- Email: [email protected]
Nanocrystalline silicon-based oligonucleotide chips
Biosensors and Bioelectronics 22 (2007) 2351–2355
Short communication
Nanocrystalline silicon-based oligonucleotide chips Z.Q. Zhu a , B. Zhu b , J. Zhang a , J.Z. Zhu a,∗ , C.H. Fan c a
b
East China Normal University, China Shanghai Baio Science and Technology Co. Ltd., China c Shanghai Institute of Applied Physics, CAS, China
Received 14 April 2006; received in revised form 9 August 2006; accepted 14 August 2006 Available online 18 September 2006
Abstract A novel oligonucleotide array sensor has been developed with nanocrystalline Si (ncSi) substrates. The ncSi was prepared by electrochemical etching technique. Our study indicated that both the binding capacity and the hybridization efficiency are dependent upon the particle size of ncSi. In contrary, the chips developed with Si substrates exhibit the lower binding capacity and hybridization efficiency. The improved performances of the sensor chips are attributed to the large specific surface area of ncSi compared to the existing conventional techniques. The sensor chips with the ncSi substrate of 13 nm-sized particle can be regenerated and reused for at least 12 times. The oligonucleotide array sensor also shows high stability, which can bear relatively the stringent conditions (e.g. 80 ◦ C, 75% of relative humidity and 3.6 klx of irradiation). © 2006 Published by Elsevier B.V. Keywords: Nanocrystalline silicon; Oligonucleotide array sensor; Binding capacity; Hybridization efficiency; Regenerated and reused times
1. Introduction The DNA array sensor technology (Fodor et al., 1991) has attracted great attention in the last decade due to its enormous application potential and good performance. DNA array sensors are normally referred to the high-density oligonucleotide arrays, or microarrays, i.e., many different DNA snippets are placed and immobilized on either Si or glass slides (Ohji et al., 2002; VoDinh, 1998; Liu et al., 2001). Glass slides and Si are usually employed as the substrates of the DNA array sensors, which have the advantages such as low cost, easy immobilization and simple detection facilities. However, loading capacity of DNA on the glass slides and Si is low due to the limited surface areas, which usually results in the poor sensitivity of the sensor. Nanomaterials have experienced a rapid development in recent years due to their existing and/or potential applications in a variety of technological areas. One unique property of nanocrystallined materials is derived from their large number of grain boundaries compared to coarse-grained polycrystalline ∗
Corresponding author. E-mail addresses: [email protected] (Z.Q. Zhu), [email protected] (B. Zhu), [email protected] (J. Zhang), [email protected] (J.Z. Zhu), [email protected] (C.H. Fan). 0956-5663/$ – see front matter © 2006 Published by Elsevier B.V. doi:10.1016/j.bios.2006.08.007
counterparts. In the nanocrystallined solids, a large fraction of atoms (up to ∼90%) can be the boundary atoms which is dependent upon the particle size. Thus, the interface structure plays an important role in determining the physical and mechanical properties of nanocrystalline materials. In the past decade, favorable applications have been found for hard and wear-resistant ceramic coatings in industrial sectors (Tjong and Chen, 2004). Recently, various kinds of nanocrystallined silicon (porous silicon, nanocrystalline silicon, etc.) had attracted the attention due to its unique properties for biosensor or gas sensor application. Up to now, a variety of methods had been developed to form nanostructured silicon with mechanical robustness and stability, such as magnetron sputtering (Charvet et al., 1999), laser induced chemical vapor deposition (Chiussi et al., 2002), Si+ ion implantation in thermally silica (Rebohle et al., 1996; Liao et al., 1996), and plasma enhanced chemical vapor deposition (Kanzawa et al., 1997; Peng et al., 2000). These methods normally presented poor control of the orientation of nanostructured silicon. Till now, only few papers concerning applications of ncSi in biochips field have been published. In this paper, ncSi was used as the substrate of oligonucleotide chip, which improved performance of oligonucleotide sensors, such as binding capacity, hybridization efficiency, linear range, detection limit and stability.
2352
Z.Q. Zhu et al. / Biosensors and Bioelectronics 22 (2007) 2351–2355
2. Experimental
The average size of ncSi particle was estimated by Raman spectroscopy (Makino et al., 2001).
2.1. Instruments 2.4. Fabrication of oligonucleotide array sensor A DNA Microarray Operating System (Cartesian Technologies Inc., USA), a ScanArray 3000 Laser Confocal Scanner (General Scanning Inc., USA), a PTC-100 PCR Amplifier (MJ Research Inc., USA), XP-2 high resolution surface profiler (AMBIOS Technology Inc., USA), an UV-2000 Ultraviolet Transmission Analyzer (Shanghai Tian Neng Analytical Instruments Factory, China), a scanning electron microscopy (SEM, JEOL-JSM-6700F), a transmission electron microscopy (TEM, JEM200CX), a X-ray diffraction (Philips X PERT-MRD fourcrystal XRD) and a Raman spectroscopy (Jobin Yvon INFINITY micro) were used in the experiments. 2.2. Chemicals and materials 3-Aminopropyltrimethoxysilane (APTMOS) was purchased from Sigma (USA). Custom oligonucleotides were synthesized by Shanghai Sangon Biological Engineering & Technology and Service Co. Ltd. (China). The sequences of DNA fragment used in this work are as follows: • Labeled probe: 5 -AGC GGA TAA CAA TTT CAC ACA GGA-Cy3-3 • Unlabeled probe: 5 -AGC GGA TAA CAA TTT CAC ACA GGA-3 • Complementary matched target: 5 -TCC TGT GTG AAA TTG TTA TCC GCT-Cy3-3 • Non-complementary target: 5 -AGC GGA TAA CAA TTT CAC ACA GGA-Cy3-3 One hundred and fifty millimolar sodium chloride solution including 15 mM sodium citrate, pH 7.0 (SSC), Sodium didecyl sulfate solution (SDS), acetone and alcohol (Analytical Pure) were used as received from Shanghai Chemical Reagent Co. Ltd. (China). All solutions were prepared with deionized water. 2.3. Preparation of ncSi The silicon wafers used in this experiments were p-type boron doped (100)-oriented with thickness of 500 m and resistivity of 0.01, 0.05 and 0.10 cm, respectively. The anodizing solutions for ncSi formation are the mixtures of hydrofluoric acid (40 wt.%), alcohol (99.7 wt.%) and the deionized water with different volume ratio. In order to understand the correlations between the experiment parameters and the film characteristics, nine different experiments were performed with the fixed etching time of 60 min and the current density of 1, 2 and 3 mA/cm2 in combination with HF concentration of 5.0, 10.0 and 25.0 wt.%, respectively. Then, the etched wafers were treated with hydrogen peroxide (3 wt.%) for 10 min in order to form a 2–4 nm-thick of SiO2 layer. The morphology and microstructure of the as-prepared samples were examined by scanning electron microscopy and transmission electron microscopy. The crystal structure of ncSi film was characterized by X-ray diffraction.
2.4.1. Modification of the substrates For comparison, two kinds of substrates had been employed in this study, including ncSi and single crystalline silicon (scSi). Before application, all substrates were immersed in 70% ethanol solution containing 2 M (mol/L) of NaOH for 2 h, followed by rinsing several times with deionized water, and then drying at 110 ◦ C. After cooling down to room temperature, they were immersed into 95% acetone solution containing 1, 2 and 3% APTMOS for 2, 4, 8, 12, 16, 20, 24, 28 and 32 min, respectively, followed by rinsing 10 times with acetone and drying at 110 ◦ C for 0.5 h. After cleaning, the background intensities of the substrates are measured. 2.4.2. Probe spotting Two kinds of oligonucleotide arrays were prepared. One was spotted by the labeled probe for determining the binding capacity of the ncSi to the oligonucleotide probe. The second was spotted by the unlabeled probe for determination of the hybridization efficiency and the dynamic range of the oligonucleotide sensor after hybridization with the complementary target. A series of probe and target solutions with different concentration were prepared by diluting the stock solution with 20 M concentration with the 5× SSC, 0.2% SDS buffer solution. The probe solutions were transferred into the 384-hole plate, which was placed on the spotting sample platform in Cartesian Microarray System to prepare the array chip with a 625 dots/cm2 -density. The array chip was placed into a wet box to hydrate for 5 min and then heated at 100 ◦ C for 30 s. 2.5. Hybridization The labeled hybridization solutions with different concentrations were transferred to the arrays without Cy3 label. It was placed in a wet box and hybridized at 65 ◦ C for 3 h. Then it was washed with the 2× SSC, 0.1% SDS solution for 5 min, followed by washing with the 0.1× SSC, 0.1% SDS solution for 5 min and the 0.1× SSC solution for several times. It was dried in air and ready for the fluorescence detection. 2.6. Detection The oligonucleotide array chips hybridized and dried were scanned at the appreciated condition by using the ScanArray 3000 laser confocal scanner to detect the fluorescent signals generated after hybridization. 3. Results and discussion 3.1. Characterization of ncSi Effects of experimental parameters, i.e., etching time, current density, and HF concentration, on the formation of ncSi were
Z.Q. Zhu et al. / Biosensors and Bioelectronics 22 (2007) 2351–2355
2353
3.4. Binding capacity The binding capacity B of oligonucleotide sensors was calculated as follows: Da × 100% (1) B= Db where Db and Da are the oligonucleotide probe densities immobilized on the substrate before and after washing, respectively. When the relationships between the fluorescence intensities and the oligonucleotide probe densities are linear and the background intensities are the negligible, the formula (1) can be rewritten as: Fa × 100% (2) B= Fb Fig. 1. SEM images of surface structure for the ncSi samples with 10 wt.% HF concentration, 1 mA/cm2 current density for 60 min etching time.
investigated. We found that porous silicon was easier to form at the lower HF concentration, the larger current density and the longer etching time. Experimentally, ncSi of good quality was obtained with current density of 1 mA/cm2 , HF concentration of 10 wt.% under etching time of 60 min (see Fig. 1). The ncSi films are strongly dependent upon the resistivity of the silicon wafers. The results showed that the ncSi films with the average particle sizes of 13, 21, 39 nm were obtained using the wafers with 0.01, 0.05 and 0.10 cm resistivity, respectively, under the above-mentioned conditions. As-prepared ncSi typically had the compact alignments, the uniform orientations, the negligible expansions of lattice constant, the good mechanical robustness and the high stability. 3.2. Fluorescence background of substrates Fluorescence background of substrates was strongly dependent on the APTMOS concentration and the treatment time. The ncSi wafers with the 13 nm-sized particle and the scSi wafers were modified by immersing in 1, 2 and 3% APTMOS solutions for 2, 4, 8, 12, 14, 20, 24, 28 and 32 min, respectively. Fluorescence backgrounds of the obtained substrates were measured. The substrates modified with 1 and 2% APTMOS exhibited the low fluorescence background. In contrast, modification with 3% APTMOS led to the high fluorescence background even with the short treatment time (∼2 min), which was thus not used in following experiments. 3.3. Fluorescence signals of oligonucleotide probe Thirty picomolar labeled probe solution was spotted on the ncSi chip with 13 nm-sized particle and the scSi substrate modified by 1 and 2% APTMOS for 2, 4, 8, 12, 16, 20, 24, 28 and 32 min, respectively. Fluorescence signals were measured. All values of fluorescence intensity shown in this paper are the mean values of 16 elements in the same array. The maximum fluorescence signals were obtained for the sample treated with the 2% APTMOS for 16 min, which were used in sequential experiments.
where Fb and Fa are the fluorescence intensities of the oligonucleotide sensors spotted by probe solution labeled with Cy3 before and after washing, respectively. The fluorescence intensities of the oligonucleotide sensors with different substrates were measured and binding capacities were calculated. The dependences of the binding capacity on particle size of ncSi were investigated. The binding capacities are 93.9 ± 1.8% for 13 nmsized ncSi, 77.7 ± 1.9% for 21 nm-sized ncSi, 71.4 ± 1.0% for 39 nm-sized ncSi and 65.8 ± 1.1% for scSi, respectively. The values behind the symbol “±” are the standard deviations (S.D.). The results show that the binding capacity of ncSi substrate to oligonucleotide increases with decreasing of the particle size. The binding capacity of the scSi substrate is the lowest among them. Of course, the binding capacity and the hybridization efficiency depend on many factors, such as modification methods, immobilization condition, and nucleic acid size, etc. This result is only the binding capacity with given conditions, in particular, in the case of short oligonucleotide. 3.5. Hybridization efficiency The hybridization efficiency of the oligonucleotide sensors was calculated as follows: F1 H1 = × 100% (3) Fa where F1 is the fluorescence intensities of the sensors after the first hybridization. Supposed that the binding capacities of the labeled and the unlabeled probe are identical, the formula (3) can be rewritten as follows: F1 H1 = × 100% (4) BFb The fluorescence intensities of the sensors with different substrates after hybridization were measured and hybridization efficiencies were calculated. The hybridization efficiencies are 90.7 ± 1.2% for 13 nm-sized ncSi, 74.3 ± 1.4% for 21 nm-sized ncSi, 64.2 ± 1.0% for 39 nm-sized ncSi, and 52.6 ± 1.1% for scSi, respectively. The results show that the hybridization efficiencies of ncSi substrate to oligonucleotide
2354
Z.Q. Zhu et al. / Biosensors and Bioelectronics 22 (2007) 2351–2355
Fig. 2. Dependence of the hybridization efficiency of the sensors with different substrate on the multiuse times.
increase with decreasing of the particle size. The hybridization efficiency of the scSi substrate is the lowest among them.
Fig. 3. Dynamic range for the detection of target molecules by using the oligonucleotide array sensor with the ncSi and scSi substrates (upper scale) and signals for non-complementary test (down scale).
3.6. Sensor regeneration and reuse
sensor with ncSi and scSi substrates are 1–35 and 5–20 pM, respectively.
After the first hybridization, the sensor was denatured by boiling at 100 ◦ C for 30 s and then was re-hybridized. The hybridization efficiency H2 was measured. Up to 12 cycles of re-hybridizations were carried out and the dependence of hybridization efficiency against the times of multiple uses was shown in Fig. 2. The hybridization efficiency for the sensor with ncSi substrate degraded slowly with multiple regenerations and reuses, while the hybridization efficiency for the sensor with scSi substrate degraded rapidly and is almost down to zero after fifth round of regeneration. Although the detailed mechanism of what the nanocrystalline Si surface grafted with DNA probes were extremely stable and could survive after repetitive denaturation and re-hybridization has not been fully elucidated yet, we assume that it is attributed to the nanostructure on the substrate surface. Some other substrates with the nanostructured surface, such as the nanocrystallined diamond (Yang et al., 2002) and porous silicon (Bessueille et al., 2005) also exhibit similar characteristics. We speculated that the nanostructured surface provides the higher density of binding sites and the more stable covalent bonds because the nanostructured surface can be regarded as a three-dimension (3d) surface instead of the initial 2d surface for the traditional ones. Further investigations are needed to probe the mechanism in-detailed. 3.7. Dynamic range and detection limit Fig. 3 shows the dynamic range of the oligonucleotide array sensor for the detection of target molecules. The oligonucleotide array sensor with ncSi substrate exhibits the wider linear range and the lower detection limit than those of scSi substrate. The detection limits are 0.5 pM for the oligonucleotide array sensor with ncSi and 1.0 pM for the sensor with scSi substrate. The linear ranges of the oligonucleotide array
3.8. Non-complementary test The oligonucleotide sensors were hybridized with the noncomplementary target solution with concentration of 1–10 M. The fluorescence signals of the sensors for both ncSi and scSi samples are very low (close to the background level), showing in Fig. 3, indicating that the developed ncSi-based oligonucleotide chip has the good selectivity. 3.9. Storage condition The oligonucleotide array sensor was tested after storage in four different conditions. The storage time was 10 days in all cases. The fluorescence intensities of the labeled probe array sensor before and after storage were measured, respectively. The results show that the temperature, the relative humidity and the irradiation that the sensor can bear are up to 80 ◦ C, 75%, 3.6 klx, respectively (signals changes are close to S.D.). It implies that the stability of the oligonucleotide chip is good. 4. Conclusion A novel oligonucleotide chip employing the ncSi as the substrates has been developed. The preliminary experimental results showed that the binding capacity, the hybridization efficiency, the dynamic range and the detection limit of the sensor were improved due to the enormous specific surface area of ncSi compared to the existing conventional techniques. In particular, the sensor can be reused for at least 12 times. Above results are limited in short-chain oligonucleotide. The evaluations for the chip with ncSi substrate using NDA snippets and real samples are undergoing.
Z.Q. Zhu et al. / Biosensors and Bioelectronics 22 (2007) 2351–2355
Acknowledgements The project is supported by National Natural Science Foundation of China) (69925409, 60276036), Shanghai Applied Material Foundation (0202) and Pujiang Project (06PJ14037). References Bessueille, F., Dugas, V., Vikulov, V., Cloarec, J.P., Souteyrand, E.J., Martin, R., 2005. Biosens. Bioelectron. 21, 908–916. Charvet, S., Madelon, R., Gourbilleau, F., Rizk, R., 1999. J. Appl. Phys. 85, 4032–4039. Chiussi, S., Serra, C., Serra, J., Gonzalezz, P., Leon, B., Urban, S., Andra, G., Bergmann, J., Falk, F., Fabbri, F., Fornarini, L., Martelli, S., Rinaldi, F., 2002. Appl. Surf. Sci. 186, 166–172. Fodor, S.P.A., Read, J.L., Pining, M.C., Stryer, L., Lu, A.T., Solas, D., 1991. Science 251, 767–773.
2355
Kanzawa, Y., Kageyama, T., Takeoka, S., Fujii, M., Hayashi, S., Yamamoto, K., 1997. Solid State Commun. 102, 533–537. Liao, L.S., Bao, X.M., Zheng, X.Q., Li, N.S., Min, N.B., 1996. Appl. Phys. Lett. 68, 850–852. Liu, H.C., He, Z.M., Rosenwaks, Z., 2001. Fertil. Steril. 75, 947–955. Makino, T., Yamada, Y., Suzuki, N., Yoshida, T., 2001. J. Appl. Phys. 90 (10), 5075–5080. Ohji, H., Izuo, S., French, J.P., Tsutsumi, K., 2002. Sens. Actuators A 97–98, 744–748. Peng, Y.C., Liu, M., He, Y.L., Li, Y.X., 2000. Chin. J. Semiconductors 21 (3), 308–312 (in Chinese). Rebohle, L., Von Borany, J., Yankov, R.A., Skorupa, W., Yankov, R.A., Tyschenko, I.E., Frob, H., Bohme, T., Leo, K., 1996. Appl. Phys. Lett. 68, 2410–2412. Tjong, S.C., Chen, H., 2004. Mater. Sci. Eng. R 45, 1–88. Vo-Dinh, T., 1998. Sens. Actuators B 51, 52–59. Yang, W.S., Auciello, O., Butler, J.E., Cai, W., Carlisle, J.A., Gerbi, J., Gruen, D.M., Knickerbocker, T., Lasseter, T.L., Russell, J.N., Smith, L.M., Hamers, R.J., 2002. Nat. Mater. 1 (4), 253–257.
Short communication
Nanocrystalline silicon-based oligonucleotide chips Z.Q. Zhu a , B. Zhu b , J. Zhang a , J.Z. Zhu a,∗ , C.H. Fan c a
b
East China Normal University, China Shanghai Baio Science and Technology Co. Ltd., China c Shanghai Institute of Applied Physics, CAS, China
Received 14 April 2006; received in revised form 9 August 2006; accepted 14 August 2006 Available online 18 September 2006
Abstract A novel oligonucleotide array sensor has been developed with nanocrystalline Si (ncSi) substrates. The ncSi was prepared by electrochemical etching technique. Our study indicated that both the binding capacity and the hybridization efficiency are dependent upon the particle size of ncSi. In contrary, the chips developed with Si substrates exhibit the lower binding capacity and hybridization efficiency. The improved performances of the sensor chips are attributed to the large specific surface area of ncSi compared to the existing conventional techniques. The sensor chips with the ncSi substrate of 13 nm-sized particle can be regenerated and reused for at least 12 times. The oligonucleotide array sensor also shows high stability, which can bear relatively the stringent conditions (e.g. 80 ◦ C, 75% of relative humidity and 3.6 klx of irradiation). © 2006 Published by Elsevier B.V. Keywords: Nanocrystalline silicon; Oligonucleotide array sensor; Binding capacity; Hybridization efficiency; Regenerated and reused times
1. Introduction The DNA array sensor technology (Fodor et al., 1991) has attracted great attention in the last decade due to its enormous application potential and good performance. DNA array sensors are normally referred to the high-density oligonucleotide arrays, or microarrays, i.e., many different DNA snippets are placed and immobilized on either Si or glass slides (Ohji et al., 2002; VoDinh, 1998; Liu et al., 2001). Glass slides and Si are usually employed as the substrates of the DNA array sensors, which have the advantages such as low cost, easy immobilization and simple detection facilities. However, loading capacity of DNA on the glass slides and Si is low due to the limited surface areas, which usually results in the poor sensitivity of the sensor. Nanomaterials have experienced a rapid development in recent years due to their existing and/or potential applications in a variety of technological areas. One unique property of nanocrystallined materials is derived from their large number of grain boundaries compared to coarse-grained polycrystalline ∗
Corresponding author. E-mail addresses: [email protected] (Z.Q. Zhu), [email protected] (B. Zhu), [email protected] (J. Zhang), [email protected] (J.Z. Zhu), [email protected] (C.H. Fan). 0956-5663/$ – see front matter © 2006 Published by Elsevier B.V. doi:10.1016/j.bios.2006.08.007
counterparts. In the nanocrystallined solids, a large fraction of atoms (up to ∼90%) can be the boundary atoms which is dependent upon the particle size. Thus, the interface structure plays an important role in determining the physical and mechanical properties of nanocrystalline materials. In the past decade, favorable applications have been found for hard and wear-resistant ceramic coatings in industrial sectors (Tjong and Chen, 2004). Recently, various kinds of nanocrystallined silicon (porous silicon, nanocrystalline silicon, etc.) had attracted the attention due to its unique properties for biosensor or gas sensor application. Up to now, a variety of methods had been developed to form nanostructured silicon with mechanical robustness and stability, such as magnetron sputtering (Charvet et al., 1999), laser induced chemical vapor deposition (Chiussi et al., 2002), Si+ ion implantation in thermally silica (Rebohle et al., 1996; Liao et al., 1996), and plasma enhanced chemical vapor deposition (Kanzawa et al., 1997; Peng et al., 2000). These methods normally presented poor control of the orientation of nanostructured silicon. Till now, only few papers concerning applications of ncSi in biochips field have been published. In this paper, ncSi was used as the substrate of oligonucleotide chip, which improved performance of oligonucleotide sensors, such as binding capacity, hybridization efficiency, linear range, detection limit and stability.
2352
Z.Q. Zhu et al. / Biosensors and Bioelectronics 22 (2007) 2351–2355
2. Experimental
The average size of ncSi particle was estimated by Raman spectroscopy (Makino et al., 2001).
2.1. Instruments 2.4. Fabrication of oligonucleotide array sensor A DNA Microarray Operating System (Cartesian Technologies Inc., USA), a ScanArray 3000 Laser Confocal Scanner (General Scanning Inc., USA), a PTC-100 PCR Amplifier (MJ Research Inc., USA), XP-2 high resolution surface profiler (AMBIOS Technology Inc., USA), an UV-2000 Ultraviolet Transmission Analyzer (Shanghai Tian Neng Analytical Instruments Factory, China), a scanning electron microscopy (SEM, JEOL-JSM-6700F), a transmission electron microscopy (TEM, JEM200CX), a X-ray diffraction (Philips X PERT-MRD fourcrystal XRD) and a Raman spectroscopy (Jobin Yvon INFINITY micro) were used in the experiments. 2.2. Chemicals and materials 3-Aminopropyltrimethoxysilane (APTMOS) was purchased from Sigma (USA). Custom oligonucleotides were synthesized by Shanghai Sangon Biological Engineering & Technology and Service Co. Ltd. (China). The sequences of DNA fragment used in this work are as follows: • Labeled probe: 5 -AGC GGA TAA CAA TTT CAC ACA GGA-Cy3-3 • Unlabeled probe: 5 -AGC GGA TAA CAA TTT CAC ACA GGA-3 • Complementary matched target: 5 -TCC TGT GTG AAA TTG TTA TCC GCT-Cy3-3 • Non-complementary target: 5 -AGC GGA TAA CAA TTT CAC ACA GGA-Cy3-3 One hundred and fifty millimolar sodium chloride solution including 15 mM sodium citrate, pH 7.0 (SSC), Sodium didecyl sulfate solution (SDS), acetone and alcohol (Analytical Pure) were used as received from Shanghai Chemical Reagent Co. Ltd. (China). All solutions were prepared with deionized water. 2.3. Preparation of ncSi The silicon wafers used in this experiments were p-type boron doped (100)-oriented with thickness of 500 m and resistivity of 0.01, 0.05 and 0.10 cm, respectively. The anodizing solutions for ncSi formation are the mixtures of hydrofluoric acid (40 wt.%), alcohol (99.7 wt.%) and the deionized water with different volume ratio. In order to understand the correlations between the experiment parameters and the film characteristics, nine different experiments were performed with the fixed etching time of 60 min and the current density of 1, 2 and 3 mA/cm2 in combination with HF concentration of 5.0, 10.0 and 25.0 wt.%, respectively. Then, the etched wafers were treated with hydrogen peroxide (3 wt.%) for 10 min in order to form a 2–4 nm-thick of SiO2 layer. The morphology and microstructure of the as-prepared samples were examined by scanning electron microscopy and transmission electron microscopy. The crystal structure of ncSi film was characterized by X-ray diffraction.
2.4.1. Modification of the substrates For comparison, two kinds of substrates had been employed in this study, including ncSi and single crystalline silicon (scSi). Before application, all substrates were immersed in 70% ethanol solution containing 2 M (mol/L) of NaOH for 2 h, followed by rinsing several times with deionized water, and then drying at 110 ◦ C. After cooling down to room temperature, they were immersed into 95% acetone solution containing 1, 2 and 3% APTMOS for 2, 4, 8, 12, 16, 20, 24, 28 and 32 min, respectively, followed by rinsing 10 times with acetone and drying at 110 ◦ C for 0.5 h. After cleaning, the background intensities of the substrates are measured. 2.4.2. Probe spotting Two kinds of oligonucleotide arrays were prepared. One was spotted by the labeled probe for determining the binding capacity of the ncSi to the oligonucleotide probe. The second was spotted by the unlabeled probe for determination of the hybridization efficiency and the dynamic range of the oligonucleotide sensor after hybridization with the complementary target. A series of probe and target solutions with different concentration were prepared by diluting the stock solution with 20 M concentration with the 5× SSC, 0.2% SDS buffer solution. The probe solutions were transferred into the 384-hole plate, which was placed on the spotting sample platform in Cartesian Microarray System to prepare the array chip with a 625 dots/cm2 -density. The array chip was placed into a wet box to hydrate for 5 min and then heated at 100 ◦ C for 30 s. 2.5. Hybridization The labeled hybridization solutions with different concentrations were transferred to the arrays without Cy3 label. It was placed in a wet box and hybridized at 65 ◦ C for 3 h. Then it was washed with the 2× SSC, 0.1% SDS solution for 5 min, followed by washing with the 0.1× SSC, 0.1% SDS solution for 5 min and the 0.1× SSC solution for several times. It was dried in air and ready for the fluorescence detection. 2.6. Detection The oligonucleotide array chips hybridized and dried were scanned at the appreciated condition by using the ScanArray 3000 laser confocal scanner to detect the fluorescent signals generated after hybridization. 3. Results and discussion 3.1. Characterization of ncSi Effects of experimental parameters, i.e., etching time, current density, and HF concentration, on the formation of ncSi were
Z.Q. Zhu et al. / Biosensors and Bioelectronics 22 (2007) 2351–2355
2353
3.4. Binding capacity The binding capacity B of oligonucleotide sensors was calculated as follows: Da × 100% (1) B= Db where Db and Da are the oligonucleotide probe densities immobilized on the substrate before and after washing, respectively. When the relationships between the fluorescence intensities and the oligonucleotide probe densities are linear and the background intensities are the negligible, the formula (1) can be rewritten as: Fa × 100% (2) B= Fb Fig. 1. SEM images of surface structure for the ncSi samples with 10 wt.% HF concentration, 1 mA/cm2 current density for 60 min etching time.
investigated. We found that porous silicon was easier to form at the lower HF concentration, the larger current density and the longer etching time. Experimentally, ncSi of good quality was obtained with current density of 1 mA/cm2 , HF concentration of 10 wt.% under etching time of 60 min (see Fig. 1). The ncSi films are strongly dependent upon the resistivity of the silicon wafers. The results showed that the ncSi films with the average particle sizes of 13, 21, 39 nm were obtained using the wafers with 0.01, 0.05 and 0.10 cm resistivity, respectively, under the above-mentioned conditions. As-prepared ncSi typically had the compact alignments, the uniform orientations, the negligible expansions of lattice constant, the good mechanical robustness and the high stability. 3.2. Fluorescence background of substrates Fluorescence background of substrates was strongly dependent on the APTMOS concentration and the treatment time. The ncSi wafers with the 13 nm-sized particle and the scSi wafers were modified by immersing in 1, 2 and 3% APTMOS solutions for 2, 4, 8, 12, 14, 20, 24, 28 and 32 min, respectively. Fluorescence backgrounds of the obtained substrates were measured. The substrates modified with 1 and 2% APTMOS exhibited the low fluorescence background. In contrast, modification with 3% APTMOS led to the high fluorescence background even with the short treatment time (∼2 min), which was thus not used in following experiments. 3.3. Fluorescence signals of oligonucleotide probe Thirty picomolar labeled probe solution was spotted on the ncSi chip with 13 nm-sized particle and the scSi substrate modified by 1 and 2% APTMOS for 2, 4, 8, 12, 16, 20, 24, 28 and 32 min, respectively. Fluorescence signals were measured. All values of fluorescence intensity shown in this paper are the mean values of 16 elements in the same array. The maximum fluorescence signals were obtained for the sample treated with the 2% APTMOS for 16 min, which were used in sequential experiments.
where Fb and Fa are the fluorescence intensities of the oligonucleotide sensors spotted by probe solution labeled with Cy3 before and after washing, respectively. The fluorescence intensities of the oligonucleotide sensors with different substrates were measured and binding capacities were calculated. The dependences of the binding capacity on particle size of ncSi were investigated. The binding capacities are 93.9 ± 1.8% for 13 nmsized ncSi, 77.7 ± 1.9% for 21 nm-sized ncSi, 71.4 ± 1.0% for 39 nm-sized ncSi and 65.8 ± 1.1% for scSi, respectively. The values behind the symbol “±” are the standard deviations (S.D.). The results show that the binding capacity of ncSi substrate to oligonucleotide increases with decreasing of the particle size. The binding capacity of the scSi substrate is the lowest among them. Of course, the binding capacity and the hybridization efficiency depend on many factors, such as modification methods, immobilization condition, and nucleic acid size, etc. This result is only the binding capacity with given conditions, in particular, in the case of short oligonucleotide. 3.5. Hybridization efficiency The hybridization efficiency of the oligonucleotide sensors was calculated as follows: F1 H1 = × 100% (3) Fa where F1 is the fluorescence intensities of the sensors after the first hybridization. Supposed that the binding capacities of the labeled and the unlabeled probe are identical, the formula (3) can be rewritten as follows: F1 H1 = × 100% (4) BFb The fluorescence intensities of the sensors with different substrates after hybridization were measured and hybridization efficiencies were calculated. The hybridization efficiencies are 90.7 ± 1.2% for 13 nm-sized ncSi, 74.3 ± 1.4% for 21 nm-sized ncSi, 64.2 ± 1.0% for 39 nm-sized ncSi, and 52.6 ± 1.1% for scSi, respectively. The results show that the hybridization efficiencies of ncSi substrate to oligonucleotide
2354
Z.Q. Zhu et al. / Biosensors and Bioelectronics 22 (2007) 2351–2355
Fig. 2. Dependence of the hybridization efficiency of the sensors with different substrate on the multiuse times.
increase with decreasing of the particle size. The hybridization efficiency of the scSi substrate is the lowest among them.
Fig. 3. Dynamic range for the detection of target molecules by using the oligonucleotide array sensor with the ncSi and scSi substrates (upper scale) and signals for non-complementary test (down scale).
3.6. Sensor regeneration and reuse
sensor with ncSi and scSi substrates are 1–35 and 5–20 pM, respectively.
After the first hybridization, the sensor was denatured by boiling at 100 ◦ C for 30 s and then was re-hybridized. The hybridization efficiency H2 was measured. Up to 12 cycles of re-hybridizations were carried out and the dependence of hybridization efficiency against the times of multiple uses was shown in Fig. 2. The hybridization efficiency for the sensor with ncSi substrate degraded slowly with multiple regenerations and reuses, while the hybridization efficiency for the sensor with scSi substrate degraded rapidly and is almost down to zero after fifth round of regeneration. Although the detailed mechanism of what the nanocrystalline Si surface grafted with DNA probes were extremely stable and could survive after repetitive denaturation and re-hybridization has not been fully elucidated yet, we assume that it is attributed to the nanostructure on the substrate surface. Some other substrates with the nanostructured surface, such as the nanocrystallined diamond (Yang et al., 2002) and porous silicon (Bessueille et al., 2005) also exhibit similar characteristics. We speculated that the nanostructured surface provides the higher density of binding sites and the more stable covalent bonds because the nanostructured surface can be regarded as a three-dimension (3d) surface instead of the initial 2d surface for the traditional ones. Further investigations are needed to probe the mechanism in-detailed. 3.7. Dynamic range and detection limit Fig. 3 shows the dynamic range of the oligonucleotide array sensor for the detection of target molecules. The oligonucleotide array sensor with ncSi substrate exhibits the wider linear range and the lower detection limit than those of scSi substrate. The detection limits are 0.5 pM for the oligonucleotide array sensor with ncSi and 1.0 pM for the sensor with scSi substrate. The linear ranges of the oligonucleotide array
3.8. Non-complementary test The oligonucleotide sensors were hybridized with the noncomplementary target solution with concentration of 1–10 M. The fluorescence signals of the sensors for both ncSi and scSi samples are very low (close to the background level), showing in Fig. 3, indicating that the developed ncSi-based oligonucleotide chip has the good selectivity. 3.9. Storage condition The oligonucleotide array sensor was tested after storage in four different conditions. The storage time was 10 days in all cases. The fluorescence intensities of the labeled probe array sensor before and after storage were measured, respectively. The results show that the temperature, the relative humidity and the irradiation that the sensor can bear are up to 80 ◦ C, 75%, 3.6 klx, respectively (signals changes are close to S.D.). It implies that the stability of the oligonucleotide chip is good. 4. Conclusion A novel oligonucleotide chip employing the ncSi as the substrates has been developed. The preliminary experimental results showed that the binding capacity, the hybridization efficiency, the dynamic range and the detection limit of the sensor were improved due to the enormous specific surface area of ncSi compared to the existing conventional techniques. In particular, the sensor can be reused for at least 12 times. Above results are limited in short-chain oligonucleotide. The evaluations for the chip with ncSi substrate using NDA snippets and real samples are undergoing.
Z.Q. Zhu et al. / Biosensors and Bioelectronics 22 (2007) 2351–2355
Acknowledgements The project is supported by National Natural Science Foundation of China) (69925409, 60276036), Shanghai Applied Material Foundation (0202) and Pujiang Project (06PJ14037). References Bessueille, F., Dugas, V., Vikulov, V., Cloarec, J.P., Souteyrand, E.J., Martin, R., 2005. Biosens. Bioelectron. 21, 908–916. Charvet, S., Madelon, R., Gourbilleau, F., Rizk, R., 1999. J. Appl. Phys. 85, 4032–4039. Chiussi, S., Serra, C., Serra, J., Gonzalezz, P., Leon, B., Urban, S., Andra, G., Bergmann, J., Falk, F., Fabbri, F., Fornarini, L., Martelli, S., Rinaldi, F., 2002. Appl. Surf. Sci. 186, 166–172. Fodor, S.P.A., Read, J.L., Pining, M.C., Stryer, L., Lu, A.T., Solas, D., 1991. Science 251, 767–773.
2355
Kanzawa, Y., Kageyama, T., Takeoka, S., Fujii, M., Hayashi, S., Yamamoto, K., 1997. Solid State Commun. 102, 533–537. Liao, L.S., Bao, X.M., Zheng, X.Q., Li, N.S., Min, N.B., 1996. Appl. Phys. Lett. 68, 850–852. Liu, H.C., He, Z.M., Rosenwaks, Z., 2001. Fertil. Steril. 75, 947–955. Makino, T., Yamada, Y., Suzuki, N., Yoshida, T., 2001. J. Appl. Phys. 90 (10), 5075–5080. Ohji, H., Izuo, S., French, J.P., Tsutsumi, K., 2002. Sens. Actuators A 97–98, 744–748. Peng, Y.C., Liu, M., He, Y.L., Li, Y.X., 2000. Chin. J. Semiconductors 21 (3), 308–312 (in Chinese). Rebohle, L., Von Borany, J., Yankov, R.A., Skorupa, W., Yankov, R.A., Tyschenko, I.E., Frob, H., Bohme, T., Leo, K., 1996. Appl. Phys. Lett. 68, 2410–2412. Tjong, S.C., Chen, H., 2004. Mater. Sci. Eng. R 45, 1–88. Vo-Dinh, T., 1998. Sens. Actuators B 51, 52–59. Yang, W.S., Auciello, O., Butler, J.E., Cai, W., Carlisle, J.A., Gerbi, J., Gruen, D.M., Knickerbocker, T., Lasseter, T.L., Russell, J.N., Smith, L.M., Hamers, R.J., 2002. Nat. Mater. 1 (4), 253–257.