+

radical ion pairs D ˙/A⁻˙ and the recombination exciplexes. Formation of alkenes as

the products of alkane photolysis is also the quoted reason for different lumines-

cence yields at different excitation wavelengths, cf. Table 1 [80]. The shorter wave-

length excitation (147 nm instead of 165 nm) results in a more efficient photolysis

and thus a higher concentration of photolysis products, which in turn effectively

quench emission thus reducing the quantum yield of luminescence. As Table 1

demonstrates, the largest difference in quantum yields of luminescence for exci-

tation at 147 nm and 165 nm is indeed observed for branched alkanes, which indi-

rectly supports a more efficient decomposition of radical cations for branched

alkanes in the conditions of X-irradiation. The work [128] reports quantum yields

of photolysis products for isooctane, which are 0.08, 0.17, 0.36, and 0.05 for H₂,

CH₄, C₄H₁₀, and C₅H₁₀, respectively. It is the efficient decomposition of an excited

isooctane molecule that the authors attribute the very low emission yield of this

alkane to. In a similar fashion, in ref. [131] the efficient decomposition of excited

states of branched alkanes was called upon to rationalize the shortening of the

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

Fraction of Spin-Correlated Pairsꢂ

ꢂ23

excited state lifetime of, e.g. squalane (τ_fꢀ=ꢀ1.4 ns) in comparison to linear alkanes

of comparable chain length, such as n-hexadecane (τ_fꢀ=ꢀ5.2 ns). The possibility of

efficient formation of alkenes and C–C cleavage products upon X-irradiation of

isooctane was also discussed in refs. [134, 135]. Also radical cation of tetramethy-

lethylene was identified to be the product of unimolecular decay of 2,3-dimeth-

ylbutane radical cation in low-temperature freon matrices [136, 137]. In ref. [138]

one can find comprehensive investigation of branched alkanes cation radicals by

means of TR MFE and their lifetime estimations.

The connection between the magnitude of the observed magnetic field effect

and the fraction of spin-correlated radical ion pairs (4) can be obtained from

expression (1) of ref. [139] by using the definition of θ given in ref. [134]:

χ_E

θ =

,

(4)

1 −(4ρ −1)⋅(χ_E+1)

0

where θ is the fraction of spin-correlated pairs, χ_Eis the magnitude of the magnetic

field effects (in the studied system 1-(phenylethynyl)-4-(trifluoromethyl)benzene

– DMA we determined it in the exciplex emission band), and ρ₀is the equilibrium

singlet state population of the pair in zero magnetic field. The value of ρ₀depends

on the situation, its limiting value of 1/4 corresponds to spin evolution completely

contained within a single radical ion pair without any operating charge trans-

fer processes [140]. Any processes of charge transfer between a radical ion and a

neutral molecule accompanying the process of recombination result in increased

+

ρ₀. For a single charge transfer upon transition of pair S ˙/A⁻˙ into pair D ˙/A⁻˙ by

+

capturing solvent hole S ˙ to a positive charge acceptor D the value of ρ₀is equal

to 5/18 [141]. Multiple charge transfer, typical for processes of degenerate charge

exchange in concentrated solutions, result in the ρ₀value of 1/3 [116]. Using these

benchmark values of ρ₀and the experimentally determined χ_Evalues, we can use

expression (4) to obtain the fractions θ in the studied alkanes. For all alkanes

except for cyclohexane, methylcyclohexane and squalane θ were calculated to

be approximately 0.19ꢀ ꢀ0.01, 0.22ꢀ ꢀ0.01 and 0.32ꢀ ꢀ0.01 for ρ₀of 1/4, 5/18 and 1/3,

respectively. For the three outliers that all share in common the presence of fast

solvent hole the values of θ are somewhat lower at 0.17ꢀ ꢀ0.01, 0.20ꢀ ꢀ0.01 and

0.28ꢀ ꢀ0.01, respectively. It is interesting to note that the values for isooctane do

not significantly deviate from the values for linear alkanes, although it is known

that the range of electron thermalization upon ionization for such “spherical”

molecules as isooctane and neopentane is significantly larger than for regular

n-alkanes [142].

The estimated θ values for a representative set of alkanes reasonably agree

with available published data obtained by other techniques: TR MFE (a relatively

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

24ꢀ

sophisticated experimental technique, typical θ values 0.2–0.4 [139, 143]), CW

magnetic field effect for non-exciplex forming systems usually comprising a mol-

ecule that simultaneously captures both electron and solvent hole (a substan-

tial contribution to emission from non-magnetosensitive channels of excitation

generation that lowers magnetic field effect, typical θ values 0.1–0.3 [143–145]),

analysis of Δg quantum beats in a special probe pair p-terphenyl-d₁₄– diphenyl-

sulphide-d₁₀with substantially shifted g-values of the two pair radical ions [134,

146]. The latter technique is not sensitive to non-recombination processes and

thus provides seemingly more reliable θ values, but it is directly applicable to

a limited range of solvents that do not imprint substantial spin evolution in the

solvent radical cation prior to its capture by the sulphur-centred radical cation,

e.g. it yields θꢀ=ꢀ0.26 for cyclohexane that satisfies this requirement, but θꢀ=ꢀ0.19

for isooctane that has substantial hyperfine couplings in its radical cation that

poorly average out by degenerate charge transfer [147, 148].

The estimated θ values should also be compared to θꢀ=ꢀ0.28 for n-dodecane

and 0.22 for cyclohexane, as determined in the previous section from TR MFE

experiments in dilute solutions of FT in alkanes without introducing the second

pair partner. It can be seen that these reference values fit well between the esti-

mates of θ calculated for the situations of single and multiple charge transfer,

which both are certainly present in the systems used to obtain the quoted χ_E

values, reproducing the lower value for the cyclic alkane. We can therefore use

the value of ρ₀between 5/18 and 1/3 to estimate the value of θ from the measured

values of χ_E, and can directly compare thus obtained estimates of θ for relative

magnitudes and trends without the necessity to consider additional complica-

tions due to limited applicability of the more advanced techniques.

4 Conclusions and outlook

The suggested method to estimate the fraction of spin-correlated radical ion pairs

θ by measuring the magnitude of magnetic field elect in the exciplex emission

band in a donor-acceptor system with optimal luminescent properties of the

forming excited complex is a relatively simple experimental technique applica-

ble in any solvent, in which excited states can be generated via recombination

of a radical ion pair, and in arbitrary concentrated solutions. On the other hand,

the knowledge of θ is needed for correct interpretation of many experiments in

radiation spin chemistry and for linking their results with calculated spin evolu-

tion in the radical ion pair. An important development of this approach is estima-

tion of the fraction of spin-correlated pairs, and in a wider sense investigation

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

Fraction of Spin-Correlated Pairsꢂ

ꢂ25

of spin-correlated radical pairs in other condensed systems as well, including

polymer matrices. While in alkane solutions all obtained values of θ turned out

to be rather close to each other and to the expected values, what would happen

in polymers is not known. In this work we applied our approach to a set of sol-

vents with viscosity differing by as much as two orders of magnitude (n-hexane

to squalane), and the latter is already close to a model of an aliphatic polymer

like polypropylene or polyisobutylene. Furthermore, as discussed in our earlier

papers, radiation-induced recombination luminescence is a convenient model

for electroluminescence in molecular systems, which forms the basis of modern

all-organic optoelectronic devices, and exciplex formation can be used for spec-

tral tuning of emission and improving optoelectric conversion efficiency of such

devices by avoiding self-absorption within the device.

Acknowledgments: The work was supported by Russian Science Foundation

(project no. 16-13-10163). A.R.M. is grateful to the Council for Grants of the Presi-

dent of the Russian Federation for awarding a personal scholarship for Support

of Young Researchers (project no. SP-81.2016.1).

References

1. G. R. Eaton, S. S. Eaton, K. M. Salikhov, Foundations of Modern EPR. World Scientiﬁc (1998).

2. V. I. Feldman, Radiat. Phys. Chem. 55 (1999) 565.

3. A. D. Milov, K. M. Salikhov, M. D. Shirov, Sov. Phys. Solid State 23 (1981) 565.

4. A. D. Milov, K. M. Salikhov, Y. D. Tsvetkov, Chem. Phys. Lett. 8 (1971) 523.

5. A. A. Obynochny, A. G. Maryasov, P. A. Purtov, K. M. Salikhov, Appl. Magn. Reson. 15 (1998)

259.

6. W. K. Subczynski, J. S. Hyde, Biophys. J. 45 (1984) 743.

7. K. Toriyama, M. Iwasaki, K. Nunome, H. Muto, J. Chem. Phys. 75 (1981) 1633.

8. J. H. Ardenkjaer-Larsen, I. Laursen, I. Leunbach, G. Ehnholm, L.-G. Wistrand, J. S. Petersson,

K. Golman, J. Magn. Reson. 133 (1998) 1.

9. H. M. Swartz, R. B. Clarkson, Phys. Med. Biol. 43 (1998) 1957.

10. Y. Iwasaki, T. Osasa, M. Asahi, M. Matsumura, Y. Sakaguchi, T. Suzuki, Phys. Rev. B. 74

(2006) 195209.

11. V. I. Krinichnyi, E. I. Yudanova, J. Phys. Chem. C. 116 (2012) 9189.

12. A. Schnegg, J. Behrends, M. Fehr, K. Lips, Phys. Chem. Chem. Phys. 14 (2012) 14418.

13. Y. Sheng, T. D. Nguyen, G. Veeraraghavan, Ö. Mermer, M. Wohlgenannt, S. Qiu, U. Scherf,

Phys. Rev. B. 74 (2006) 045213.

14. E. G. Bagryanskaya, Y. A. Grishin, R. Z. Sagdeev, Y. N. Molin, Chem. Phys. Lett. 114 (1985) 138.

15. E. L. Frankevich, M. M. Tribel, I. A. Sokolik, A. I. Pristupa, Phys. Status Solidi B 87 (1978) 373.

16. Y. N. Molin, O. A. Anisimov, V. M. Grigor’yants, V. K. Molchanov, K. M. Salikhov, J. Phys.

Chem. 84 (1980) 1853.

17. A. D. Trifunac, J. P. Smith, Chem. Phys. Lett. 73 (1980) 94.

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

26ꢀ

18. K. M. Salikhov, Y. N. Molin, R. Z. Sagdeev, A. L. Buchachenko, Spin Polarization and Magne-

tic Eﬀects in Radical Reactions, Netherlands, Elsevier (1984).

19. A. J. Hoﬀ, Phys. Rep. 54 (1979) 75.

20. K. M. Salikhov, A. J. Van der Est, D. Stehlik, Appl. Magn. Reson. 16 (1999) 101.

21. M. C. Thurnauer, Rev. Chem. Intermed. 3 (1979) 197.

22. S. G. Zech, W. Lubitz, R. Bittl, Ber. Bunsen-Ges. Phys. Chem. 100 (1996) 2041.

23. K. M. Salikhov, J. H. Golbeck, D. Stehlik, Appl. Magn. Reson. 31 (2007) 237.

24. V. I. Borovkov, I. S. Ivanishko, V. A. Bagryansky, Y. N. Molin, J. Phys. Chem. A. 117 (2013)

1692.

25. K. L. Ivanov, M. V. Petrova, N. N. Lukzen, K. Maeda, J. Phys. Chem. A. 114 (2010) 9447.

26. J. A. Jones, K. Maeda, P. J. Hore, Chem. Phys. Lett. 507 (2011) 269.

27. P. A. Purtov, Chem. Phys. Lett. 496 (2010) 335.

28. A. I. Kruppa, T. V. Leshina, R. Z. Sagdeev, K. M. Salikhov, F. S. Sarvarov, Chem. Phys. 67

(1982) 27.

29. S. Richert, A. Rosspeintner, S. Landgraf, G. Grampp, E. Vauthey, D. R. Kattnig, J. Am. Chem.

Soc. 135 (2013) 15144.

30. O. A. Anisimov, Ion Pairs in Liquids, (Eds. A. Lund, M. Shiotani), Netherlands, Springer,

(1991), P. 285.

31. I. A. Shkrob, A. D. Trifunac, Radiat. Phys. Chem. 50 (1997) 227.

32. J. Klein, R. Voltz, Phys. Rev. Lett. 36 (1976) 1214.

33. Y. N. Molin, Bull. Korean Chem. Soc. 20 (1999) 7.

34. E. V. Kalneus, D. V. Stass, Y. N. Molin, Appl. Magn. Reson. 28 (2005) 213.

35. W. M. Bartczak, A. Hummel, J. Chem. Phys. 87 (1987) 5222.

36. V. V. Lozovoy, S. V. Anishchik, N. N. Medvedev, O. A. Anisimov, Y. N. Molin, Chem. Phys. Lett.

167 (1990) 122.

37. M. Wojcik, M. Witold, Bartczak, J. Kroh, Chem. Phys. Lett. 177 (1991) 184.

38. V. I. Borovkov, K. A. Velizhanin, Radiat. Phys. Chem. 76 (2007) 988.

39. V. I. Borovkov, K. A. Velizhanin, Radiat. Phys. Chem. 76 (2007) 998

40. A. R. Melnikov, E. V. Kalneus, V. V. Korolev, I. G. Dranov, A. I. Kruppa, D. V. Stass, Photochem.

Photobiol. Sci. 13 (2014) 1169.

41. A. R. Melnikov, E. V. Kalneus, V. V. Korolev, I. G. Dranov, D. V. Stass, Dokl. Phys. Chem. 452

(2013) 257.

42. J. B. Birks, Nature 214 (1967) 1187.

43. E. A. Chandross, C. J. Dempster, J. Am. Chem. Soc. 92 (1970) 3586.

44. D. R. Kattnig, A. Rosspeintner, G. Grampp, Phys. Chem. Chem. Phys. 13 (2011) 3446.

45. N. L. Lavrik, J. Mol. Struct. 115 (1984) 477.

46. Y. H. Lee, M. Lee, Bull. Korean Chem. Soc. 18 (1997) 1054.

47. P. P. Levin, V. A. Kuz’min, B. Acad. Sci. USSR Ch. 35 (1986) 1303.

48. N. K. Petrov, V. N. Borisenko, M. V. Alﬁmov, T. Fiebig, H. Staerk, J. Phys. Chem. 100 (1996)

6368.

49. M. A. F. Tavares, Trans. Faraday Soc. 66 (1970) 2431.

50. M. Ottolenghi, Acc. Chem. Res. 6 (1973) 153.

51. A. J. Bard (Ed.), Electrogenerated Chemiluminescence, New York, Marcel Dekker Inc. (2004).

52. J. B. Ketter, R. M. Wightman, J. Am. Chem. Soc. 126 (2004) 10183.

53. A. Weller, K. Zachariasse, Chem. Phys. Lett. 10 (1971) 590.

54. V. I. Borovkov, I. S. Ivanishko, Radiat. Phys. Chem. 80 (2011) 540.

55. Y. Hirata, T. Okada, N. Mataga, T. Nomoto, J. Phys. Chem. 96 (1992) 6559.

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

Fraction of Spin-Correlated Pairsꢂ

ꢂ27

56. C. Ferrante, U. Kensy, B. Dick, J. Phys. Chem. 97 (1993) 13457.

57. K. Bhattacharyya, M. Chowdhury, Chem. Rev. 93 (1993) 507.

58. P. P. Levin, P. F. Pluzhnikov, V. A. Kuzmin, Chem. Phys. Lett. 147 (1988) 283.

59. N. Mataga, H. Miyasaka, Electron Transfer and Exciplex Chemistry, in: Advances in Chemical

Physics: Electron Transfer from Isolated Molecules to Biomolecules, Vol. 107 (Part 2).,

(Eds. J. Jortner, M. Bixon), New York, Wiley (1999) 431.

60. D. N. Nath, M. Chowdhury, Chem. Phys. Lett. 109 (1984) 13.

61. N. K. Petrov, A. I. Shushin, E. L. Frankevich, Chem. Phys. Lett. 82 (1981) 339.

62. H. J. Werner, Z. Schulten, K. Schulten, J. Chem. Phys. 67 (1977) 646.

63. S. G. Fedorenko, A. I. Burshtein, J. Chem. Phys. 141 (2014) 114504.

64. S. P. Van, G. S. Hammond, J. Am. Chem. Soc. 100 (1978) 3895.

65. S. T. Cheung, W. R. Ware, J. Phys. Chem. 87 (1983) 466.

66. K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 16 (1975) 4467.

67. R. D. Stephens, C. E. Castro, J. Org. Chem. 28 (1963) 3313.

68. E. V. Kalneus, A. R. Melnikov, V. V. Korolev, V. I. Ivannikov, D. V. Stass, Appl. Magn. Reson. 44

(2013) 81.

69. Y. N. Molin, O. A. Anisimov, Radiat. Phys. Chem. 21 (1983) 77.

70. D. V. Stass, N. N. Lukzen, B. M. Tadjikov, Y. N. Molin, Chem. Phys. Lett. 233 (1995) 444.

71. S. V. Anishchik, V. M. Grigoryants, I. V. Shebolaev, Y. D. Chernousov, O. A. Anisimov, Y. N.

Molin, Instrum. Exp. Tech. 32 (1990) 813.

72. V. I. Borovkov, S. V. Anishchik, O. A. Anisimov, Chem. Phys. Lett. 270 (1997) 327.

73. K. G. Thakur, G. Sekar, Synthesis 16 (2009) 2785.

74. J. Saltiel, V. R. Kumar, J. Phys. Chem. A. 116 (2012) 10548.

75. M. Wierzbicka, I. Bylinska, C. Czaplewski, W. Wiczk, RSC Adv. 5 (2015) 29294.

76. Y. Hirata, T. Okada, T. Nomoto, Chem. Phys. Lett. 209 (1993) 397.

77. K. Okuyama, T. Hasegawa, M. Ito, N. Mikami, J. Phys. Chem. 88 (1984) 1711.

78. A. M. Brouwer, Pure Appl. Chem. 83 (2011) 2213.

79. I. B. Berlman, Handbook of Fluorescence Spectra of Aromatic Molecules. New York, Acade-

mic Press (1971).

80. W. Rothman, F. Hirayama, S. Lipsky, J. Phys. Chem. 58 (1973) 1300.

81. K. Shinsaka, H. Koizumi, T. Yoshimi, N. Kouchi, Y. Nakamura, M. Toriumi, M. Morita, Y. Hata-

no, S. Asaoka, H. Nishimura, J. Phys. Chem. 83 (1985) 4405.

82. A. R. Melnikov, E. V. Kalneus, V. V. Korolev, P. S. Sherin, V. I. Borovkov, D. V. Stass, Photo-

chem. Photobiol. Sci. 15 (2016) 767.

83. D. W. Tweeten, K. Lee, S. Lipsky, Int. J. Radiat. Appl. Instrum., Part C. Radiat. Phys. Chem. 34

(1989) 771.

84. H. T. Choi, J. A. Haglund, S. Lipsky, J. Phys. Chem. 87 (1983) 1583.

85. C. D. Jonah, M. C. Sauer Jr, Int. J. Rad. Appl. Instrum. Part C. Radiat. Phys. Chem. 34 (1989) 497.

86. J. M. Warman, K. Asmus, R. H. Schuler, J. Phys. Chem. 73 (1969) 931.

87. V. O. Saik, N. N. Lukzen, V. M. Grigoryants, O. A. Anisimov, A. B. Doktorov, Y. N. Molin,

Chem. Phys. 84 (1984) 421.

88. V. A. Bagryansky, O. M. Usov, N. N. Lukzen, Y. N. Molin, Appl. Magn. Reson. 12 (1997) 505.

89. K. M. Salikhov, Y. N. Molin, J. Phys. Chem. 97 (1993) 13259.

90. K. M. Salikhov, Y. Sakaguchi, H. Hayashi, Chem. Phys. 220 (1997) 355.

91. B. M. Tadjikov, A. V. Astashkin, Y. Sakaguchi, Chem. Phys. Lett. 284 (1998) 214.

92. A. V. Koptyug, V. O. Saik, O. A. Animisov, Y. N. Molin, Chem. Phys. 138 (1989) 173.

93. L. R. Mursalimov, K. Salikhov, Appl. Magn. Reson. 21 (2001) 223.

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢀꢀꢀꢁ

ꢀA.R. Melnikov et al.

28ꢀ

94. Y. E. Kandrashkin, K. M. Salikhov, D. Stehlik, Appl. Magn. Reson. 12 (1997) 141.

95. Y. E. Kandrashkin, K. M. Salikhov, A. Van der Est, D. Stehlik, Appl. Magn. Reson. 15 (1998) 417.

96. V. V. Lozovoy, V. M. Grigoryants, O. A. Anisimov, Y. N. Molin, P. V. Schastnev,

L. N. Shchegoleva, I. I. Bilkis, V. D. Shteingarts, Chem. Phys. 112 (1987) 463.

97. D. R. Duling, J. Magn. Reson. B. 104 (1994) 105.

98. K. Pal, G. Grampp, D. R. Kattnig, ChemPhysChem. 14 (2013) 3389.

99. Y. N. Molin, K. M. Salikhov, K. I. Zamaraev, Spin Exchange. Principles and Applications in

Chemistry and Biology. Berlin, Springer-Verlag (1980).

100.

D. V. Stass, N. N. Lukzen, B. M. Tadjikov, V. M. Grigoryantz, Y. N. Molin, Chem. Phys.

Lett. 243 (1995) 533.

101. O. A. Anisimov, V. M. Grigoryants, S. V. Kiyanov, K. M. Salikhov, S. A. Sukhenko,

Y. N. Molin, Theor. Exp. Chem. 18 (1983) 256.

102. V. N. Verkhovlyuk, S. V. Anishchik, Y. N. Molin, O. A. Anisimov, Mol. Phys. 113 (2015) 113.

103. M. Justinek, G. Grampp, S. Landgraf, P. J. Hore, N. N. Lukzen, J. Am. Chem. Soc. 126 (2004)

5635.

104. K. L. Ivanov, D. V. Stass, E. V. Kalneus, R. Kaptein, N. N. Lukzen, Appl. Magn. Reson. 44

(2013) 217.

105. S. A. Sukhenko, P. A. Purtov, K. M. Salikhov, Sov. J. Chem. Phys. 2 (1985) 29.

106. D. V. Stass, B. M. Tadjikov, Y. N. Molin, Chem. Phys. Lett. 235 (1995) 511.

107. V. O. Saik, A. E. Ostaﬁn, S. Lipsky, J. Chem. Phys. 103 (1995) 7347.

108. B. M. Tadjikov, D. V. Stass, Y. N. Molin, Chem. Phys. Lett. 260 (1996) 529.

109. B. M. Tadjikov, D. V. Stass, Y. N. Molin, J. Phys. Chem. A. 101 (1997) 377.

110. B. M. Tadjikov, D. V. Stass, O. M. Usov, Y. N. Molin, Chem. Phys. Lett. 273 (1997) 25.

111. V. M. Grigoryants, S. D. McGrane, S. Lipsky, J. Chem. Phys. 109 (1998) 7354.

112. E. V. Kalneus, A. A. Kipriyanov, P. A. Purtov, D. V. Stass, Y. N. Molin, Appl. Magn. Reson. 30

(2006) 549.

113. E. V. Kalneus, D. V. Stass, K. L. Ivanov, Y. N. Molin, Mol. Phys. 104 (2006) 1751.

114. E. V. Kalneus, A. A. Kipriyanov, P. A. Purtov, D. V. Stass, Y. N. Molin, Dokl. Phys. Chem. 415

(2007) 170.

115. N. V. Sergey, V. N. Verkhovlyuk, E. V. Kalneus, V. V. Korolev, A. R. Melnikov, A. B. Burdukov,

D. V. Stass, Y. N. Molin, Chem. Phys. Lett. 552 (2012) 32.

116. K. Schulten, P. G. Wolynes, J. Chem. Phys. 68 (1978) 3292.

117. N. E. Polyakov, P. A. Purtov, T. V. Leshina, M. B. Taraban, R. Z. Sagdeev, K. M. Salikhov,

Chem. Phys. Lett. 129 (1986) 357.

118. P. A. Purtov, K. M. Salikhov, Theor. Exp. Chem. 23 (1987) 192.

119. V. N. Verkhovlyuk, D. V. Stass, N. N. Lukzen, Y. N. Molin, Chem. Phys. Lett. 413 (2005) 71.

120. V. A. Bagryansky, V. I. Borovkov, Y. N. Molin, Russ. Chem. Rev. 76 (2007) 493.

121. V. I. Borovkov, D. V. Stass, V. A. Bagryansky, Y. N. Molin, Study of Spin-Correlated Radical

Ion Pairs in Irradiated Solutions by Optically Detected Epr and Related Techniques, (Eds. A.

Lund, M. Shiotani), Springer International Publishing (2014), P. 629.

122. V. I. Borovkov, Radiat. Phys. Chem. 77 (2008) 1190.

123. V. I. Borovkov, V. A. Bagryansky, I. V. Yeletskikh, Y. N. Molin, Mol. Phys. 100 (2002) 1379.

124. S. J. Rzad, R. H. Schuler, A. Hummel, J. Chem. Phys. 51 (1969) 1369.

125. J. M. Warman, P. P. Infelta, M. P. De Haas, A. Hummel, Chem. Phys. Lett. 43 (1976) 321.

126. E. Zador, J. M. Warman, A. Hummel, Chem. Phys. Lett. 23 (1973) 363.

127. V. A. Bagryansky, O. M. Usov, V. I. Borovkov, T. V. Kobzeva, Y. N. Molin, Chem. Phys. 255

(2000) 237.

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

ꢁꢀꢀꢀ

Fraction of Spin-Correlated Pairsꢂ

ꢂ29

128. R. Hermann, R. Mehnert, L. Wojnárovits, J. Lumin. 33 (1985) 69.

129. F. B. Sviridenko, D. V. Stass, Y. N. Molin, Chem. Phys. Lett. 297 (1998) 343.

130. Y. Katsumura, Y. Yoshida, S. Tagawa, Y. Tabata, Radiat. Phys. Chem. 21 (1983) 103.

131. I. A. Shkrob, M. C. Sauer, A. D. Trifunac, J. Phys. Chem. 100 (1996) 5993.

132. V. M. Tatevski (Ed.): Physical-Chemical Properies of Individual Hydrocarbons, Gostoptekhi-

zdat, Moscow (1960) (in Russian).

133. V. I. Borovkov, S. V. Anishchik, O. A. Anisimov, Radiat. Phys. Chem. 67 (2003) 639.

134. O. M. Usov, V. M. Grigoryants, B. M. Tadjikov, Y. N. Molin, Radiat. Phys. Chem. 49 (1997)

237.

135. D. W. Werst, M. F. Desrosiers, A. D. Trifunac, Chem. Phys. Lett. 133 (1987) 201.

136. T. Miyazaki, H. Tsuruta, Y. Fujitani, K. Fueki, J. Phys. Chem. 86 (1982) 970.

137. Y. Saitake, T. Miyazaki, Z. I. Kuri, J. Phys. Chem. 77 (1973) 2418.

138. V. I. Borovkov, P. A. Potashov, L. N. Shchegoleva, V. A. Bagryansky, Y. N. Molin, J. Phys.

Chem. A. 111 (2007) 5839.

139. J. A. LaVerne, B. Brocklehurst, J. Phys. Chem. 100 (1996) 1682.

140. E. W. Knapp, K. Schulten, J. Chem. Phys. 71 (1979) 1878.

141. W. R. S. Appleton, B. Brocklehurst, Chem. Phys. Lett. 136 (1987) 199.

142. W. F. Schmidt, Can. J. Chem. 55 (1977) 2197.

143. B. Brocklehurst, Int. Rev. Phys. Chem. 4 (1985) 279.

144. R. S. Dixon, V. J. Lopata, Can. J. Chem. 57 (1979) 3023.

145. R. S. Dixon, F. P. Sargent, V. J. Lopata, E. M. Gardy, B. Brocklehurst, Can. J. Chem. 55 (1977)

2093.

146. S. V. Anishchik, O. M. Usov, O. A. Anisimov, Y. N. Molin, Radiat. Phys. Chem. 51 (1998) 31.

147. S. V. Anishchik, V. I. Borovkov, V. I. Ivannikov, I. V. Shebolaev, Y. D. Chernousov, N. N. Luk-

zen, O. A. Anisimov, Y. N. Molin, Chem. Phys. 242 (1999) 319.

148. V. N. Verkhovlyuk, V. A. Morozov, D. V. Stass, A. B. Doktorov, Y. N. Molin, Chem. Phys. Lett.

378 (2003) 567.

Supplemental Material: The online version of this article (DOI: 10.1515/zpch-2016-0819) offers

supplementary material, available to authorized users.

- 10.1515/zpch-2016-0819

Downloaded from De Gruyter Online at 09/29/2016 11:55:14AM

via Cornell University Library

Article Doi

Estimation of the Fraction of Spin-Correlated Radical Ion Pairs in Irradiated Alkanes using Magnetosensitive Recombination Luminescence from Exciplexes Generated upon Recombination of a Probe Pair

DOI: 10.1515/zpch-2016-0819

Source and publish data:

Authors:

Article abstract of DOI:10.1515/zpch-2016-0819

Full text of DOI:10.1515/zpch-2016-0819

Products guided by the article

R&D Labs maybe for 370-99-0

Relevant to this article

Hot Product