Hatano Y., Katsumura Y., Mozumder A. (Eds.) Charged Particle and Photon Interactions with Matter - Recent Advances, Applications, and Interfaces

Подождите немного. Документ загружается.

680 Charged Particle and Photon Interactions with Matter

In deaerated solutions containing triethylamine (TEA), Bz

•+

is scavenged by TEA and the excess

electrons

react with polysilane molecules (PS) according to

Bz TEA Bz TEA

e PS PS

• •

•

+ +

− −

+  →

[ ]…

(25.7)

giving polysilane anion radicals (PS

•+

) without contribution from counter cations. The excited

states and excess electrons are rapidly scavenged in oxygen-saturated solutions ([O

] = 11.9mM

at 1atm, 25°C), giving singlet oxygen molecules (

) and oxygen anions (

−

) (Candeias et al.,

2000; Kawaguchi et al., 2003). PS in the solution react with Bz

•+

via

•

, yielding polymer cation

radicals(PS

•−

Bz PS Bz PS

• •

+ +

+  → +

(25.8)

Tagawa and Seki have already discussed the transient spectra of radical ions of polysilanes, including

PH1,

PD6, etc. (Ban et al., 1987; Seki et al., 1998, 1999a, 2000, 2001a, b, 2002, 2004a; Kawaguchi

et al., 2003). PS

•−

and PS

•+

were found to exhibit two absorption bands in the near-UV (350–400 nm)

and infrared (IR) (>1600nm) regions. The transient spectra of PS

•−

and PS

•+

suggest the presence

of an interband level occupied by an excess electron (IBL

−

) or by a hole (IBL

). The UV and IR

bands in PS

•−

are due to the transition from VB to IBL

−

, and from IBL

−

to the conduction band

(CB), respectively. This is also the case of PS

•+

, attributing the UV and IR bands to transition from

IBL

to CB, and from VB to IBL

, respectively. The presence of IBL

−

or IBL

is interpreted well

as a delocalized negative (positive) polaron state (Rice and Phillpot, 1987; Abkowitz et al., 1990;

Tachibana et al., 1990; Seki et al., 1999a, 2004a) and/or the Anderson localization state on a Si seg-

ment (Ichikawa et al., 1999a,b). Figure 25.4 shows a series of transient absorption spectra of anion

radicals observed in PH1-12 as examples of transient absorption spectra. The spectra observed for

anion radicals of MEn and PDn are almost identical, with only a small dependence on the chain

length of n-alkyl substituents. In contrast, PHn, in Figure 25.4, exhibits dramatic changes in the

IR band with a considerable red shift upon elongation of the n-alkyl chains. The IR band of cation

radicals of PHn also tends to shift dramatically with an elongation of n-alkyl substituents, despite a

negligibly small change observed for the IR bands of MEn and PDn anion radicals [55]. The peak

energy of the IR band in relation to the binding energy of negative or positive polarons on the Si

chains

(Pitt, 1977; Abkowitz et al., 1987) is given by

δε

≈

























∆V

(25.9)

where

δε

denotes the binding energy of a polaron

a denotes a lattice unit of a trans-chain segment

is the polaron width

V is the matrix element describing the interaction between two atomic orbitals consisting of a

covalent

bond

denotes the matrix element between two atomic orbitals of a Si atom

The

parameter A is given by

A V l

≡ −









( )

sin

∆

(25.10)

Nanoscale Charge Dynamics and Nanostructure Formation in Polymers 681

where 2θ is the tetrahedral bond angle. Thus, Δ is a parameter specifying the degree of delo-

calization of σ electrons on a σ-conjugated segment, while V describes the localization of a pair

of electrons in a local bond. The relative polaron width on the polymers can be estimated based

on Equation 25.8. With the previously reported values of Δ in poly(dimethylsilane) (Kumada and

Tamao, 1968; Pollard and Lucovsky, 1982) and V in poly(dichlorosilane) (Koe et al., 1998), the value

of (ΔV/2A) can be estimated to be ca. 1.6eV. The transition energy of the IR band observed for PH2

cation radicals is 0.77eV, the highest example value for all the polysilanes, yielding ξ

= ∼2 Si repeat-

ing units as the minimum value. The considerable red shift of the IR band observed for PHn induces

a dramatic increase in the polaron width, and hence a highly delocalized negative (anion radicals) or

positive

(cation radicals) charge state.

Figure

25.5 shows the transient UV band observed for the ME3 solution. The spectra appear to

suggest the simultaneous formation of the radical anion and cation in the Ar-saturated solution, lead-

ing to overlap of the two transient absorption bands at 358 and 348nm, which have been determined

independently (Seki et al., 2001b, 2004a; Kawaguchi et al., 2003). In contrast to the Ar-saturated

solution, only the radical anion or the radical cation is selectively formed in the Ar-saturated solu-

tion with 20mM TEA or in the O

-saturated solution. The kinetic trace is recorded over a wide

range of observation time from a few ns to μs. The extinction coefcients of anion radicals (ε

•−

)

were

∼ determined by using the electron-transfer reaction between PS and pyrene (Py) as follows:

PS Py PS Py

• •

− −

+ → + (25.11)

The kinetic traces of PS

•−

and Py

•−

(ε

•−

of Py = 5.0 × 10

−1

at 492nm (Gill et al., 1964)) gave

the ε

•−

values of polysilanes, as summarized in Table 25.1. The extinction coefcient of the cation

Wavelength (nm)

0.4

0.8

0.4

0.8

(d)

(e)

(a)

Optical density

0.4

0.8

0.4

0.8

0.4

0.8

(b)

(c)

Wavelength (nm)

1800200

14001000600

1800200 14001000600

•

Figure 25.4 Transient absorption spectra of anion radicals of (a) PH1, (b) PH2, (c) PH4, (d) PH6, and

(e)PH8, in Bz at 5.0 × 10

−3

mol dm

−3

(base mol unit) with TEA at RT = 2.0 × 10

−2

mol dm

−3

conc. Spectra were

recorded

after electron pulse irradiation.

682 Charged Particle and Photon Interactions with Matter

radicals can also be determined by the CT reaction between PS

•+

and N,N,N′,N′-tetramethyl-p-

phenylene-diamine

(TMPD) as follows:

PS T MPD PS TMPD

• •

+ +

+  → + (25.12)

The formation process of TMPD

•+

can be observed as an increasing process in the transient absorp-

tion at 570nm, at which TMPD

•+

in Bz peaks with an extinction coefcient of 1.2 × 10

−1

(Stegman and Cronkright, 1970). The tting of the formation of the kinetic trace at 570 nm and the

decay trace at the absorption maximum of PS

•+

based on the extinction coefcient of TMPD

•+

gives

•+

extinction coefcients (ε

•+

) of 3.3 × 10

–2.0 × 10

−1

, as listed in Table 25.1.

In the quantitative elucidation of the degree of charge delocalization on the Si skeleton, the

authors have already reported the degree of charge delocalization (n

del

) determined through simul-

taneous observation of transient bleaching of the lowest excitonic backbone peak (

∆

) and the

formation

of the UV bands (

∆

•

), as given by (Seki et al., 2001a, 2004a; Kawaguchi et al., 2003)

del

OD ES

⋅

•

∆

(25.13)

The empirical relationship between the apparent oscillator strength of the UV band (f

•±

) and the

value

of n

del

has been obtained, expressed as (Seki et al., 1999a, 2001a,b, 2004a)

n f

del

∝

•

(25.14)

f d

• •

± − ±

= ×

∫

4 32 10

. ε ν (25.15)

The values of f

•−

and f

•+

can be obtained by a numerical integration of the Gaussian t to the IBL

–

CB transitions. The obtained values of f

•−

and f

•+

are summarized in Table 25.1. The half-Gaussian

(high-energy side), half-Lorentzian (low-energy side) t gives lower values of f

•−

and f

•+

, but the

difference between the two ttings is less than 20% for all spectra. The values of f

•+

for a series of

Wavelength (nm)

ΔA (10

–4

–1

)

400330

Ar-saturated

-saturated

(cation radicals)

8.0

Ar-saturated

with 20 mM TEA

(anion radicals)

Figure 25.5 UV band in the transient absorption spectra of ME3 in Bz at 5.0 × 10

−3

mol dm

−3

(base mol unit).

Spectra were recorded 100ns after electron pulse irradiation. Solid (squares), dashed (diamonds), and dot-dashed

(circles) lines are spectra observed in O

-saturated, Ar-saturated with TEA at 2.0 × 10

−3

mol dm

−3

, and Ar-saturated

solutions, respectively. (Reprinted from Seki, S. et al., J. Am. Chem. Soc., 126, 3521, 2004a. With permission.)

Nanoscale Charge Dynamics and Nanostructure Formation in Polymers 683

MEn, PDn, PHn, and poly(arylsilane)s are plotted as a function of α in Figure 25.6. The values of

•+

principally reect the values of α, and exhibit a similar dependence on the degree of delocal-

ization of ES on the Si backbone. However, ε

•+

and f

•+

values observed for the phenyl-substituted

polysilanes are clearly higher than those for the dialkyl-substituted polysilanes. In contrast to cation

radicals, the anion radicals of all the polysilanes represent a clear correlation between the values

•−

and α, without contribution from the presence of phenyl pendant groups in their side chains.

The values of both q and L derived from Equations 25.4 and 25.5 clearly indicate that the physical

segmentation length in Si backbones of the ground-state or excited-state polysilanes is well inter-

preted in terms of the molecular stiffness, α. In the case of cation radicals, the dependenceof f

•+

, or

del

, on the molecular stiffness can be apparently categorized into three series: polysilanes bearing

no, one, or two phenyl rings, despite the no categories presenting in anion radicals. It should be noted

that the transient spectra in the present study were recorded after intra-chain CT processes, which

were expected to occur within a few ns (Matsui et al., 2002b). Thus, the spectra occur mainly due to

energetically favorable segments in polysilane backbones. Despite the fact that PH2 exhibits almost

identical Stokes shifts to those of any poly(dialkylsilanes), a relatively high value of f

•+

is observed

for PH2. This is strongly suggestive of a crucial contribution from the phenyl rings only in the delo-

calization of positive charges. Takeda et al. suggested, based on theoretical calculations, that the

valence state and the ES of steady-state polysilanes bearing phenyl rings are considerably affected

by σ-π mixing. The present results also indicate that this is the case, giving rise to IBL

. A delocal-

ized IBL

over the phenyl ring is a good explanation for the increase in f

•+

for poly(n-alkylphenyl-

silane), though an IBL

−

is considered to be localized onto Si chains without spreading over phenyl

substituents. This explanation is supported by models of the spread of the singly occupied molecular

orbital (SOMO) state into substituents, as calculated theoretically only for polysilane cation radicals

(Kumagai et al., 1995; Seki et al., 1999a, 2002). Therefore, the UV band includes both the transition

from IBL

to CB, and the transition from IBL

to pseudo-π at an energy of ∼0.2eV below the CB.

In summary, results of the spectroscopic measurement on ion radicals of σ-conjugated poly-

mers were quantitatively discussed in terms of oscillator strength, hence the degree of delocaliza-

tion of excess charges along their σ-conjugated skeletons. The principal role of side-chain pendant

groups was suggested for the delocalization of positive charge, despite the insulating nature of the

0.2

0.4 0.5 0.6 0.7 0.8 0.9 1

0.4

0.6

0.8

Diaryl-substituted

(DPA and DPB)

Aryl-substituted

(PHn)

Dialkyl-substituted

(MEn, PDn, and PCHMS)

Viscosity index, α

Oscillator strength, f

•

Figure 25.6 Oscillator strength (f

•

) of UV band versus viscosity index (α). (Reprinted from Seki, S.etal.,

J. Am. Chem. Soc.,

126, 3521, 2004a. With permission.)

684 Charged Particle and Photon Interactions with Matter

side chains for electrons delocalized over Si catenations. The quantitative analysis of the transient

absorption

spectra gave precise number (concentration) of charge carriers generated in the media.

25.2.3 nanoMeter-Scale dynaMicS of the hole and the electron in dna

The direct ionization of DNA produces electrons and holes randomly throughout the DNA structure

and its intimate surroundings nearly in proportion to the number of valence electrons at each site.

However, electron spin resonance (ESR) experiments have shown that the nal damage to DNA is

not a random distribution (Bernhard and Close, 2004; Schuster and Landman, 2004; Becker et al.,

2007). The sugar and phosphate radicals are produced in smaller abundance than expected from

the 50% fraction of ionization that occurs on the sugar–phosphate backbone. Rather, the majority of

the radicals are on the DNA bases. Furthermore, the radical is not randomly distributed among the

bases. Among the bases, the localization is not statistical, but shows a high degree of specicity. The

electrons are trapped initially on cytosine (C) and later on thymine (T) moieties. The electron “hole”

is localized almost exclusively on guanine (G), which has the lowest oxidation potential, among

the four bases. The identication of the DNA sites that trap these holes and electrons is essential

to understand radiation-induced damage. This section describes the dynamics of cation and anion

radicals of DNA in an aqueous solution at room temperature using the technique of pulse radiolysis

(Kobayashi

and Tagawa, 2003; Kobayashi et al., 2008; Yamagami et al., 2008).

25.2.3.1

radical

Cation

For

the observation of the radical cation generated in DNA, aqueous solutions of oligonucleotides

(ODNs) are irradiated in the presence of persulfate (

S O

2 8

2−

). The hydrated electron (

−

) reacts

with

S O

2 8

2−

to give

−i

. Since the

−i

(2.5–3.1V) is powerful enough to oxidize all four bases

(Candeias and Steenken, 1993; Steenken and Javanovic, 1997), the radical cations of bases are

formed unselectively within double-stranded ODN. However, the guanine radical cation (G

+•

) is

exclusively formed, whereas the spectra in the single-stranded ODN exhibit a composite spectra

of four nucleic acid base species. The essentially quantitative formation of G

+•

, based on the initial

yield of

−i

, suggests that hole transfer occurs efciently through 13–15 mer ODN. This indicates

that hole transfer from the initially formed loss centers to G sites occurs along double-stranded

ODN very rapidly (<2 × 10

−1

). By introducing mismatch, the formation of G

+•

was diminished,

suggesting

that local perturbation of the duplex structure results in incomplete hole migration.

The

+•

of deoxyguanosine (dG) has a pK

of 3.9, and it rapidly loses the N1 proton with a rate

constant of 1.8 × 10

−1

at a pH of 7.0 (Kobayashi and Tagawa, 2003). In the guanine radical cation/

cytosine base pair (G

+•

-C), it is not clear whether or not the proton of N1 of G shifts to C. Based on

the pK

values of the N1 proton of G

+•

(pK

= 3.9) and N3-protonated C (pK

= 4.3), the equilibrium

for the proton transfer lies slightly to the right. The value of K

for proton transfer suggests that the

base pair (G

+•

-C) in DNA has only partial proton transfer and both cation radical (30%) and depro-

tonated neutral radical (70%) forms exist. There exists extensive discussion regarding the proton-

ation state of the base pair (G

+•

-C) (Steenken, 1989, 1997; Kobayashi et al., 2003, 2008; Lee et al.,

2008). Recently, ESR studies at 77K clearly show that one-electron-oxidized G in double-stranded

ODN

exists as the deprotonated neutral radical G(−H)

•

(Adhikary et al., 2009).

–

•

(25.16)

Nanoscale Charge Dynamics and Nanostructure Formation in Polymers 685

The G

+•

in ODN, produced by oxidation with

−i

, deprotonates to form G(−H)

•

with a half

time of 35ns (Kobayashi et al., 2003, 2008). The spectrum of G

+•

exhibits absorption around

450nm, whereas G(−H)

•

has absorption around 380–500nm, together with a characteristic shoulder

in the spectral range of 600–700nm (Figure 25.7). The deprotonation of G

+•

is monitored at 625 nm

(Figure

25.8).

The

deprotonation of G

+•

in double-stranded DNA is composed of several steps, including proton

transfer from G to C in the base-pair radical cation (G

+•

-C) (step i) and the release of the proton

from C(+H

) into the solution (step ii) (Scheme 25.1): K

(=k

−i

) is the equilibrium constant and k

is the rate constant of the release of proton. Since proton transfer from the N1 proton of G to the N3

proton of C could, in principle, occur very rapidly on the order of 10

−1

(Douhal et al., 1995), the

rate-limiting step in these processes is the deprotonation of C(+H

) by a watermolecule(k

≪k

−i

•

G(–H) C(+H

)]

(i)

O H

G(–H) C

(ii)

•

sCheme 25.1

350 400 450 500 550 600 650 700

0.02

0.04

0.06

0.08

Wavelength (nm)

ΔA

Figure 25.7 Kinetic difference spectra of pulse radiolysis of ODN G

3AA

monitored at 50ns (•) and 500ns

(○) after pulse radiolysis.

200 ns

ΔA =0.02

625 nm

Figure 25.8 Kinetics of absorbance change after pulse radiolysis of ODN G

3AA

686 Charged Particle and Photon Interactions with Matter

The rate constant of the deprotonation can be expressed as k = k

. This suggests that the absor-

bance changes observed would correspond to the proton loss from the oxidized proton-shifted reso-

nance structure (G

+•

:C ↔ G(−H)

•

:C(+H

)) by a water molecule. This is further supported by the

effect of the substitution by the methyl or bromine group for the cytosine C5 hydrogen in selected

positions complementary to G (Table 25.3) and by the solvent deuterium isotope effect (Kobayashi

etal., 2008). A bromo substituent on C, as an electron-accepting group, suppressed the process

ca. 20-fold compared with a methyl substituent on C, as an electron-donating group. The proton

transfer from G

+•

(pK

= 3.9) to

C is thermodynamically unfavorable, since the pK

dC is 2.8.

The equilibrium of

C in Equation 25.1 is far to the left, and the equilibrium constant (K

) of the

proton transfer in G

+•

C is estimated to be 0.08, suggesting that a large portion of the oxidized G in

C remains protonated after oxidation. Thus, the difference in the rate constants can be explained

primarily by the pK

values of C. For the solvent deuterium isotope effect, it should be noted that a

kinetic isotope effect (k

= 3.8) of the deprotonation in ODN was about twice that of the kinetic

isotope effect on the deprotonation of free dG (k

= 1.7). This suggests that the formation of

G(−H)

•

in ODN may be associated with the loss of tightly bound protons in DNA. Thus, this process

consistent with step (ii) in Scheme 25.1.

is of particular interest to study ODN sequences showing the spectroscopic changes and the

rate constants of the deprotonation of G

+•

(Kobayashi et al., 2008). The characteristic absorption

maxima of G

+•

intermediate around 450nm were found to differ. These spectra shifted to longer

wavelengths in the order of G < GG < GGG. The spectral shift can be understood as the stacking

interaction of two or three consecutive guanine bases. These results demonstrate that the tran-

siently formed G

+•

is stabilized by base pairing with C and the stacking interaction of neighborhood

nucleobases. These ndings provide direct spectroscopic evidence of the delocalization of the posi-

tive charge along the extended π orbitals of DNA bases (Saito et al., 1998; Yoshioka et al., 1999;

Shao et al., 2004). In contrast, the spectra of G(−H)

•

in ODNs are essentially identical to that of dG

and are not affected by the ODN sequence. This strongly suggests that the radical orbital of G(−H)

•

essentially localized on a preferential specic guanine base site.

table 25.3

sequences

of the odn

3AA

TTTTTCCCTTTTT 5

AAAAAGGGAAAAA 3

AAAAAAGAAAAAA 3

TTTTTT

CTTTTTT 5

TTTTTT

CTTTTTT 5

OH OH

Nanoscale Charge Dynamics and Nanostructure Formation in Polymers 687

The rate constants of deprotonation are affected by ODN sequences. The differences among

G, GG, and GGG were not distinct, but rather depended on the neighboring bases. The rate con-

stant increases in the order CGC > TGT > AGA > AGG > GGG, an order that may correlate with

thecalculated ionization potential of the nucleobases (Yoshioka et al., 1999; Voityuk et al., 2000).

The nding that the rate decreases as the oxidation potential decreases may reect on the stability

of the radical cation.

25.2.3.2 radical

nion

The radical anions of nucleotides in DNA are produced by the reaction of hydrated electrons

(

−

) with ODN. Hydrated electrons (

−

) can react with all four nucleobases, and the result-

ing products of the nucleic acids have been identied spectroscopically. The differences in the

dynamics of ODNs and corresponding isolated nucleotides are compared. However, unlike posi-

tive holes, electrons are not trapped at specic sites of DNA. Here, we present the dynamics of

electron adducts of ODNs containing A·T and G·C base pairs with those of single-stranded ODNs

and the isolated nucleotides.

Radical anions of double-stranded ODNs containing A·T bases were produced by pulse radioly-

sis of a deaerated aqueous solution. Figure 25.9 shows transient absorption spectra after pulse radi-

olysis of ODN AT. For comparison, the spectra of single-stranded ODN (5′-TAATTTAATAT-3′),

dT, and dA are shown in Figure 25.10. The spectrum of single-stranded ODN, which has no

absorption around 480 nm, is essentially similar to that of dT. In single-stranded ODN, elec-

tron transfer from the initially formed transients to T sites is completed within 10ns. In double-

stranded ODN, on the other hand, the initial transient obtained after the complete decay of

−

and characterized by an absorption at around 480nm and a broad absorption from 600 to 900 nm,

is apparently complex because of the presence of A and T bases. However, this spectrum was not

reproduced as a set of benchmark spectra for dA and dT. These ndings, therefore, suggest that the

transiently formed (A·T)

•−

is stabilized by hydrogen-bond base-pairing, and the unpaired electron

is delocalized to the A site.

Following the spectral change, other spectral changes took place in the time range of micro-

seconds. The absorption around 480 nm, associated with the charge delocalization on the A site,

disappeared with the decay of absorption at 350 nm (Figure 25.9). The spectrum of the resulting

product at 20 μs after the pulse, which has an absorption maximum at 350 nm, is similar to that

of T

•

−

. This process follows a decrease of absorbance at 480 nm (inset of Figure 25.9) and obeys

300 400 500 600 700 800 900 1000

0.1

0.2

0.3

ΔA

Wavelength (nm)

Figure 25.9 Kinetic difference spectra of pulse radiolysis of ODN AT at 50ns (○), 1μs (•), and 20μs (□).

688 Charged Particle and Photon Interactions with Matter

rst-order kinetics with a rate constant for ODN AT of 1.6 × 10

−1

, and it is independent of

ODN concentration. Therefore, the spectroscopic changes in Figure 25.8 may reect progress

from the transient (A·T)

•−

to T

•−

in the A·T anion complex. The nature of the species generating

the 20 μs spectrum suggests that the electron adduct center became localized mainly at the T

site (Gu et al., 2005).

Proton-transfer reactions are important for the stabilization of initial DNA ion radicals (Barnes

etal., 1991; Steenken et al., 1992; Steenken, 1997; Kobayashi et al., 2008). In regard to excess elec-

trons of DNA, attention has been focused on the protonation state and protonation sites of C andT. In

a G-C pair, C

−•

rapidly acquires a proton at N3 through hydrogen bonding with the HN1 of G. From

the pK

values of the N3-protonated C(H)

•

(pK

> 13) and the N1 proton of G (pK

=9.5) (Steenken

et al., 1992), the equilibrium of GC

•−

in ODN GC is far to the right, and the spectrum observed

is regarded as corresponding to G(−H)

−

-C(H)

•

. Moreover, the protonation of free dC occurs before

10ns. Such a rapid reaction may be due to the protonation by water-bound N3 of C. Similar results

were observed using conductance techniques (Hissung and von Sonntag, 1979) and in ESR/electron

nuclear double resonance (ENDOR) studies of cytosine monohydrate single crystals irradiated at 10K

(Sagstuen etal., 1992). In contrast, since T

−•

is only a weak base, it was present transiently in a neutral

aqueous solution and protonated to form the neutral radical T(H)

•

. Thus, the protonation of dT occurs

reversibly at O4 by the direct combination of T

−•

with protons in solution. In double-stranded ODNs,

since the pK

values of the O4 of protonated T(H)

•

(pK

= 6.9) and the N6 of A (pK

> 13.75), T

−•

not protonated by its complementary base, A. Thus, the spectrum of the electron adduct of ODN AT

corresponds to A-T

−•

. The process of protonation of T in ODN AT is thus consistent with a slower

process of spectroscopic change observed on the scale of hundreds of microseconds. Remarkably,

however, this process revealed a substantial difference between the ODN AT and free dT. At a pH of

7.0, the rate constant of the former (4 × 10

−1

) was 10-fold slower than that of the latter (4.3 × 10

−1

In addition, the rates of ODN AT are independent of a change in the pH from 5.8 to 8.3. This differ-

ence may be due to the differences in the protonation sites of T in DNA and free dT. ESR studies using

low-temperature glasses yield strong evidence that when DNA is exposed to ionizing radiation at 77K,

an ESR doublet associated with T

−•

is converted into a readily identiable eight-line spectrum of the

5,6-dihydrothymine-5-yl radical, a conversion due to the irreversible protonation of T

•−

at C6. A water

molecule bound to DNA would participate in this protonation. Thus, irreversible protonation by water

at C6 is both a pH independent and a relatively slow process.

0.1

0.2

0.3

Wavelength (nm)

300 400

ε (M

–1

)

500 600 700 800 900 1000

1000

2000

3000

4000

5000

6000

7000

8000

ΔA

-TAATTTAATAT-3

Figure 25.10 Absorption spectra at 200 ns after pulse radiolysis of dA (○), dT (

⚫

), and single-stranded AT

(△). Note that the right-hand ordinate is in units of dA corresponding to the kinetic difference spectrum of

ODN,

whereas the left-hand ordinate is in units of extinction coefcients of dA and dT.

Nanoscale Charge Dynamics and Nanostructure Formation in Polymers 689

25.3 nanosCale dynamiCs oF Charge Carriers

by

ime-r

esolved m

iCrowave

Condu

Ctivity

The π- and σ-conjugated materials are well-known organic semiconductors due to their robust

nature, multiplicity of chemical constitution, and capability of charge transport that can be tailored

via judicious functionalization. They have demonstrated feasibility for industrial applications, such

as organic thin-lm transistors [TFT] (Sirringhaus et al., 1998, 1999), light-emitting diodes [LED]

(Kido et al., 1995; Sheats et al., 1996), photovoltatic cell [PV] (Sariciftci et al., 1992; Yu et al., 1995),

and radio-frequency identication [RFID] tag (Baude et al., 2003; Myny et al., 2008). The pos-

sibility of solution process opens a versatile route toward the fabrication of large-area devices on

exible substrates. Alongside these benets, considerable efforts have been devoted not only to the

development of novel conjugated materials, but also to the understanding of their optical, electric,

and optoelectronic properties. The mobility of charge carriers as well as their polarity (positive

or negative) play a key role in the performance of organic electronic devices, as these relate to the

response speed, the efciency of charge carrier transport, and the realization of complementary

circuits. Generally, the mobility is measured by direct-current (DC) techniques, for example, FET

and TOF; however, these are considerably dependent on the lm morphology and sample purity like

grain boundary, impurities, and structural defects, which are undesirable features hiding the intrin-

sic nature of charge carrier transport. The interface between the electrodes and the semiconductors

causes problems of barrier against charge injection and strong electric eld, which disturbs the ther-

mal motion of charge carriers. The long-range transport property estimated by the DC technique is

important for the actual electrical devices; however, to reveal the principal character, without being

affected by complicated factors, will offer chances to mitigate barriers for the development of novel

organic electronics and for their performance optimization.

In order to perform direct observation of the drift mobility of charge carriers, free from the

disturbance at interfaces, such as electrode contacts and domain boundaries, the TRMC technique

based on microwave absorption has been developed (de Haas et al., 1975; Warman et al., 2002).

TRMC, where the probe is an alternating-current (AC) electromagnetic wave, can overcome the

drawbacks of DC measurements, as follows: (1) As it does not necessitate the fabrication of elec-

trodes in contact with polymers, contact and surface interactions can be avoided. (2) The high

frequency (GHz) of microwaves extends the observable region of charge carrier mobility to a very

short regime (nanometer scale). Thus, the mobility obtained at the end of a pulse is not signicantly

affected by the grain and/or domain boundary conditions. (3) The low magnitude of the electric eld

makes it possible to observe intrinsic charge-transport phenomena in well-organized and/or highly

extended conjugated segments of polymers without perturbing the thermal drift of the charge car-

riers. Because of these unique properties, the mobilities of polymer matrices obtained by TRMC

are typically orders of magnitude higher than those obtained by the DC technique, in which charge

migration

is dominated by interchain transport and other undesirable factors.

The

microwave absorption technique was initially proposed by Biondi et al. in 1949, who mea-

sured electron concentrations in the gas phase (Biondi and Brown, 1949; Biondi, 1951). Time-

resolved experiments with an enhanced S/N ratio and wide dynamic ranges were realized by the

improvement of microwave circuits and the use of pulsed high-energy radiations (mainly, EB or

converted x-ray). Fessenden et al. at the University of Notre Dame have conducted investigations

on dissociative electron attachment in the gas phase (Warman and Fessenden, 1968; Warman etal.,

1972). Shimamori and Hatano reported an electron thermalization process in the gas phase and

the reaction between electrons and molecules (Shimamori and Hatano, 1977). The researches were

expanded to transient change in the inter- or intramolecular dipole moment (Fessenden et al., 1979;

Fessenden et al., 1982). The separation of real and imaginary parts of the signals has been discussed

in a gas phase (Suzuki and Hatano, 1986a,b). Investigations on energy loss processes (Shimamori

and Sunagawa, 1997) and excited states in the liquid phase (Fessenden and Hitachi, 1987) have been

performed. Using pulse-radiolysis time-resolved microwave conductivity (PR-TRMC), Warmanetal.