Nanocopying of Individual DNA Strands and Formation of the

DNA was first spread on a solid substrate, and its molecular surface was coated with an ultrathin titania layer by 3 cycles of the surface sol−gel p...
0 downloads 0 Views 551KB Size
Download PDF
Langmuir 2005, 21, 8899-8904

8899

Nanocopying of Individual DNA Strands and Formation of the Corresponding Surface Pattern of Titania Nanotube Shigenori Fujikawa,* Rie Takaki, and Toyoki Kunitake Topochemical Design Laboratory, Spatio-Temporal Function Materials Research Group, Frontier Research System, RIKEN, Wako, Saitama 351-0198, Japan Received June 12, 2005 Titania nanotube was prepared by nanocopying of the individual DNA double strand as template. DNA was first spread on a solid substrate, and its molecular surface was coated with an ultrathin titania layer by 3 cycles of the surface sol-gel process. Fluorescence microscopic images before and after titania coating of the DNA/YOYO-1 complex were essentially identical, showing that the titania coating did not change the chemical properties of the complex. Titania coating effectively prohibited chemical degradation of titania-coated DNA with DNase I and physically separated the DNA strand from the surrounding environment with an ultrathin titania barrier. The morphology of the DNA strand was preserved, as confirmed by microscopic and spectroscopic observations. The presence of the hollow (tubular) structure was confirmed by a silver staining experiment coupled with scanning transmission electron microscopyenergy-dispersive X-ray spectroscopy (STEM-EDX) analysis. The present finding shows the effectiveness of nanocopying of the individual DNA strand.

Introduction One-dimensional structures in the nanometer regime have been studied from both scientific and industrial aspects in the past decade. Carbon nanotube is a representative material and now is utilized in molecular electronics parts,1-7 as a hydrogen absorbent,8-11 and as a probe tip in atomic force microscopy.12-14 Non-carbon nanotubes and wire-shaped nanostructures have been also used as new functional materials for developing optical, magnetic, and electronic nanodevices. Among several approaches for the synthesis of onedimensional (1D) nanostructures,15 template synthesis would be a particularly useful one, if precise replication is achieved in the nanometer precision. There have been * Corresponding author: e-mail [email protected]. (1) Tans, S. J.; Devoret, M. H.; Dai, H.; Thess, A.; Smalley, R. E.; Geerligs, L. J.; Dekker, C. Nature 1997, 386, 474-477. (2) Bockrath, M.; Cobden, D. H.; McEuen, P. L.; Chopra, N. G.; Zettl, A.; Thess, A.; Smalley, R. E. Science 1997, 275, 1922-1925. (3) Yao, Z.; Postma, H. W. Ch.; Balents, L.; Dekker, C. Nature 1999, 402, 273-276. (4) Fuhrer, M. S.; Nygård, J.; Shih, L.; Forero, M.; Yoon, Y.-G.; Mazzoni, M. S. C.; Choi, H. J.; Ihm, J.; Louie, S. G.; Zettl, A.; McEuen, P. L. Science 2000, 288, 494-497. (5) Tans, S. J.; Verschueren, A. R. M.; Dekker, C. Nature 1998, 393, 49-52. (6) Martel, R.; Schmidt, T.; Shea, H. R.; Hertel, T.; Avouris, Ph. Appl. Phys. Lett. 1998, 73, 2447-2449. (7) Postma, H. W.; Teepen, T.; Yao, Z.; Grifoni, M.; Dekker, C. Science 2001, 293, 76-79. (8) Dillon, A. C.; Jones, K. M.; Bekkedahl, T. A.; Kiang, C. H.; Bethune, D. S.; Heben, M. J. Nature 1997, 386, 377-379. (9) Ye, Y.; Ahn, C. C.; Witham, C.; Fultz, B.; Liu, J.; Rinzler, A. G.; Colbert, D.; Smith, K. A.; Smalley, R. E. Appl. Phys. Lett. 1999, 74, 2307-2309. (10) Liu, C.; Fan, Y. Y.; Liu, M.; Cong, H. T.; Cheng, H. M.; Dresselhaus, M. S. Science 1999, 286, 1127-1129. (11) Dillon, A. C.; Heben, M. J. Appl. Phys. A 2001, 72, 133-142. (12) Dai, H.; Hafner, J. H.; Rinzler, A. G.; Colbert. D. T.; Smalley, R. E. Nature 1996, 384, 147-150. (13) Wong, S. S.; Harper, J. D.; Lansbury, P. T., Jr.; Lieber, C. M. J. Am. Chem. Soc. 1998, 120, 603-604. (14) Wong, S. S.; Joselevich, E.; Woolley, A. T.; Cheung, C. L.; Lieber, C. M. Nature 1998, 394, 52-55. (15) As a comprehensive review, see Xia, Y.; Yang, P.; Sun, Y.; Wu, Y.; Mayers, B.; Gates, B.; Yin, Y.; Kim, F.; Yam, H. Adv. Mater. 2003, 15, 353-389.

several attempts at template synthesis by using virus,16-18 molecular assemblies,19-22 inorganic crystals,23-28 and carbon nanotubes,29,30 as templates. Individual organic molecules and polymers possess ultimate small sizes, and they can act as the smallest templates. We already showed that it was possible to wrap the individual chain of a cationic polymer with a silicate sheath.31 However, the template polymer was polydisperse with a wide molecular weight distribution, and removal of the template polymer was not conducted. We herein report preparation and structural characterization of a titania nanotube by using DNA molecules as template. The DNA chain is a complex molecule that is most readily designed, in term of sequence, chain length, and morphology, and is highly attractive as a template. Ultrathin films of metal oxides are superior candidates for copying organic molecules with nanometer precision.32,33 This process is based on the stepwise chemi(16) Shenton, W.; Douglas, T.; Young, M.; Stubbs, G.; Mann, S. Adv. Mater. 1999, 11, 253-256. (17) Flowler, C. E.; Sheton, W.; Stubbs, G.; Mann, S. Adv. Mater. 2001, 13, 1266-1269. (18) Fujikawa, S.; Kunitake, T. Langmuir 2003, 19, 6545-6552. (19) Kwan, S.; Kim, F.; Arkana, J.; Yang, P. Chem. Commun. 2001, 447-448. (20) Huang, L.; Wang, H.; Wang, Z.; Mitra, A.; Bozhilov, K. N.; Yan, Y. Adv. Mater. 2002, 14, 61-64. (21) Jung, J. H.; Ono, Y.; Hanabusa, K.; Shinkai, S. J. Am. Chem. Soc. 2000, 122, 5008-5009. (22) van Bommel, K. J. C.; Jung, J. H.; Shinkai, S. Adv. Mater. 2001, 13, 1472-1476. (23) Mayya, K. S.; Gittins, D. I.; Dibaj, A. M.; Caruso, F. Nano Lett. 2001, 1, 601-603. (24) Yin, Y.; Lu, Y.; Sun, Y.; Xia, Y. Nano Lett. 2002, 2, 427-430. (25) Song, J. H.; Wu, Y.; Messer, B.; Kind, H.; Yang, P. J. Am. Chem. Soc. 2001, 123, 10397-10398. (26) Sun, Y.; Mayers, B. T.; Xia, Y. Nano Lett. 2002, 2, 481-485. (27) Wen, X.; Yang, S. Nano Lett. 2002, 2, 451-454. (28) Sun, Y.; Mayers, B. T.; Xia, Y. Adv. Mater. 2003, 15, 641-646. (29) Satishkumar, B. C.; Govindaraj, A.; Voli, E. M.; Masumallic, L.; Rao, C. N. R. J. Mater. Res. 1997, 12, 604-606. (30) Ajayan, P. M.; Stephan, O.; Redlich, Ph.; Colliex, C. Nature 1995, 375, 564-567. (31) Ichinose, I.; Kunitake, T. Adv. Mater. 2002, 14, 344-366. (32) Huang, J.; Ichinose, I.; Kunitake, T. Chem. Commun. 2002, 18, 2070-2071. (33) Hunag, J.; Kunitake, T. J. Am. Chem. Soc. 2003, 125, 1183411835.

10.1021/la051554o CCC: $30.25 © 2005 American Chemical Society Published on Web 07/28/2005

8900

Langmuir, Vol. 21, No. 19, 2005

Fujikawa et al.

Figure 1. Schematic representation of the surface sol-gel process.

Figure 2. Schematic representation of DNA nanocopying.

sorption of metal alkoxides from solution onto substrate surfaces and subsequent hydrolysis of the alkoxide moiety to give nanometer-thick oxide films (Figure 1). The DNA molecule possesses phosphate groups of nucleotides along the double helix, and would readily react with titania layers on the molecular surface in the absence of additional linker materials and molecules. Figure 2 is a schematic representation of our approach. First, DNA strands are spread onto solid substrates by casting of their solutions (Figure 2a). Then, the surface sol-gel process is conducted to coat DNA chains with ultrathin titania layers (Figure 2b), and finally, the template DNA is removed by oxygen plasma treatment to produce cavities in the titania nanochannel (Figure 2c). Experimental Section Materials. λ DNA [Nippon Gene, 48 502 base pairs, 0.620 µg/µL in 10 mM Tris-HCl, pH 7.9, 1.0 mM ethylenediaminetetraacetic acid (EDTA), and 1 mM aqueous MgCl2] and pBR322 (Takara, 4361 bp, 0.5 µg/µL in 10 mM Tris-HCl, pH 8.0, and 1 mM EDTA) were employed as templates. For purification of λ DNA, a 500 µL portion of the original λ DNA solution was mixed with 80 µL of 4 M NaCl (aqueous) and a given amount of ethanol, and the mixture was stored in a freezer (-20 °C) for 1.5 h. The precipitate was separated from the solution by centrifugation with a rotation speed of 1300 rpm for 40 min (Avanti, HP-30I) and was washed with a 70% aqueous solution of cold ethanol. Further centrifugation (1300 rpm, 20 min) was applied to separate the precipitate, which was then freeze-dried for 30 min to obtain solid λ DNA. The purified λ DNA was dissolved in ion-exchanged water to prepare a DNA solution of 6.0 × 10-7 µM in base pairs. In the case of pBR322 plasmid, the as-purchased solution was diluted with aqueous MgCl2 (1 mM) to adjust its concentration to 4 ng/µL. Microscopic Observation. Atomic force microscopy (AFM) observation (Veeco, Explorer TMX2100) was performed in noncontact mode at room temperature. In general, simple casting of DNA solutions was employed to spread DNA strands onto solid substrates. Five microliters of a pBR322 solution (4 ng/µL) was dropped onto a mica surface at room temperature. After 30 s, the mica was washed with ion-exchanged water and was dried by nitrogen gas flushing. Transmission electron microscopy (TEM) observation was conducted on a JEOL JEM-2000EX instrument. A silicon monoxide-coated Au-TEM grid (Ted Pella, silicon monoxide type

A) was set on filter paper. The λ DNA solution was cast onto the TEM grid, and the extra solution was absorbed in the filter paper, and then the TEM grid was washed with ion-exchanged water. The unstained specimen was subjected to TEM observation at an acceleration voltage of 100 kV and an original magnification of 10000. Scanning TEM (STEM)-energy-dispersive X-ray spectroscopy (EDX) was conducted on a Hitachi HF-2200 at an acceleration voltage of 100 kV in the dark-field observation mode. For fluorescence microscopic (FM) observation before and after deoxyribonulease I treatment, YOYO-1 (Molecular Probes, YOYO-1 iodide 1 mM in dimethyl sulfoxide, DMSO) was mixed with the λ DNA solution, prior to casting of DNA. The concentration of λ DNA and YOYO-1 in the mixture were 6.0 × 10-7 µM in base pairs and 6.0 × 10-8 M, respectively. Ethanol (1 µL) was added to a 9 µL portion of this solution. The resultant solution was spread on a cover glass, and the solvent was evaporated by heating the glass plate at 45 °C for 5 min. In fluorescence microscopy (Olympus BX60-34-FLBD1), the magnifications of objective and ocular lens were 50× and 10×, respectively, and the filter set of exciter XF1087 and emitter XF3105 (Optoscience) was employed to acquire FM images through a digital camera system. Glass plates were used as a substrate. X-ray photoelectron spectroscopy (XPS) experiments were carried out on an Escalab 250 (VG) with Al KR (1486.6 eV) radiation. The applied power was operated at 15 kV and 20 mA. The base pressure in the analysis chamber was less than 10-8 Pa. Peak positions were normalized to the carbon peak at 285 eV. Titania Coating. DNA molecules were spread on substrates, as described above, and the substrate was immersed in ethanol or toluene solutions of titanium n-butoxide [Ti(OnBu)4, Gelest], in ethanol and ion-exchanged water, in this order, for 1 min each and then dried by blowing nitrogen gas. The concentration of Ti(OnBu)4 was 10 mM in ethanol, and 100 mM in toluene in the case of AFM observation, and 100 mM in toluene for other experiments. The immersion and drying processes were repeated three times to ensure formation of ultrathin titania layers. In the DNA degradation experiment, both titania-coated and noncoated λ DNA with YOYO-1 doping as dispersed on solid substrates were immersed in aqueous deoxyribonuclease I (DNase I) (Invitrogen, 41 units/µL in 1 mM MgCl2) for 10 min and then washed with ion-exchanged water and observed by fluorescence microscopy. Removal of Template DNA. Oxygen plasma treatment (South Bay Technology: PE-2000 plasma etcher) was carried out by placing the sample directly on the RF electrode. The applied radio frequency (RF) was 13.56 MHz, and the base pressure in

Nanocopying of Individual DNA Strands

Langmuir, Vol. 21, No. 19, 2005 8901

Figure 4. Fluorescence microscopic images of the DNA/ YOYO-1 complex on the glass plate after DNase I treatment: (a) without titania coating; (b) with titania coating.

Figure 3. Fluorescence microscopic images of the DNA/ YOYO-1 complex on the glass plate: (a) on mica; (b) after surface sol-gel process; (c) after oxygen plasma treatment (nothing was observed). the reactor was 75 mTorr. The oxygen (industrial grade) pressure during plasma treatment was about 180 mTorr, and the RF power dissipated to the sample was 20 W for several minutes at room temperature. Silver staining was conducted to characterize the interior space of titania nanochannels. After oxygen plasma treatment of the titania-coated λ DNA on a TEM grid, the substrate was immersed in AgNO3 solution (0.2 M, H2O/EtOH ) 1/1 v/v) for 30 min under room light and then washed with ion-exchanged water. A silverstained sample without removal of DNA was also prepared for comparison. These samples were subjected to TEM observation.

Results and Discussion Fluorescence Microscopic Observation of the DNA/YOYO-1 Complex. YOYO-1 is a fluorescent dye that can bind in the major groove of double-stranded DNA through electrostatic interaction, and the resulting DNA/ YOYO-1 complex exhibits strong green fluorescence. In our study, YOYO-1 was premixed with an aqueous solution of λ DNA, prior to casting of DNA solution. Simple casting of DNA/YOYO-1 in solution on a glass plate gave dotlike morphologies in FM observation (data not shown), and clear distinction between the free YOYO-1 aggregate and the globular state of DNA/YOYO-1 complex was difficult. To avoid this problem, we heated the glass plate at 45 °C during casting (5 min), expecting that DNA strands would be stretched perpendicular to the evaporating front of the DNA solution. The parallel-aligned line morphology was widely observed on the glass plate as a result (Figure 3a). These bright line structures were not observed when YOYO-1 and DNA were separately cast, indicating the lines are composed of the λ DNA/YOYO-1 complex. λ DNA has 48 502 base pairs and the corresponding chain length is about 16 µm. Some of the observed lines are clearly shorter than the length of the fully elongated λ DNA. We suspected that fragmentation of the DNA chain occurred during the casting process. This morphology is not basically affected by the surface sol-gel process, and preservation of the original line structure is noticed in the fluorescence image after the titania coating (Figure 3b). Notably, the florescence intensity was not affected even after titania coating. Therefore, the titania coating apparently did not induce any changes in the chemical structure and electronic properties of the DNA/YOYO-1 complex. After oxygen plasma treatment, the fluorescent

image was not observed at all (Figure 3c). Oxygen plasma treatment is known to decompose organic moieties into gaseous species, such as CO2 and H2O. Therefore, the organic moiety in the DNA/titania complex would be oxidatively decomposed to leave a titania shell behind. Under the same plasma conditions, the organic moiety was completely decomposed in titania-coated polystyrene particles (diameter 500 nm).18 The plasma treatment conditions we employed here seem to be sufficiently strong for decomposition of the organic moiety in the titaniacoated DNA/YOYO-1 complex. Enzymatic Degradation of Coated and Bare DNA. The preceding fluorescence study indicates that linear morphologies of the template DNA strand are preserved after the surface sol-gel process and that oxygen plasma treatment removed the organic components completely from DNA/ YOYO-1 complex. However, these results do not show whether the titania layer fully covers the DNA molecular surface. Chemical degradation of DNA is an effective way to prove whether DNA chains are physically and chemically isolated from reactants of the surrounding environment. Deoxyribonuclease I (DNase I) was used for this purpose. DNase I is an endonuclease that degrades DNA molecules into short fragments. Uncoated DNA chains will be readily decomposed by DNase, while titaniacoated DNA would not be affected if physical isolation is effective. After casting of the DNA/YOYO-1 complex on a solid substrate, three cycles of the surface sol-gel process were conducted, and the substrate was immersed in aqueous DNase I (41 units/mL, 1 mM aqueous MgCl2) for 10 min. Without the surface sol-gel process, fragmented bright spots were observed after DNase treatment (Figure 4a). In contrast, titania-coated DNA showed linear structures abundantly, even after DNase treatment (Figure 4b). It is clear that the titania shell prevented DNA degradation by DNase I, though bare DNA was readily fragmented. These contrasting data are a persuasive indication of effective protection of the DNA chain from enzymatic attack. AFM and TEM Confirmation of Nanocopying of DNA Strands. pBR322 is a circular plasmid DNA with a circumference of ca. 1500 nm. In the AFM observation, closed circular strands were abundantly observed on the mica surface before the surface sol-gel process (Figure 5a). Although titania coating makes the circular morphology bleary (Figure 5b), the width of the strand remains unchanged. The latter result suggests that the thickness of the titania layer on the template DNA is very thin. Small dots with diameters of several tens of nanometers appeared after titania coating (Figure 5b, arrow). They are probably physically adsorbed titania particles that have been formed in the solution of Ti(OnBu)4. After the

8902

Langmuir, Vol. 21, No. 19, 2005

Fujikawa et al.

Figure 5. AFM images of surface morphology: (a) DNA on mica; (b) after surface sol-gel process; (c) after oxygen plasma treatment.

Figure 6. TEM images of titania nanowires: (a) titania-coated DNA; (b) after oxygen plasma treatment. (Insets) Magnified images of the areas surrounded with dotted lines.

oxygen plasma treatment, the line structure becomes clearer again, and the circular morphology of the original plasmid DNA is clearly seen (Figure 5c, arrow). The complementary elemental analysis was performed for DNA specimens before and after titania coating. λ DNA was cast on a quartz plate, and its surface was covered with titania layers by three cycles of the sol-gel process and then subjected to oxygen plasma treatment. XPS elemental analysis was carried out for both the titaniacoated DNA and its plasma-treated sample. The atomic concentration ratio of nitrogen to titanium decreased to 25% of the original value upon plasma treatment (before plasma treatment, N/Ti ) 0.13; after plasma treatment, N/Ti ) 0.033). Although a small amount of the nitrogen was detected by XPS measurement, most of the organic moiety must have been removed from the titania/ DNA complex. Further plasma treatment would remove the organic component completely from the DNA/titania complex. The line structure observed after plasma treatment is composed essentially of titanium oxide, and titania nanotubes are thus formed. The schematic illustration given in Figure 5 shows how a DNA strand is converted to a titania nanotube via titania/DNA complex. Further structural analysis was made by TEM observation. Unstained DNA produced no image prior to titania coating. Titania coating of λ DNA spread on the TEM grid gave a network structure of nanowires in the whole area of the TEM grid (Figure 6a). The wires have widths from 10 to 180 nm. The minimum width observed in the magnified image (Figure 6a, inset) is about 10 nm. From our previous results, one cycle of the surface sol-gel

process gives about a 1 nm thick titania layer on template surfaces,34,35 and three cycles of the sol-gel process would give a titania layer of about 3 nm thick. The observed minimum width of the wire structure is in good agreement with that of the titania-coated single DNA strand, if we assume DNA (about 2 nm diameter) is covered by 2∼3 nm titania wall. Some of the DNA chain apparently formed bundles during the casting process. The larger wires probably correspond to such DNA bundles (black arrows in Figure 6a). The network structure was not observed when titania coating was conducted without DNA templates or when DNA was subjected to TEM observation without titania coating. Thus, such observation of DNA morphology requires both the presence of the template DNA and subsequent titania coating. This morphology is preserved even after oxygen plasma treatment (Figure 6b), though small dots with diameters of several tens of nanometers appeared (indicated by the white triangles). These dots appear to be dust species generated during the plasma process. Elemental Mapping of Titania Nanotubes. Oxygen plasma decomposes the inner organic moiety of the titaniacoated DNA, and the resulting tube interior provides hollow space that can encapsulate ions and molecules. We employed silver ions for the incorporation, since silver ions are easily reduced to form metallic silver by exposure to room light.36 The plasma-treated sample that was (34) Ichinose, I.; Senzu, H.; Kunitake, T. Chem. Mater. 1997, 9, 12961298. (35) Ichinose, I.; Senzu, H.; Kunitake, T. Chem. Lett. 1996, 10, 831832.

Nanocopying of Individual DNA Strands

Langmuir, Vol. 21, No. 19, 2005 8903

Figure 7. TEM and STEM images of the titania nanowires after silver staining. (a) Bright-field image obtained by TEM. The inset is the magnified image of the area surrounded with dotted lines. (b) Dark-field image obtained by STEM. The circled areas a and b are the points to be analyzed for their elemental composition by STEM-EDX point analysis. Point a is on the bright dot on the line; point b does not include the dot. (c) Schematic representation of the nanowires of points a and b in panel b.

immersed in the aqueous AgNO3 for 30 min showed dark dots with a diameter of a few nanometers only in the region of the titania tubes (indicated by white triangles in the enlarged image of Figure 7a). Especially, the linearly aligned arrangement of the particles is clearly found within the network area (indicated by a white oval in the magnified image of Figure 7a), and the diameter of the individual dot was close to the cross-section of DNA. These dots were not clearly observed in the other area. Therefore, we conclude that silver nanoparticles are formed in the cavity within the titania network. From the result of XPS analysis, the cavity that is formed inside the titania network by plasma treatment may not be fully connected to give a tubular hollow structure, and organic residue could remain in the titania. Unfortunately, the extent of the doping is not sufficient for producing continuous silver moiety on/in the titania nanotube. We have to find a way to greatly enhance the amount of silver selectively on the titania layer. From the size similarity between the diameter of the particle and DNA strand, and the observed particle alignment, the morphology of the individual DNA chain has been successfully replicated, though the detailed structural analysis has not been done at the molecular scale. A dark-field image of scanning TEM (STEM) observation (Figure 7b) gave clearer contrasts between the nanotubes and dots. The point elemental analysis by STEM-EDX reveals that the atomic ratio of silver and titanium (Ag/ Ti) in the gray area (point b in Figure 7b) and the white particle (point a in Figure 7b) is 0.89 and 1.85, respectively. The spatial resolution of the STEM-EDX analysis is ca. 1 nm, and clear distinction between the white dot and the wire portion in EDX analysis is rather difficult. Therefore, silver atom was detected even in the network area, in addition to the dot species. (36) Huang, M. H.; Choudrey, A.; Yang, P. Chem. Commun. 2000, 1063-1064

A schematic representation of the cross section of the titania nanotube is illustrated in Figure 7c. The atomic ratio of Ag and Ti at point a (Ag/Ti ) 1.89) is higher than at point b (Ag/Ti ) 0.89). This means that the white dot in the titania nanotube contains many more silver atoms compared with the other area. When silver staining was conducted before removal of the DNA template, such particle morphology was not formed in the domain of the titania nanotube (data not shown). Therefore, the white dot must be silver particles that are produced in the hollow interior of the titania nanotube. The preservation of the network morphology of the titania nanotube and the size agreement between the silver particle and the DNA template are strong indications that the nanostructure of the individual DNA strand was replicated as a titania nanotube. Nanocovering/Nanocopying of Native DNA. In this report, we successfully demonstrated that DNA strands spread on a solid substrate could be covered with an ultrathin titania layer via a surface sol-gel process and that removal of the template DNA led to nanotubes as a positive replica of the individual DNA: see Figure 2. From microscopic and spectroscopic observations, it was clear that the morphology of the DNA strand was preserved. Particularly interesting is the fact that the covered DNA appears to maintain its original molecular features. This was suggested by fluorescence microscopy of the DNA/YOYO-1 complex. The fluorescence microscopic images before and after titania coating were essentially unchanged, showing that titania coating did not change the electronic properties of the complex. The titania layer may chemically interact with surface functional groups of the DNA double helix, but the interaction does not alter the fluorescence property of the intercalated YOYO-1. Another unique feature of titania coating is that the coating can physically isolate the DNA chain from the surrounding medium. Thus, chemical degradation with

8904

Langmuir, Vol. 21, No. 19, 2005

DNase I did not progress after the coating. The titania coating can separate template molecules from the surrounding environment with an ultrathin metal oxide layer. The titania layer is supposedly porous, and permeation of small molecules through the layer to the DNA chain would not be inhibited. In fact, the organic DNA component was almost removed after oxygen plasma treatment, indicating that the small gaseous product arising from the plasma treatment escaped to the outer environment. In contrast, larger protein molecules such as DNase I cannot physically come in contact with the chain, being inactive toward the coated DNA. In addition, silver nanoparticles were not removed by water washing after the incorporation of silver ions. This result indicates that the titan and oxygen bond network is dense enough not to release nanoparticles. The thickness of the titania layer is estimated as 3 nm in the present case. Therefore, the coated DNA gives morphologies that are essentially identical with that of the native (template) DNA. After DNA removal, the remaining titania nanotube should be a loyal replica of the original template, since the thickness of the titania shell is extremely small and this ultrathin wall can sustain the morphology of a hollow DNA replica. It is possible to create unique morphologies of double-stranded DNA by molecular design. Formation of two-dimensional networks37 and lattice38 structures of DNA was reported. When the titania coating approach is applied to such DNA structures, the corresponding nanotube will be fabricated. The design of molecular electronic circuits is a longstanding challenge in nanotechnology research. For this purpose, metal nanowires coated with ultrathin films of nonconductive materials are necessary. Molecular coating with an ultrathin layer of metal oxide will offer a solution for this challenge. Further efforts to achieve this purpose will be discussed in the future. Two-Dimensional Covering vs Three-Dimensional Wrapping. The preceding discussion is basically concerned with covering of 2D molecular patterns on surface. This approach has been applied to 3D molecular systems. We recently reported wrapping of a cationic viologen polymer with ultrathin silicate in solution. In this case, oligomeric sodium silicate was employed as a coating material.31 From suppression of fluorescence quenching as a result of the wrapping, cationic polymer was shown to be surrounded by a silicate sheath, and the stoichiometric wrapping was inferred from the ratio of Si atom (37) Winfree, E.; Liu, F.; Wenzler, L. A.; Seeman, N. C. Nature 1998, 394, 539-544. (38) Yan, H.; Park, S. H.; Finkelstein, G.; Reif, J. H.; LaBean, T. H. Science 2003, 301, 1882-1884.

Fujikawa et al.

and viologen monomer unit. Unfortunately, the molecular weight distribution of the template polymer was wide and the resulting composite was a mixture. Shinkai and coworkers also reported fabrication of DNA-embedded silica by using tetraethoxysilane.39 In both cases, electrostatic interaction between template molecular surfaces and silicate compounds was indispensable for effective coating, and positive charges are needed for template polymers to strongly interact with anionic silica compounds. In the case of the DNA-embedded silica, a cationic amphiphile was used as a binder to adsorb negatively charged silica species onto the DNA surface. This modification may cause a less precise wrapping process, since the spacer (binder amphiphile) molecules will alter the outer shape of the DNA strand. In the current surface sol-gel process, both monolayer adsorption of monomeric titanium butoxide and lateral condensation of the adsorbed titania precursor along the DNA surface occurs at every adsorption cycle. This would produce molecularly precise replication of the individual DNA strand. Conclusion Our present finding shows the effectiveness of nanocopying of the individual DNA strand, as in the structural replication of submicrometer-sized natural substances.18,33 It is noteworthy that simple and structurally featureless metal alkoxide precursors can create structural diversity in a wide range from molecular to macroscopic sizes. This is made possible by outstanding flexibility of amorphous metal oxides to adjust their morphology to the template shape. Thus, metal oxides are superior materials for nanostructure construction by the copying approach. With nanometer precision, the shapes of nanoarchitectures play important roles in producing new phenomena, functions, and properties, and the shape design is, thus, extremely important. Our copying approach can extract the structural information of the template into copying materials with molecular precision. In addition, metal oxides are chemically and physically stable and are convenient materials for practical applications. Acknowledgment. We acknowledge Mr. Yasumitsu Ueki in Hitachi High-Technologies Corporation for skillful measurements in STEM-EDX experiment. We are grateful to Dr. Aiko Nakao for XPS measurement. Associate Professor Koji Nakano at Kyushu University is appreciated for helping in AFM observation. Professor Kuniharu Ijiro at Hokkaido University is acknowledged for useful discussion about DNA spreading procedure. LA051554O (39) Numata, M.; Sugiyasu, K.; Hasegawa, T.; Shinkai, S. Angew. Chem., Int. Ed. 2004, 43, 3279-3283.