Electronic states of the DNA polynucleotides poly(dG)-poly(dC) in the presence of iodine

The electronic states of chemically doped DNA polynucleotides [poly(dG)-poly(dC)] have been studied by photoelectron spectroscopy, x-ray absorption spectroscopy, and Raman spectroscopy, in order to understand the charge conduction mechanism of this material. Upon iodine doping (i.e., hole doping), we clearly observed radical cation formation on DNA bases, and the appearance of additional occupied and unoccupied electronic states within the band gap of poly(dG)-poly(dC). Referring to the analogous studies for $\pi$-conjugated conductive polymers, we point out that the charges induced by iodine doping and their derived electronic states are important origins of the hole-conductive electrical properties of poly(dG)-poly(dC).

Figure 1

X-ray PE spectra ($\mathrm{Mg}K\alpha$, m $h\nu =1253.6\mathrm{eV}$) for nondoped poly(dG)-poly(dC) (NGC) and iodine-doped poly(dG)-poly(dC) (IGC-1 and IGC-2) films in the BE regions of (a)–(c) $\mathrm{I}3d$ and (d)–(f) $\mathrm{N}1s$ core levels. All of the film samples were fabricated onto ${\mathrm{SiO}}_2∕p\text{−}\mathrm{Si}\left(111\right)$ substrates with a thickness of approximately $50\mathrm{nm}$. The measurements were done at room temperature with energy resolution of approximately $1.0\mathrm{eV}$. For all of the displayed $\mathrm{N}1s$ x-ray PE spectra, the raw data are plotted by closed diamonds. In the peak assignments for each $\mathrm{N}1s$ spectrum, the contribution of secondary electrons was first subtracted from the raw data by means of the Proctor-Sherwood method. The individual $\mathrm{N}1s$ components were evaluated by a least squares fitting. The iodine-doping effect is observed as the appearance of a higher-BE component attributed to the radical cationic nitrogen species.

Figure 2

Valence-band electronic states (occupied states) for (a) NGC (dotted line), (b) IGC-1 (dashed line), and (c) IGC-2 (solid line), characterized by vuv PES $(h\nu =170\mathrm{eV})$. These samples are the same ones utilized for the x-ray PES measurements shown in Fig. 1. The spectra were measured at room temperature with energy resolution of $150\mathrm{meV}$. Overall the shapes of the spectra are comparable to each other, indicating that the chemical structure of poly(dG)-poly(dC) is not significantly modified upon iodine doping.

Figure 3

Expanded VUV PE spectra in the range close to the valence band maximum for NGC (dotted line), IGC-1 (dashed line), and IGC-2 (solid line), in which the energy resolution is identical to that in Fig. 2. These samples are also the same ones utilized at the x-ray PES (Fig. 1) and VUV PES (Fig. 2) measurements. In the inset, $\mathrm{P}2p$ core levels of these samples are displayed with their full width at half maximum values, which clearly indicate the broadening by charge-up or other effects was neglected in these measurements. The iodine-doping (hole-doping) effect is observed as a shift of the valence band edge toward the Fermi level (indicated by the arrow). This feature implies the appearance of additional occupied electronic states above the intrinsic $\pi$ band of poly(dG)-poly(dC).

Figure 4

Unoccupied electronic states for (a) NGC, (b) IGC-1, and (c) IGC-2 characterized by nitrogen $K$-edge XAS. These samples are the same ones utilized in Figs. 1, 2, 3. The spectra were measured at room temperature with a resolution of $300\mathrm{meV}$, and the intensities were normalized in the regions of 395–398 and $420–425\mathrm{eV}$ (not shown here). A typical iodine-doping (hole-doping) effect is observed as the appearance of a shoulder component below the ${\pi }_{\mathrm{N}}^{*}$ resonance peak (indicated by the arrow). This feature implies the appearance of additional unoccupied electronic states below the intrinsic ${\pi }^{*}$ band of poly(dG)-poly(dC).

Figure 5

Raman spectra (Ar ion laser excitation, $\lambda =514.5\mathrm{nm}$) of (a) nondoped poly(dG)-poly(dC), and (b)–(e) iodine-doped poly(dG)-poly(dC) at various doping levels. For these measurements, the films were fabricated onto ${\mathrm{Al}}_2{\mathrm{O}}_3\left(0001\right)$ substrates with a thickness of the order of micrometers. The measurements were carried out under atmospheric conditions with a resolution of approximately $4{\mathrm{cm}}^{-1}$. Intensities of the obtained spectra were normalized with the Raman peak at $788{\mathrm{cm}}^{-1}$ ($\mathrm{O}\text{−}\mathrm{P}\text{−}\mathrm{O}$ stretch vibration mode in the backbone of the DNA strand, denoted “Bk”). The $2{\nu }_1$ and $3{\nu }_1$ (“Region $X$”) Raman bands correspond to the second and third harmonics of the symmetrical stretching mode of the ${{\mathrm{I}}_3}^{-}$ ion, respectively $({\nu }_1\approx 107{\mathrm{cm}}^{-1})$. The Raman band of “Region $Y$” is mainly contributed by the deformation mode of guanine (partly, that of cytosine). The inset plots the Raman intensity integral of Region $Y$ versus that of Region $X$. From this relationship, it is clear that the evolution of the Region $Y$ Raman band strongly depends on the amount of ${{\mathrm{I}}_3}^{-}$ ion in the sample.

Figure 6

(a) Difference in Raman intensities between nondoped and iodine-doped poly(dG)-poly(dC) spectra [Figs. 5a, 5e, respectively]. The signal change originating from the iodine doping is clearly seen in the higher-wave-number regions. Calculated Raman spectra $(\mathrm{B}3\mathrm{LYP}∕6\text{−}31{\mathrm{G}}^{*})$ are shown for (b) radical cation phase of guanine $\left({\mathrm{G}}_{\mathbf{\bullet }}^{+}\right)$, (c) neutral guanine (G), (d) radical cation phase of cytosine $\left({\mathrm{C}}_{\mathbf{\bullet }}^{+}\right)$, and (e) neutral cytosine (C). A scale factor of 0.96 was used for the calculated vibration frequencies.