Phase, current, absorbance, and photoluminescence of double and triple metal ion-doped synthetic and salmon DNA thin films

The essence of bionanotechnology involves combining nanotechnology with biology to manipulate structures on the nanometer scale with appreciable accuracy. Researchers have proven biological molecules (e.g., nucleic acids, proteins, and membranes) to be reliable and simple tools for fabrication and construction of nanostructures and devices. Among biological molecules, DNA (a genetic information carrier at a fundamental level for all species) has been highlighted as a new engineering material. Due to the capability for programming the DNA base sequence (the Watson–Crick complementarity base pairing rule) to obtain complex structural assemblies and the feasibility of using DNA for attaching functionalized nanomaterials (e.g., nanoparticles, ions, carbon-based materials, fluorescence molecules, proteins, and drugs) by exploiting the intrinsic characteristics or simple modification of DNA, researchers today consider DNA as one of the most promising biomaterials. By properly considering these advantages, researchers can exploit DNA for use in many potential applications [1].

Functionalized DNA molecules appear advantageous due to the specificity displayed for binding of the ligand to the components of DNA, imparting new properties to DNA having functionalized materials. DNA can be used as efficient templates of various functionalized materials, such as: assembling gold nanoparticles into aggregates using DNA as linkers, transition metal complexes, doxorubicin, vitamin B₂, and organic dye molecules as DNA binding agents [2–8]. Considering their simple fabrication process for electrical and optical applications, researchers consider the use of metal ions (M²⁺) as specifically functionalized materials in conjunction with DNA to be one of the promising candidates. We can treat M²⁺ (e.g., Cu²⁺, Ni²⁺, and Co²⁺) as ligands that bind to DNA selectively (intercalation between a base pair and electrostatic binding to ${{\rm{PO}}}_{4}^{-}$) to construct a system with novel functionalities [9–15]. In addition, metal complexes could regulate DNA functions by targeting gene sequences, acting as checkpoints for transcription and translation [16–18].

Despite reports of semiconducting and superconducting electronic behaviors in specific environments, DNA is generally known to be an insulator at room temperature [19–22]. Previously, we developed a simple method to dope individual M²⁺ (having distinct physical characteristics) to synthetic DNA lattices, but not in combinations of M²⁺ (which can provide multifunctional physical characteristics in an M²⁺-doped DNA thin film) [23–25]. Consequently, developing a method to find the optimum concentration of various combinations of M²⁺ in DNA, as well as establishing their basic characterization, represent crucial steps toward developing applications using multiple types of different M²⁺-doped DNA thin films (either synthetic or natural).

Here, we present three main aspects: first, fabrication of the synthetic double-crossover (DX) DNA lattices and natural salmon DNA (SDNA) thin films doped with 3 combinations of double M²⁺-doped groups (Co²⁺–Ni²⁺, Cu²⁺–Co²⁺, and Cu²⁺–Ni²⁺) and a triple M²⁺-doped group (Cu²⁺–Ni²⁺–Co²⁺) at various concentrations of M²⁺ ([M²⁺]); second, evaluation of the optimum concentration of M²⁺ ([M²⁺]_O) at which the phase of M²⁺-doped DX DNA lattices changes from crystalline to amorphous; and lastly, delineation of the optoelectric characteristics of multiple M²⁺-doped SDNA thin films. We confirmed a phase diagram (containing information on the degree of phase change of the DX DNA lattices, controlled by combinations of [M²⁺]) by atomic force microscope (AFM) images. In addition, we measured the current (for measuring conductivity behavior), absorption (to determine the binding of M²⁺ to DNA), and photoluminescence (PL, for an understanding of fluorescence characteristics) by the probe station, spectrophotometer, and fluorimeter, respectively.

2.1. Preparation of the M²⁺-doped DX DNA lattices grown by the substrate-assisted growth (SAG) method

High-performance liquid chromatography purified synthetic oligonucleotides for DX lattices, divalent metal ions (copper ion solution Cu(NO₃)₂, nickel ion NiCl₂, and cobalt ion CoCl₂) for doping, and substrates (O₂ plasma treated glass and fused silica) were purchased from Bioneer (Daejeon, Korea), Sigma-Aldrich (St. Louis, MO, USA), and Dasom (Gyeonggi, Korea), respectively. An equimolar concentration (50 nM) of 8 different individual DX DNA strands, and proper concentrations of M²⁺ (see figure 2), were added and an O₂ plasma treated glass substrate was submerged into an AXYGEN test-tube containing physiological 1× TAE/Mg²⁺ buffer (40 mM Tris, 20 mM acetic acid, 1 mM EDTA (pH 8.0), and 12.5 mM magnesium acetate) before annealing. The total volume of the mixture was maintained at 250 μl. The glass substrate (5.0 × 5.0 mm²) was immersed into the test-tube in order to facilitate the growth of M²⁺-doped DX DNA lattices during the course of the annealing. Finally, the test-tube was placed in a Styrofoam box with 2 L of boiled water cooled from 95 °C to room temperature over a period of 24 h, for the growth of M²⁺-doped DX DNA lattices on the glass substrate [26–29] (first row in figure 1).

Zoom In Zoom Out Reset image size

2.2. Preparation of the M²⁺-doped SDNA thin films by the drop-casting method

2.3. AFM imaging

The morphological stability of M²⁺-doped DX DNA lattices grown on glass was studied by AFM imaging. The M²⁺-doped DX DNA sample substrate was fixed on a metal puck using instant glue and 30 μl of 1× TAE/Mg²⁺ buffer was pipetted onto the substrate, while another 20 μl of 1× TAE/Mg²⁺ was dispensed on a silicon nitride AFM tip (NP-S10, Veeco Inc., NY, USA). The AFM images were obtained by using a Multimode Nanoscope III (Veeco Inc., NY, USA) in fluid tapping mode. (Figures 2 and S2 is available online at stacks.iop.org/NANO/28/405702/mmedia in the online supplementary data (OSD).)

Zoom In Zoom Out Reset image size

2.4. Current–voltage (I–V) measurement

Silver contacts were deposited as electrodes (channel gap distance of ∼1 mm) on the glass substrate having an M²⁺-doped SDNA thin film. I as the function of applied V was measured within a range between −3.0 and 3.0 V, using a probe station with two terminals (4200–SCS, Keithley Instruments Inc., OH, USA) under atmospheric pressure and room temperature. (Figure 3).

Zoom In Zoom Out Reset image size

2.5. Ultraviolet–visible (UV–vis.) spectroscopy measurement

The optical absorbance of SDNA thin films (on fused silica in the visible and ultraviolet regions) was measured without and with various [M²⁺] by a spectrophotometer (Cary 5G, Varian, CA, USA). The spectrophotometer was equipped with two light sources (a deuterium arc lamp for near-infrared and Vis., and a quartz W–halogen lamp for UV) and two detectors (a cooled PbS detector for the near-infrared region, and a photomultiplier tube for the Vis. and UV regions) (figure 4).

Zoom In Zoom Out Reset image size

2.6. PL measurement

The PL emission and excitation spectra of SDNA thin films were collected without and with various [M²⁺] on fused silica by using a fluorimeter (FS-2, Scinco, Seoul, Korea), equipped with the Xe-arc lamp at a power of 25 W and room temperature. The emission spectra were acquired by exciting an M²⁺-doped SDNA thin film at a wavelength of 290 nm, and the excitation spectrum was obtained at a fixed emission wavelength, 330 nm (figures 5, S3 in OSD).

Zoom In Zoom Out Reset image size

Doping of DX DNA lattices (a representative synthetic DNA nanostructure with relatively higher production yield) and SDNA thin films (natural DNA available in large quantities at low cost) were carried out by using Cu²⁺, Ni²⁺, and Co²⁺ (chosen due to relatively high conductivities and distinctive optical spectra, with stable ionic states compared to the pristine DNA). A DX tile—dimension of 4 nm × 12 nm in width and length—based DNA lattice used in this study had two repeating DX tiles (DX-1 and DX-2, as shown in figure S1 and tables S1, S2 in OSD). The DX DNA lattices were grown on a given substrate by the SAG method, with details of sample preparation explained in the experimental section. During the course of annealing via SAG, DNA crystallization was achieved by a growth process (including tile seeding, nucleation, and growth), resulting in polycrystalline DX DNA lattices with 50 nM of (DX) (enough for full coverage on the substrate). In addition, we used pristine DNA derived from marine salmon for fabricating SDNA thin films via the drop-casting method. During the doping process, M²⁺ could be freely intercalated between a base pair and bound into the negatively charged phosphate backbones (shown in figure 1).

Previously, we constructed DX DNA lattices doped with single M²⁺, studying their structural stabilities in a relationship with various [M²⁺] [24, 25, 30]. [M²⁺]_O,S was estimated from single M²⁺-doped DX DNA lattices (e.g., [Cu²⁺]_O,S = 6.0 mM, [Ni²⁺]_O,S = 1.5 mM, and [Co²⁺]_O,S = 1.0 mM). Up to [M²⁺]_O,S, the crystalline phase of the DX DNA lattices with single M²⁺ was guaranteed, while the amorphous phase of the lattices occurred above [M²⁺]_O,S [24]. In this study, multiple M²⁺-doped DX DNA lattices were fabricated at a fixed $[{{\rm{M}}}_{1}^{2+}]$ with a varying $[{{\rm{M}}}_{2}^{2+}]$ for double doping, and at fixed $[{{\rm{M}}}_{1}^{2+}]$ and $[{{\rm{M}}}_{2}^{2+}]$ with a varying $[{{\rm{M}}}_{3}^{2+}]$ for triple. Here ${{\rm{M}}}_{1}^{2+},$ ${{\rm{M}}}_{2}^{2+},$ and ${{\rm{M}}}_{3}^{2+}$ represented different species of M²⁺ (i.e., Cu²⁺, Ni²⁺, and Co²⁺). Analyzing the obtained AFM images gave ${[{{\rm{M}}}_{2}^{2+}]}_{{\rm{O}}}$ for double doping and ${[{{\rm{M}}}_{3}^{2+}]}_{{\rm{O}}}$ for triple doping, followed by the completion of the phase diagram (a graphical representation of the crystalline and amorphous phases as a function of [M²⁺]). DX DNA lattices with multiple M²⁺ dopants might have the inherent characteristics of each species of M²⁺ while fine-tuning the functionality of the DX DNA lattices by controlling the combinations of M²⁺ was possible (advantageous compared to single doping).

We could expect the crystalline domain region with multiple M²⁺-doped DX DNA lattices to follow a relationship between the each species of [M²⁺] and [M²⁺]_O (for single M²⁺-doping, [M²⁺]_O was estimated from the data of a single M²⁺-doped DX DNA lattice). For double ${{\rm{M}}}_{1}^{2+}$ and ${{\rm{M}}}_{2}^{2+}$-doping, the crystalline, and amorphous phases were to meet the following conditions, $\tfrac{[{{\rm{M}}}_{1}^{2+}]}{{[{{\rm{M}}}_{1}^{2+}]}_{{\rm{O}},{\rm{S}}}}+\tfrac{[{{\rm{M}}}_{2}^{2+}]}{{[{{\rm{M}}}_{2}^{2+}]}_{{\rm{O}},{\rm{S}}}}\leqslant 1,$ and $\tfrac{[{{\rm{M}}}_{1}^{2+}]}{{[{{\rm{M}}}_{1}^{2+}]}_{{\rm{O}},{\rm{S}}}}+\tfrac{[{{\rm{M}}}_{2}^{2+}]}{{[{{\rm{M}}}_{2}^{2+}]}_{{\rm{O}},{\rm{S}}}}\gt 1,$ respectively. The expected lines in a phase diagram of ${{\rm{M}}}_{1}^{2+}$ and ${{\rm{M}}}_{2}^{2+}$-doped DX DNA lattices as shown in figures 2(a)–(c) separated the crystalline and amorphous phases of the DX DNA lattices in 2-dimensional phase diagrams. Here, the ratio of ${[{{\rm{M}}}_{2}^{2+}]}_{{\rm{O}},{\rm{S}}}$ to ${[{{\rm{M}}}_{1}^{2+}]}_{{\rm{O}},{\rm{S}}}$ indicated the slope magnitude of the expected line. In the 3-dimensional phase diagram shown in figure 2(d), a crystalline triangular surface was defined by a condition, $\tfrac{[{{\rm{M}}}_{1}^{2+}]}{{[{{\rm{M}}}_{1}^{2+}]}_{{\rm{O}},{\rm{S}}}}+\tfrac{[{{\rm{M}}}_{2}^{2+}]}{{[{{\rm{M}}}_{2}^{2+}]}_{{\rm{O}},{\rm{S}}}}+\tfrac{[{{\rm{M}}}_{3}^{2+}]}{{[{{\rm{M}}}_{3}^{2+}]}_{{\rm{O}},{\rm{S}}}}=1.$ Even though the expected surface was not drawn for simplicity, but it is expected to be similar to the light yellow shaded region obtained by the experimental data.

Figure 2 depicted the characteristic AFM images for double and triple M²⁺-doped DX DNA lattices with phase diagrams. AFM images verified the crystalline and amorphous phases controlled by different combinations of [M²⁺] in the DX DNA lattices. [M²⁺] for double-doping was used in the following combinations: (i) Co²⁺ 0.5 mM fixed, Ni²⁺ varied from 0.2 to 1.0 mM by increments of 0.2 mM, and phase transition occurred from crystalline (up to 0.8 mM ≡ [Ni²⁺]_O) to amorphous (at 1.0 mM of [Ni²⁺]) (indicated by a red dot located on the red line); (ii) Cu²⁺ 3.0 mM fixed, and Co²⁺ varied from 0.2 mM to 0.8 mM, and phase transition occurred between 0.6 mM (≡[Co²⁺]_O) and 0.8 mM of [Co²⁺]; and (iii) Cu²⁺ 3.0 mM fixed, and Ni²⁺ varied 0.2 to 0.8 mM, while phase transition occurred between 0.6 mM (≡[Ni²⁺]_O) and 0.8 mM of [Ni²⁺]. By using the definition for percentage of discrepancy between the expected and the experimental values, calculated as (∣[M²⁺]_O,Expected − [M²⁺]_{O,Experimental}∣/[M²⁺]_O,Expected) ×100, we obtained relatively larger percentages of discrepancy, (i) 6.7%, (ii) 20.0, and (iii) 20.0, due to small sample sets (we analyzed 3 sets of each M²⁺ combination). Similarly, we prepared two cases of triple doping: (i) Co²⁺ varied from 0.2 mM to 0.8 mM by increments of 0.2 mM, at fixed Cu²⁺ 2.0 mM and Ni²⁺ 0.8 mM, and phase transition from crystalline to amorphous occurred between 0.6 mM (≡Co²⁺]_O) and 0.8 mM of [Co²⁺]) (indicated by a blue dot located on the surface of the yellow shaded region); and (ii) Co²⁺ varied from 0.2 mM to 1.0 mM, at a fixed Cu²⁺ 2.0 mM and Ni²⁺ 0.2 mM, where phase transition occurred between 0.8 mM (≡[Co²⁺]_O) and 1.0 mM of [Co²⁺]) (indicated by a green dot placed on the surface of the yellow shaded region).

The light yellow shaded and the white regions denoted in graphs (figure 2) indicated the crystalline and amorphous domains of M²⁺-doped DX DNA lattices, respectively. The yellow lines in the AFM images in the crystal domain showed the M²⁺-doped DX DNA lattice boundaries, while the insets in those graphs contained noise-filtered reconstructed images by fast Fourier-transformation that indicated the periodicities in the DX lattices (crystalline phase). The periodicity of DX DNA lattices was not visible in the amorphous phase (confirmed by fast Fourier-transformation). Additional AFM images for M²⁺-doped DX DNA lattices for the rest of [M²⁺] are shown in figure S2 in OSD. In the presence of the excess M²⁺(=[M²⁺] –[M²⁺]_O), deformation of DX DNA lattices occurred due to the improper and nonspecific bindings of excess M²⁺ into the DNA molecules [24, 25, 30, 31].

The measurements of M²⁺-doped SDNA thin films (∼1.5 μm of thickness) carried out for I (conductivity), UV–vis. absorption (binding of M²⁺ to DNA), and PL (fluorescence characteristic) showed consistent and reproducible characteristics compared to the M²⁺-doped DX DNA lattices (∼2.0 nm of thickness) due to the substantial sample volume at a given area. SDNA served as a readily available natural source of DNA, extracted from salmon fish and processed enzymatically to yield relatively higher amounts of pure DNA molecules. Figure 3 showed the I–V characteristics (measured by a two terminal source-drain configuration) and resistance (R, calculated from Ohm's law, R = V/I) of multiple M²⁺-doped SDNA thin films. I curves for multiple M²⁺-doped SDNA thin films were roughly linearly proportional to the applied V, majorly due to the hopping of charge carriers through the dopants in DNA duplex [32, 33]. From figure 3, we noticed that I heavily relied on the various combinations of [M²⁺], as well as the different types of M²⁺. Figure 3(f) showed a graphical representation for R against varying [M²⁺] of multiple M²⁺-doped SDNA thin films at a fixed 3 V. The minimum R values obtained, for a double doping group with varying Ni²⁺ (Co²⁺, Ni²⁺) from 0.2 mM to 1 mM at fixed Co²⁺ 0.5 mM (Cu²⁺ 3.0 mM, Cu²⁺ 3.0 mM) were 0.12 GΩ (0.75 GΩ, 0.13 GΩ) at 0.8 mM of [Ni²⁺] (0.6 mM of [Co²⁺], 0.6 mM of [Ni²⁺])—coincidently identical to [Ni²⁺]_O ([Co²⁺]_O, [Ni²⁺]_O) obtained from phase diagrams discussed in figure 2. Similarly the minimum R obtained for a triple doping group, with varying Co²⁺ (Co²⁺) from 0.2 mM to 1.0 mM at fixed Cu²⁺ 2.0 mM and Ni²⁺ 0.2 mM (Cu²⁺ 2.0 mM and Ni²⁺ 0.8 mM) were 0.10 GΩ (0.91 GΩ) at 0.8 mM of [Co²⁺] (0.6 mM of [Co²⁺])—also identical to [Co²⁺]_O ([Co²⁺]_O). From these observations, we could redefine [M²⁺]_O measured through SDNA thin films as [M²⁺] at which R was minimum at a fixed V. We observed that R decreased (I increased) with increasing [M²⁺] up to [M²⁺]_O (due to the proper DNA site binding of M²⁺), while R increased (I decreased) above [M²⁺]_O (improper binding of excess M²⁺ on the DNA) [24].

Figure 4 depicted the absorbance (as the function of the scanned wavelength) and relative peak intensity (as a function of varying [M²⁺]) of multiple M²⁺-doped SDNA thin films (DNA molecules were known to exhibit absorption in the range of 250–280 nm). We measured 5 sets of M²⁺-doped combinations on SDNA thin films, 3 for double doping (Co²⁺0.5 mM, Cu²⁺ 3.0 mM, and Cu²⁺ 3.0 mM fixed; while Ni²⁺, Co²⁺, and Ni²⁺ varied, respectively), and 2 for triple doping (Cu²⁺ 2.0 mM and Ni²⁺ 0.2 mM fixed, while Co²⁺ varied; and Cu²⁺ 2.0 mM and Ni²⁺ 0.8 mM fixed, while Co²⁺ varied). Figure 4(f) summarized the absorbance peak intensities as a function of [M²⁺], by analyzing absorbance peaks at 260 nm. Absorbance spectra of SDNA thin film revealed quenching of the absorbance intensity after M²⁺-doping, due to the binding of M²⁺ with DNA (which compromised the stability of the DNA duplex). Absorption intensity gradually decreased up to a certain [M²⁺]_O,ABS {for a double doping group, minimum peak intensity obtained with varying Ni²⁺ (Co²⁺, Ni²⁺) at fixed Co²⁺ 0.5 mM (Cu²⁺ 3.0 mM, Cu²⁺ 3.0 mM) at 0.4 mM of [Ni²⁺] (0.8 mM of [Co²⁺], 0.8 mM of [Ni²⁺]); similarly for a triple doping group, minimum peak intensity obtained with varying Co²⁺ (Co²⁺) at fixed Cu²⁺ 2.0 mM and Ni²⁺ 0.2 mM (Cu²⁺ 2.0 mM and Ni²⁺ 0.8 mM) at 0.4 mM of [Co²⁺] (0.4 mM of [Co²⁺])}, and increased above [M²⁺]_O,ABS. Despite slight differences between [M²⁺]_O,ABS and [M²⁺]_O (measured from AFM and I–V), overall trends followed as expected.

The typical excitation spectra monitored at a fixed emission wavelength, 330 nm for pristine SDNA and M²⁺-doped SDNA thin films (data not shown) revealed a strong peak at ∼290 nm. The emission spectra for SDNA thin films without M²⁺ and with various combinations of [M²⁺] were collected at an excitation wavelength of 290 nm. PL involved the proper excitation energy to excite an electron to a singlet excited state from the singlet ground state upon absorbing the photon energy. A desirable root for relaxation was an internal conversion in which relaxation from a higher vibrational energy level of an excited state occurred toward the zero level of that excited state by the non-radiative energy transfer. The inter-system crossing was a type of non-radiative energy transfer that occurred when an electron relaxed to the triplet excited state from the singlet excited state. PL could be observed from the release of a photon upon relaxation of the electron from the triplet excited state. Thus, energy transfer occurred through an internal conversion process within the excited singlet state, then from the excited singlet state to the triplet state through the inter-system crossing, and finally to emissive states [25].

Figure 5 showed the PL emission spectra of double and triple doped M²⁺-doped SDNA thin films (Co²⁺ fixed and Ni²⁺ varied, Cu²⁺ fixed and Co²⁺ varied, and Cu²⁺ fixed and Ni²⁺ varied, respectively for double doping; Cu²⁺, Ni²⁺ fixed and Co²⁺ varied for triple). The PL spectra for M²⁺-doped SDNA thin films, as well as for pristine SDNA thin film, showed the peak centered at ∼330 nm, while we observed enhancing and quenching of PL intensities by controlling [M²⁺]. PL intensities were enhanced at Co²⁺ 0.5 mM with Ni²⁺ 0.8 mM followed by quenching with increasing [Ni²⁺] (figure 5(a)), and Cu²⁺ 3 mM with Co²⁺ 0.8 mM followed by quenching with increasing [Co²⁺] (figure 5(b)) for double doping; and Cu²⁺ 2 mM and Ni²⁺ 0.2 mM with Co²⁺ 1 mM (figure 5(c)), and Cu²⁺ 2 mM and Ni²⁺ 0.8 mM with Co²⁺ 0.6 mM (figure 5(d)) for triple, followed by quenching with increasing [Co²⁺]. Based on the PL graphs shown in figure 5, we expected the energy transfer from the excited states of the SDNA to the emissive states of the bound M²⁺ to take place from the lowest triplet excited state in M²⁺-doped SDNA thin films up to [M²⁺]_O at a given condition. Above [M²⁺]_O, the energy transfer might occur from M²⁺ to the SDNA through cross relaxation due to the excess M²⁺ bound non-specifically to SDNA molecules. Consequently, the energy transfer results revealed the PL quenching of M²⁺-doped SDNA thin films.

We developed a simple method to dope various M²⁺ to synthetic DNA lattices which provided multifunctional physical characteristics in an M²⁺-doped DNA thin film. We (i) fabricated synthetic DX DNA lattices and natural SDNA thin films doped with combinations of double and triple M²⁺-doped groups at various [M²⁺]; (ii) evaluated the optimum [M²⁺] at which phase of M²⁺-doped DX DNA lattices changed from crystalline to amorphous; and lastly, (iii) studied the optoelectric characteristics of multiple M²⁺-doped SDNA thin films. We analyzed a phase diagram by the theoretical estimation of the [M²⁺], confirmed experimentally by AFM. In addition, the measurement of current, absorption, and PL lead to gain an understanding concerning conductivity, chemical binding, and fluorescence. We realized that the current and PL intensities were observed maximum but the absorbance intensities showed minimum at around [M²⁺]_O for double and triple M²⁺-doped SDNA thin films. Consequently, multiple M²⁺-doped DNA thin films may be used in physical devices and chemical sensors with tunable multiple functionalities in the near future.

This work was supported by the Center for Advanced Meta-Materials (CAMM) as Global Frontier Project (CAMM-2014M3A6B3063707, CAMM-2015M3A6B3071411) and by the National Research Foundation (NRF) of Korea (2016R1D1A1B03933768, 2017R1A2B4010955, 2017R1D1A1B03035053) funded by the Ministry of Science, ICT and Future Planning of the Korean government.