Enhanced bioconjugation on sputtered palladium nano-thin-film electrode

Bio-conjugation is an essential process that allows biomolecules to immobilize evenly on the surface of an electrode with a suitable density. An effective bio-conjugation process can increase the sensitivity and specificity of biosensing devices.¹ For example, a nanoparticle-modified electrode exhibits a high surface area-to-volume ratio that can increase biomolecule attachment and enhance the signal-to-noise ratio.^2–4 This observation suggests that the surface roughness of an electrode enhances the surface activity during bio-conjugation.⁵ Interestingly, the surface of a lattice (111) exhibits partially coordinated atoms and step edges that reassemble the roughness in nano-scales and provide active sites for the binding of molecules. This postulation was demonstrated by the discovery that more molecules were bound to the surface of Au (111)⁶ and Pt₃Ni (111)⁷ than those of surfaces with other lattices. Accordingly, the electrochemical signal of the electrode might be enhanced, which in turn increases the sensitivity and stability of electrochemical measurements. The surface lattice and roughness of a thin film can be formed on a substrate with a similar surface lattice and roughness. For example, Au (111)⁸ and Ag (111)⁹ were used as templates to generate Si (111).

Direct current (DC) sputtering has been used to generate thin films and nanostructured coatings of various materials, including Au and Ag.^10,11 In this process, metal vapors were evenly sputtered in the ionic liquids to form metal nanoparticles via self-assembly in the liquid.¹² However, metal vapors at high temperature might cause damage to temperature sensitive substrates, such as polyethylene terephthalate (PET). The epitaxial growth of thin films may follow Stranski-Krastanov (S-K)-like, Frank-Van der Merwe-like, or Volmer-Weber-like growth mechanisms.¹³ Among these mechanisms, the S-K-like growth mode is the most acceptable for thin film growth at low temperatures and leads to a two dimensional thin film that exhibits a layer-plus-island structure. Hence, we propose that under certain conditions, a thin film composed of the conducting layer and nano-granules with 1–2 nanoparticles can be generated by sputtering, via the S-K-like mechanism.

The physical nature of DNA-small molecules or DNA-protein complexes is important for the design and screening of anticancer compounds (ACCs).^3,14–16 Interactions between ACCs and the target DNA probe have been studied by various physical and chemical techniques, such as nuclear magnetic resonance,¹⁷ optical,¹⁸ mass spectrometry,¹⁹ and electrochemical approaches.³ Compared to other methods, electrochemical impedance spectroscopy (EIS) is desirable for drug screening, clinical diagnosis, and point-of-care testing.^3,20,21

It has been postulated that surface roughness plays an important role in enabling efficient bioconjugation on the surface of electrodes. The efficiency of the bio-conjugation was shown to be increased on surfaces modified with nanoparticles or that exhibited lattice (111).^8,9,11 Although many metals can possess a similar surface roughness and (111) lattice structure, palladium (Pd) is relatively reactive with a lower work function²² and smaller lattice constant²³ than Au and Pt. Therefore, the generation of Pd nano-thin films (NTFs) via sputtering may retain the surface roughness of the PET substrate and further enhance the bioconjugation on the working electrode.

In order to generate an active metal surface exhibiting roughness at the nano-scale, a flexible PET sheet, U483-125, was used as a substrate to deposit the Pd-NTF with a nanocrystalline structure by sputtering at room temperature. The Pd-NTF-deposited PET sheet was generated using a modified roll-to-roll control sputtering system (Join Well Thin-Film Corp, Ltd., Jhunan, Taiwan) and the Pd target was 80 mm away with a base pressure lower than 8 × 10⁻⁶Torr. The resulting Pd-NTF electrode exhibited a surface resistance of 40 Ω/□.

The surface texture of Pd-NTF deposited on the PET substrate was monitored and analyzed using a field-emission scanning electron microscope (FE-SEM; Hitachi SU8010, Tokyo, Japan), an X-ray diffractometer (XRD, D8, Bruker, Billerica, MA, USA), and an atomic force microscope (AFM, SPA-300HV, Seiko Instruments Inc.) (Fig. 1). The sputtered Pd-NTF exhibited a smooth surface at a low magnification [Fig. 1(a)]. However, under high magnification, numerous nano-islands of 20–100 nm in size were observed [Fig. 1(a), inset]. The thickness of Pd-NTF was determined to be ∼20.8 nm using an SEM cross-sectional image [Fig. 1(b)]. As mentioned above, the surface of PET contains a nano-corrugated structure [Fig. 1(c), inset], becoming a natural template to guide and support the formation of the Pd thin film in a layer-plus-island-like nanostructure under the low temperature sputtering system [Fig. 1(a)]. Hence, the Pd thin film can be stabilized by a flexible PET substrate. Accordingly, the formation of the Pd-NTF on the PET substrate (Pd-NTF electrode) by sputtering may follow the S-K-like growth model. Furthermore, XRD analysis shows that the Pd surface contained Pd (111) in the nanocrystalline form [Fig. 1(c)] and that the granule size [B(2θ)] of Pd in the nano-thin-film, after analysis by the Scherrer equation, i.e., B(2θ) = Kλ/cos(θ), was found to be approximately 10 nm. These results indicate that the sputtering Pd-NTF is formed by two layers of Pd granules and that the surface roughness of PET can be retained.

Similar results can be observed while performing AFM imaging of PET sheet surfaces with and without Pd sputtering. A root mean square (RMS) deviation of ∼0.78 nm was observed for the surface roughness before Pd sputtering [Fig. 1(d), inset]. However, after Pd sputtering, the surface roughness of Pd-NTF at the Pd lift-off side (PET substrate) and at Pd was ∼0.88 nm [Fig. 1(e), upper panel] and ∼0.61 nm [Fig. 1(e), lower panel], respectively. These results suggest that the sputtered Pd-NTF exhibited a surface texture resembling that of the PET substrate.

As shown in Fig. 2(b), that Pd-NTF electrodes exhibit a peak potential difference (ΔE_p) of ∼74.02 mV of cyclic voltammetry (CV) spectra for the [Fe(CN)₆]⁺³/[Fe(CN)₆]⁺⁴ redox couple which is similar to that of the standard electrode (Au electrode) [Fig. 2(a)]. These results suggest that the developed Pd-NTF electrode exhibits electrochemical properties that are comparable to those of the Au electrode.

The bio-molecule conjugation is postulated to be greatly enhanced on the Pd-NTF electrode, which exhibits both nano-scaled roughness and (111) lattices on the surface. EIS is a Faradaic impedance technique that is generally used to detect the surface conductance change²⁴ on the surface of electrodes. Hence, the capability and efficiency of the Pd-NTF electrode for the bio-conjugation of the SP1 probe were investigated by EIS. Using EIS, the efficiency of the conjugation of ssDNA on Pd-NTF electrodes can be evaluated, by measuring the relative changes in the charge transfer resistance, which is designated as the ratio of the net changes in the charge transfer resistance [ΔR_{ct(ssDNA − bare)}] and the charge transfer resistance of the bare electrode [Eq. (1)]

Here, ΔR_{ct(ssDNA−bare)} is the difference in charge transfer resistance between the bare electrode and the electrode with ssDNA: ΔR_{ct (ssDNA − bare)} = [R_{ct (ssDNA)} − R_{ct (bare)}].

In this study, the fabrication of an EIS-based SP1 biosensor was proposed through the immobilization of the SP1 responding element containing a DNA probe (SP1 probe) on the Pd-NTF electrode. The illustration model of the Pd-NTF working electrode for sensing the anti-cancer compound and DNA interaction is shown in Fig. 1(f). The DNA probe without the SP1 responding element (non-SP1 probe)-integrated Pd-NTF electrode (non-SP1 biosensor) was used as a control (Table I). Functionalization was performed by removing 1 μl of 10 μM thiolated ssDNA (5′-SH) and adding it to the Pd-NTF electrodes at room temperature for 15–300 min, for the formation of a DNA self-assembled monolayer (SAM), after which SP1/non-SP1 complementary ssDNA (10 μM) was added onto the SP1/non-SP1 ssDNA-modified electrode at room temperature for 10–15 min, to allow the dsDNA to form, and this was followed by the washing of the dsDNA-coated electrode with double distilled water (d.d.H₂O) twice. Therefore, the optimal time for DNA SAM formation can be estimated via the crossing point of linear extrapolation lines of the growth and saturated regions of the titration curve. As shown in Figs. 3(a) and 3(b), the crossing point of two linear exploration lines is ∼20 min. The excess reagents were removed by rinsing with d.d.H₂O. In contrast, the formation of the DNA SAM on the Au electrodes occurred in approximately 8 h.²⁵ These results indicated that the conjugation of biomolecules on the Pd-NTF electrode is much faster than that on the conventional Au electrodes.

Tetra-O-methyl nordihydroguaiaretic acid (M₄N) has been previously demonstrated to bind specifically at the major groove of the SP1 responding element, a G/C-rich sequence.²⁶ SP1 is a transcription factor that regulates the expression of Cdc2 and surviving genes.²⁷ Hence, it was proposed that M₄N could be used in potential clinical applications in combination with temozolomide and radiation, for glioblastoma treatment.²⁸ To investigate the capability of the SP1 biosensor during DNA drug recognition and affinity analysis, the binding of M₄N was tested.

In order to normalize the variations in different electrodes, a relative charge transfer resistance ratio [Eq. (2)] was used in this study, and the level of binding of drugs at various concentrations can be compared

Here, ΔR_{ct(M4N − DNA)} is the difference in charge transfer resistance between the M₄N-DNA bound electrode and the electrode with double stranded DNA [R_ct(DNA)].

One microliter of M₄N was used to prepare solutions with various concentrations that were incubated with SP1 or non-SP1 dsDNA SAM electrodes for 5 min, which were then washed with d.d.H₂O twice. The changes in impedance in the Pd-NTF electrodes were measured. As shown in the Nyquist plot [Fig. 3(c)], the addition of M₄N (1.20 μM) leads to a decrease in R_ct [blue line and star, in Fig. 3(c)] values at a scale similar to that of ssDNA [orange line and circle, in Fig. 3(c)] on the SP1 biosensor. This result suggests that the activity of M₄N can deform the DNA SAM structure and reduce the surface impedance of the DNA SAM electrode. Further studies showed that the linear range of detection of M₄N binding was between ∼0.1 μM and 10 μM and that the lowest detection limit was ∼0.1 μM [Fig. 3(d)].

Moreover, the binding between M₄N and SP1 on Pd-NTF-based SP1 biosensors [Fig. 4(a), closed square] exhibits a sigmoidal binding curve. In contrast, the relative change in the ΔR_ct ratio was not observed when M₄N was incubated with a non-SP1 DNA probe (closed circles). These results indicate that M₄N interacts with the SP1 element specifically. The binding affinity of M₄N to the SP1 probe was analyzed systematically, and the affinity of M₄N to the SP1 probe was determined by the Hill equation. The apparent dissociation constant (K_d) value for M₄N and SP1 on Pd-NTF based biosensors was found to be 21.55 ± 0.56 μM [Fig. 4(c), closed square], with ∼5 M₄N molecules competing for binding to one SP1 site on the DNA probe.

Doxorubicin (DOX), a known DNA binding ACC,^28–30 and Taxol, a microtubule-binding ACC, were used as positive and negative controls, respectively. DOX can bind to both SP1 [Fig. 4(b), closed square] and non-SP1 probes [Fig. 4(b), closed circle] on the Pd-NTF-based biosensor, during which the K_d values were 2.03 ± 0.10 μM and of 1.86 ± 0.01 μM, respectively. These findings are consistent with those observed in previous studies.^30,31 Taxol has been reported to bind to DNA at an approximate K_d value that was at the mM level.³² Hence, little or no interactions could be seen between Taxol and SP1 or non-SP1 probes on the Pd-NTF-based biosensors if the K_d value was below 10 μM [Fig. 4(b), close triangle for SP1 and open diamond for non-SP1].

In this study, a simple method to generate roughness at an active nano-scale surface in the Pd-NTF on the PET substrate via sputtering has been demonstrated. The generated Pd-NTF allows biomolecules to be conjugated rapidly within 20 min. In addition to bioconjugation, the electric field of the electrode can be enhanced over a thousand fold by altering the degree of roughness of the metal surface to ∼5–20 nm.^33,34 This was demonstrated by the high sensitivity of the fabricated Pd-NTF electrode-based SP1 biosensor for the detection of the SP1 specific drug M₄N, for which the linear range of detection was between ∼0.1 μM and 11 μM, and the lowest detection limit with M₄N was ∼0.1 μM (∼36 pg) can be detected in 1 μl of solution. Furthermore, this SP1 biosensor could effectively distinguish between specific DNA binding drugs, e.g., M₄N, from those that are non-specific, e.g., DOX.

In summary, we have demonstrated that the Pd-NTF-deposited PET electrode exhibits many quality characteristics, including low surface resistance, fast conjugation, high reactivity, and high detection sensitivity, which makes it a potential electrode for the development of electrochemistry-based biosensors with high performance.

This study was supported by an academy and industry cooperation project (MOST 104-2622-M-009-002-CC2; MOST 105-2622-M-009-003-CC2) by MOST, Taiwan and Cheeshin Technology Co. Ltd. (Renamed as Join Well Thin-Film Co., Ltd.). We thank the NSRRC, Taiwan, for the use of 04B SRCD. We also thank and the Center for Advanced Instrumentation, NCTU for the use of XRD. We thank Professor Ming-Hua Hsu, Department of Chemistry, National Changhua University of Education, Changhua, Taiwan for the synthesis of the M₄N compound. This work was particularly supported by the Ministry of Education through the SPROUT Project and Center for Intelligent Drug Systems and Smart Biodevices (IDS²B) of NCTU, Taiwan.