- Email: [email protected]
Kelvin probe force microscopy for characterizing doped semiconductors for future sensor applications in nano- and biotechnology
Applied Surface Science 281 (2013) 24–29
Contents lists available at SciVerse ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Kelvin probe force microscopy for characterizing doped semiconductors for future sensor applications in nano- and biotechnology Heidemarie Schmidt a,∗ , Stefan Habicht b,c , Sebastian Feste b,c , Anne-Dorothea Müller d , Oliver G. Schmidt a,e a Chemnitz University of Technology, Department of Materials for Nanoelectronics, Faculty of Electrical Engineering and Information Technology, 09126 Chemnitz, Germany b Forschungszentrum Jülich, Peter Grünberg Institute 9 (PGI-9-IT), 52425 Jülich, Germany c JARA-FIT, Fundamentals of Future Information Technology, Germany d Anfatec Instruments AG, Melanchthonstr. 28, 08606 Oelsnitz, Germany e Institute for Integrative Nanosciences, IFW Dresden, Helmholtzstraße 20, 01069 Dresden, Germany
a r t i c l e
i n f o
Article history: Received 20 October 2012 Received in revised form 25 March 2013 Accepted 17 April 2013 Available online 2 May 2013 Keywords: Kelvin probe force microscopy Electrical biosensors Doped semiconductors Biomaterials Immobilization Transport
a b s t r a c t Kelvin probe force microscopy (KPFM) is one of the most promising non-contact electrical nanometrology techniques to characterize doped semiconductors. By applying a recently introduced explanation of measured KPFM signals, we show the applicability of KPFM to determine and control surface-near electrostatic forces in planar doped silicon and in doped silicon nanostructures. Surface-near electrostatic forces may be used for the immobilization of nano- and biomaterials in future sensor applications in nano- and biotechnology. Additionally, the influence of the electrostatic potential distribution in doped semiconductor nanostructures, e.g. in horizontal Si nanowires, and its influence on the surface-near electrostatic forces are discussed. It is explained how drift and diffusion of injected electrons and holes in intrinsic electric fields influence the detected KPFM signal. For example KPFM is successfully employed to locate p+ p and n+ p junctions along B-doped and As-doped p-Si nanowires, respectively. As an outlook the physical immobilization and the transport of biomaterials above arrays of separately addressable doped semiconductor cells will be discussed. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Biosensors possess an unprecedented potential for the immobilization, modification, transport, and detection of a wide range of analytes in health care, food industry, and environmental monitoring [1]. The interaction between analytes and biosensors can be driven by physical adsorption, chemical covalent immobilization, and by biological immobilization. The chemical covalent and the biological immobilization are considered to be practical and convenient, but are difficult to be controlled and sometimes deactivated. The most widely used biosensors are based on the functionalization of substrates so that only specific analytes bind to the functionalized substrate similar to a lock-and-key mechanism [2]. In this paper we focus on the detection, control, and application of surfacenear electrostatic forces above doped semiconductors for future sensor applications in nano- and biotechnology. Electrically polarizable nano- and biomaterials may differ with respect to their size,
∗ Corresponding author. Tel.: +49 371 531 32481. E-mail address: [email protected] (H. Schmidt). 0169-4332/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apsusc.2013.04.080
shape, electrical polarizability, mass and their tendency to align in electric gradient fields. Therefore, the proposed electrical sensors may be used to immobilize and transport electrically polarizable nano- and biomaterials. Only recently a newly developed magnetic biosensor has been presented [3]. This biosensor benefits from the fact that biological fluids are inherently nonmagnetic and thus provide a low magnetic background. Magnetic detection can be done sensitively even in complex fluids such as whole blood, saliva, or tissue extracts. Furthermore, magnetic actuation by an external field can be used to accelerate the speed of reactions and provide force discrimination by removing non-specifically bound labels for increased selectivity. Typically, the range of surface-near magnetic stray fields above magnetic biosensors is comparatively as large as the range of surface-near electrostatic forces above doped semiconductors with an ultrathin insulating surface layer. A problem which cannot be solved completely is that the vast majority of biomaterials is not magnetizable and must be labelled for use with magnetic biosensors. It has already been shown that electric gradient forces may be used to move single ions above a solid if an electrode is attached to the back side of the solid and if a positionable metallic
H. Schmidt et al. / Applied Surface Science 281 (2013) 24–29
Fig. 1. Schematic illustration of a Kelvin probe force microscopy cantilever above a planar doped semiconductor with a thin oxide layer (grey–blue atomic layer) showing occupied surface states at the interface between the oxide layer and the semiconductor (animated in red) and the same number of unscreened ionized immobile dopant atoms (animated in light-blue). Picture: Sander Münster, 3DKosmos.
conductive tip of an atomic force microscopy is scanned over the solid [4]. As shown by Chiou et al. [5] electrically polarizable nanoand biomaterials may be moved outside a solid carrier between the surface of the carrier and a large-area front side electrode on the m length scale by dielectrophoresis. There the front side electrode is mounted at a distance from the surface of the carrier and does not touch the solid carrier. Note that dielectrophoresis does not use surface-near electrostatic forces, but the electrical gradient established between two electrodes. Nesterov et al. [6] describe an arrangement with a array of electrodes on a chip which allows the generation of a specific pattern of gradient forces when large voltages of up to 100 V are applied between at least two electrodes of the array. The electrical field lines are mainly formed in parallel to the micro-chip. Due to the strong electric field at the edges of the electrodes, it is expected that nano-sized biomaterials are destroyed at the edges of the electrodes. In contrast, the surface-near electrostatic forces above doped semiconductors can be controlled by a small voltage not larger than the bandgap of the semiconductor. For example, in order to control surface-near electrostatic forces above planar doped silicon with a backside electrode, a maximum voltage of 1.1 V has to be applied. In this paper we present a solid carrier without front electrode for the locally controlled immobilization and transport of electrically polarizable nano- and biomaterials using surface-near electrostatic forces (Section 2). As shown in Sections 2–4 Kelvin probe force microscopy (KPFM) is one of the most promising non-contact electrical nanometrology techniques to determine surface-near electrostatic forces above planar doped silicon and above doped horizontal silicon nanowires. As an outlook the physical immobilization and transport of biomaterials on a rectangular array of separately addressable doped semiconductor cells will be discussed in Section 5. 2. Surface-near electrostatic forces The starting point for the development of new solid carriers for the immobilization and transport of nano- and biomaterials is a doped semiconductor (Fig. 1). Optional on the semiconductor surface, a thin insulating layer can be deposited. The occupied surface states at the interface between the oxide layer and the semiconductor (animated in red) and the unscreened ionized immobile dopant atoms (animated in light-blue) form an asymmetric electric
25
dipole with the electric field gradient normal to the semiconductor surface. As explained in the following the surface-near electrostatic forces may be detected by KPFM measurements because the electric field gradient influences the oscillation at the operation frequency fac of the conductive cantilever. This operation frequency is used as the KPFM feedback signal (left, Fig. 1). By applying the appropriate KPFM bias mobile majority charge carriers are injected into the semiconductor surface (animated in orange) and screen the unscreened ionized immobile dopant atoms (centre, Fig. 1). Without the electric field gradient normal to the semiconductor surface, the asymmetric electric dipole and the oscillation of the conductive cantilever at the operation frequency fac is not influenced by surface-near electrostatic forces (right, Fig. 1). To date, KPFM is a standard electrical nanometrology technique applied in various research fields, e.g. to investigate the interface dipole layer formed between a metal surface and alkali chloride thin films [7], surface defects in chalcopyrite solar cell devices [8] and other semiconductors [9,10], dopant profiles in semiconductors [11–13], especially with atomic resolution [14], and doping junctions in semiconductors [13,15–18]. KPFM has also been applied in the field of biotechnology research comprising organic solar cells as well as biomolecules and their interaction. Due to its non-destructive character and high lateral resolution, i.e. 2 nm achieved by means of UHV KPFM reported by Spadafora et al. [19], KPFM qualifies in particular for the investigation of nanostructures. Recently, low temperature UHV KPFM measurements have been performed with a CO-terminated tip on a naphtalocyanine molecule and the charge distribution within a single-molecule charge-transfer complex has been imaged [20]. In this work, a Level-AFM from Anfatec Instruments AG is employed for detecting surface-near electrostatic forces above doped semiconductors by KPFM measurements. It is demonstrated, that the transport of majority charge carriers to the measurement position during the KPFM measurement is crucial for the correct interpretation of the recorded KPFM bias. This knowledge allows a quantitative correlation of the probed lateral KPFM bias variation with the dopant distribution in planar doped semiconductors (Section 3) and in doped semiconductor nanowires (Section 4).
3. Planar doped semiconductors In contrast to state of the art solid carriers for nano- and biomaterials [5] that make use of chemical covalent or biological immobilization, surface-near electrostatic forces of the carrier presented in Fig. 2 are not affected by the environment close to the surface. That is because those surface-near electrostatic forces are chemically and biologically isolated from the environment. The direction of the surface-near electric gradient field is indicated by the direction of the arrows and depends on the charge of occupied surface states at the interface between the oxide layer and the semiconductor. The strength of the surface-near electric gradient field depends on the equal number of occupied surface states and unscreened ionized immobile dopant atoms. Close arrows indicate a strong electric gradient field (Fig. 2). The proposed new carrier is a doped semiconductor with surface-near electrostatic fields and utilizes electrical polarizability of nano- and biomaterials for immobilization and transport of electrically polarizable nano- and biomaterials. The electrical polarizability is a measure of the displacement of a positive relative to a negative charge and for the formation of electric dipoles. For more complex nano- and biomaterials also electric multipoles may form. Electric forces are long-range and strongest on the nanometer-length scale. Without external electric fields the orientation of electrically polarizable nano- and biomaterials is determined by dipole–dipole interactions. Electric dipoles or multipoles act as a torque in a uniform
26
H. Schmidt et al. / Applied Surface Science 281 (2013) 24–29
Fig. 2. Electrically polarizable particles on the surface above (a) an n-type and (b) a p-type semiconductor and an insulating layer (grey). The direction and strength of the electric gradient field are indicated by arrows. A partial negative and positive charge can be found on the surface of the blue and red particle, respectively.
external electric field and orient the electric dipole in the direction of the electric field. In a non-uniform electric field the electrically polarizable nano- and biomaterials experience an electric force which draws them into regions of higher electric field strength and may be used to immobilize them. The dynamics of the electric forces in uniform and non-uniform electric fields determines whether the inertia of immobilized nano- and biomaterials can be overcome and whether the nano- and biomaterials may be transported. In the following we assume that the nano- and biomaterials are electrically polarizable particles with a partial negative charge (−) and a partial positive charge (+). The partial charge which is indicated with the larger of both symbols mainly sits on the particle surface. Therefore, under equilibrium conditions particles with different electrical polarizability can align with different strength and orientation near the surface of the carrier material. In general, a large density of surface states in doped semiconductors leads to a large amount of adsorbed, electrically polarizable particles. On the other hand the species of dopants (donors/acceptors) and the dopant concentration in the locally doped semiconductor determines the direction and strength of surface-near electrostatic forces, respectively. Typically charge densities ranging between 1015 and 1021 cm−3 can be realized by doping and the density of localized charges in surface states typically ranges between 1010 and 1014 cm−2 . According to Coulomb’s law the forces typically decrease with decreasing distance from the carrier surface. Therefore, the thickness of the insulating layer also determines the strength of the surface-near electrostatic forces. Fig. 2 shows the use of a locally doped n-type semiconductor, n+ , n (Fig. 2a) and a locally doped p-type semiconductor, p+ , p (Fig. 2b) with an insulating layer and with surface-near electrostatic forces of opposite sign. If a backside electrode is fixed to this carrier (not shown), the surface-near electrostatic forces can be minimized by applying the corresponding Kelvin bias [13]. Furthermore, in the surface-near region of the n-type semiconductor, G unscreened ionized donors and of the p-type semiconductor, G unscreened acceptors are formed. The G occupied interface states and the G unscreened acceptors and donors form an asymmetric electrostatic dipole. The strength of the surface-near electrostatic forces above the n-type and p-type semiconductor increases with decreasing donor concentration ND and acceptor concentration NA , respectively. The properties of the interface between semiconductor and insulating layer with respect to surface charge per area and time
Fig. 3. Cross-sectional view of a silicon-on-insulator (SOI) structure including top Si layer, 145 nm thick buried oxide and p-Si substrate. The 88 nm thick Si top layer contains the Si pad and horizontal Si nanowires fabricated by means of electron beam lithography, implantation, and annealing. Courtesy of C. Baumgart.
constant of interface states can be controlled by physical, chemical, and thermal treatment of the semiconductor surface prior to the deposition of the insulating layer on top. 4. Doped semiconductor nanowires as future carriers 4.1. Preparation Si nanowires (NWs) were fabricated on (001) silicon-oninsulator (SOI) with a top Si layer thickness of 88 nm and a buried oxide (BOX) of 145 nm thickness. A cross-sectional view of the SOI structure is presented in Fig. 3. The Si substrate is p-type conducting with a small acceptor concentration of approximately 1 × 1015 cm−3 . Owing to the fabrication process, the top Si layer features a low p-type background doping of less than 1 × 1015 cm−3 . Note, that the top Si layer is covered by a native oxide layer of around 1 nm thickness. Thus surface-near electrostatic forces are chemically and biologically isolated from the environment. To provide good electrical contact for KPFM, a 200 nm thick Al layer is deposited on the Si top layer and electrically connected to the biased sample back contact. The length of the Si pad amounts to 10 m. Nanowires with widths ranging from 10 nm to 2 m were patterned by means of electron-beam lithography (EBL) and reactive ion etching. The use of EBL and top-down processing provides several advantages over chemical vapour deposition (CVD) methods, in particular an enhanced control of the location on the sample, and thus controlled alignment. As a consequence, all NWs are very uniform in length and diameter. This makes lithographically fabricated NWs more precise and reproducible which simplifies their integration into a device architecture. Transmission electron microscopy measurements indicate that the NWs have a slightly trapezoidal shape after processing. After transferring the structures into the substrate, a photo-resist implantation mask was defined by EBL in order to protect certain segments of the patterned Si top layer from implantation. For each dopant type, implantation masks have been employed to prepare completely implanted and unimplanted NWs, respectively. The implantation of the structure was carried out by means of ion implantation of As and B at an energy of
H. Schmidt et al. / Applied Surface Science 281 (2013) 24–29
27
10 keV and 2.5 keV and a dose of 2 × 1015 cm−2 and 1 × 1015 cm−2 , respectively. Implantion conditions were chosen such that a sufficiently thick crystalline seed layer remained after the implantation in order to obtain full recrystallization of the top Si layer after annealing. Before activation of dopants, the implantation mask was removed wet chemically, followed by a rapid thermal annealing for 5 s at 1000 ◦ C. Athena simulations of planar SOI structures reveal that for the applied implantation and annealing conditions both, n-type and p-type samples, feature a box-like dopant distribution with comparable concentration of activated dopants. However, the implantation depth differs remarkably for As and B implantation. The simulated acceptor concentration after B implantation amounts to 1 × 1020 cm−3 with an implantation depth of 50 nm. After As implantation the simulated donor concentration amounts to 5 × 1020 cm−3 with an implantation depth of 30 nm. Note, that the 88 nm thick top Si layer is implanted only in the near-surface region while deeper regions remain unimplanted. From the Athena simulations an acceptor concentration of 1 × 1016 cm−3 and a donor concentration of 1 × 1014 cm−3 is deduced in the top Si layer at a depth of 88 nm after B and As implantation, respectively. Note, that KPFM is sensitive to the near-surface asymmetric electric dipole in semiconductors. Thus, no conclusion about the depth-dependent dopant distribution can be drawn from KPFM measurements. In a final preparation step, 200 nm Al was deposited as metallization via a lift-off process. The Al layer enables applying an electrical contact to the top Si layer containing pad and NWs which is required for the KPFM measurements. To short-circuit the insulating BOX layer, the Al layer is connected to the sample back contact by means of silver conductive paint. The deposited Al layer is illustrated schematically in the cross-sectional view of the investigated SOI structure in Fig. 3.
4.2. Detection and control of surface-near electrostatic forces The results of the KPFM measurements across the sample with a B-implanted Si pad and the unimplanted NWs with a low background doping of less than 1015 cm−3 are presented in Fig. 4. The simultaneously probed surface topography and KPFM bias are given in Figs. 4a and b, respectively. The investigated section line across the pad and along the NW is marked as a light blue dashed line in the KPFM bias image and plotted in Fig. 4c. A lateral KPFM bias variation of 590 mV is probed between the B-implanted Si pad and the unimplanted Si NW. To understand this result, the transport of majority charge carriers to the respective measurement positions above the pad and the NW has to be discussed. The B-implanted sample exhibits a horizontal p+ p doping junction at the pad-NW interface due to the fact that the unimplanted NWs are lightly p-type conducting. When probing above the highly p-type conducting Si pad, majority charge carriers, i.e. holes, are injected via the Al contact. The acceptor concentration in the pad amounts to 1 × 1020 cm−3 and is correlated with the KPFM bias by the energy difference Ev − EF (p) = 0 meV. The acceptor concentration in the unimplanted NWs is smaller than 1 × 1015 cm−3 due to the SOI preparation process. When probing above an unimplanted NW, holes are injected via the Al contact and have to be transported through the horizontal doping junction at the padNW interface to reach the measurement position. Due to the fact that the Si pad is highly p-type conducting, the horizontal doping junction works in forward direction regarding transport of holes. Therefore, holes can be accumulated at the measurement position by applying the KPFM bias related to the energy difference of Ev − EF (p) = −560 meV. No additional bias is required for transport through the horizontal p+ p junction. Note, that additional vertical drift due to the depth-dependent dopant distribution only occurs in the implanted Si pad but is not expected in the unimplanted NW. In the KPFM bias section line in Fig. 4c it can be seen that
Fig. 4. KPFM results of the B-implanted Si pad and the unimplanted Si NW. (a) Surface topography, (b) KPFM bias with marked investigated section line (light blue), and (c) KPFM bias section line (averaged over 10 scan lines) compared to the calculated energy difference required for accumulation of holes in the Si NW. Courtesy of C. Baumgart.
the KPFM bias increases by around 50 mV above the Si pad with increasing distance to the Al contact. Above the unimplanted NW the KPFM bias value is reasonably constant and independent on the distance to the Al contact. In conclusion, in the B-implanted sample with the unimplanted NWs a p+ p doping junction occurs, through which majority charge carriers have to be transported. This doping junction works in forward direction regarding holes and the KPFM bias probed above the horizontal Si NW remains correlated with the energy difference required for accumulation of holes. Therefore, the lateral KPFM bias variation probed between the B-implanted pad and unimplanted NWs reflects the energy difference required for respective accumulation of majority charge carriers, i.e. the energy difference between EV − EF (p) = 0 meV and EV − EF (p) = −560 meV. On the contrary to this, for the As-implanted sample, a horizontal n+ p junction has to be considered at the pad-NW interface. In Fig. 5 the results of the KPFM measurements across the As-implanted Si pad and an unimplanted horizontal Si NW are presented. In Fig. 5a and b, the simultaneously probed surface topography and KPFM bias are illustrated, respectively. The light blue dashed line in the KPFM bias image marks the investigated section line along the nanowire, while the corresponding data are plotted in Fig. 5c. Remainders of the photoresist are observed, i.e. at the edge between Si pad and SiO2 (Fig. 5a). This contamination locally distorts the KPFM bias probed across the n+ p junction between the pad and
28
H. Schmidt et al. / Applied Surface Science 281 (2013) 24–29
Fig. 6. Cross-sectional view of an array carrier with four, separately addressable cells 1–4 with the same direction of surface-near electrostatic forces. Here only the charged surface states, the insulating surface layer (grey) and a particle with a partial negative charge (−) and a partial positive charge (+) in the particle centre and on the particle surface, respectively, are shown. Note that the doped semiconductor and the structured bottom electrode are not shown.
Fig. 5. KPFM results of the As-implanted Si pad and the unimplanted Si NW. (a) Surface topography, (b) KPFM bias with marked investigated section line (light blue), and (c) KPFM bias section line (averaged over 5 scan lines) compared to the calculated energy difference required for inversion of the Si pad and accumulation of holes in the Si NW. Courtesy of C. Baumgart.
the NW at the lateral position of 0 m (Fig. 5c). Despite this local distortion, the total lateral KPFM bias variation between pad and NW can be deduced from the plotted section line and amounts to around 1300 mV. For understanding this large KPFM bias variation, the horizontal n+ p junction at the pad-NW interface and its influence on the transport of injected majority charge carriers has to be discussed. When probing across the n+ -Si pad, majority charge carriers, i.e. electrons, are injected via the Al contact. The donor concentration of 5 × 1020 cm−3 is compared with a KPFM bias related to the calculated energy difference EC − EF (n) = 0 meV. When probing across the NW, majority charge carriers, i.e. holes, need to be injected via the Al contact and accumulated at the measurement position. Note, that now the n+ p junction between the n+ -type Si pad and the p-Si NW is not working in forward direction regarding the injection of holes. Thus, to enable transport of holes through the n+ -type Si pad, it needs to be inverted. This is achieved by applying a KPFM bias of −560 meV. Injected holes can now be transported through the inverted Si pad. For additional accumulation of holes in the Si NW, a KPFM bias related to the energy difference EV − EF (p) = −560 meV is required. In total, due to padinversion and hole accumulation in the NW, a lateral KPFM bias
variation of 1120 mV is theoretically expected between the n+ -Si pad and the p-Si NW. In conclusion, in the As-implanted sample with the unimplanted NWs a n+ p doping junction occurs, through which majority charge carriers have to be transported. The probed lateral KPFM bias variation between the As implanted Si pad and the unimplanted NWs reflects the energy necessary to invert the Si pad, i.e. −560 meV, and the KPFM bias required for accumulation of holes in the Si NW, i.e. −560 meV. These findings will have a beneficial impact on the design of doped silicon nanowires and of other doped semiconductor nanostructures for the immobilization and the transport of electrically polarizable biomaterials. 5. Summary and outlook We have presented a new design of a solid carrier with surface-near electrostatic forces for the immobilization and for the transport of electrically polarizable nano- and biomaterials. The carrier can be fabricated from locally doped planar semiconductors or from doped semiconductor nanostructures and is compatible with standard micro-electronics. One advantage is the use of a thin insulating, chemically inert and biocompatible surface layer. Furthermore, the presented carrier allows for recycling. A main advantage is the controllability of time-dependent, locally varying surface-near electrostatic forces by means of a bias applied to a conductive electrode attached to the carrier. By structuring the electrode and the differently implanted regions of the carrier material, nano- and biomaterials of different mass and electrical polarizability can be transported above separately addressable doped semiconductor cells of the array carrier by using the slipand stick mechanism. Preferentially translational transport (Fig. 6) is expected above an array with the same direction of surface-near electrostatic forces in adjacent cells of the carrier. The surface-near electrostatic forces are minimized by applying an external bias to
H. Schmidt et al. / Applied Surface Science 281 (2013) 24–29
29
Acknowledgments H.S. and A.-D. Müller acknowledge funding from Sächsische Aufbaubank (SAB59948/245). H.S. also thanks for financial support from the Deutsche Forschungsgemeinschaft (DFG SCHM1663/4) and C.B. for KPFM measurements and fruitful discussions.
References
Fig. 7. Cross-sectional view of an array carrier with four, separately addressable cells 1–4 with opposite direction of surface-near electrostatic forces in adjacent cells. Here only the charged surface states, the insulating surface layer (grey) and a particle with partial negative charge (−) and a partial positive charge (+) in the particle centre and on the particle surface, respectively, are shown.
the 2nd–4th cell (Fig. 6a), to the 3rd and 4th cell (Fig. 6b) and to the 1st, 3rd and 4th cell (Fig. 6c). Preferentially rotational transport of electrically poalrizable nano- and biomaterials (Fig. 7) is expected above an array carrier with opposite direction of surface-near electrostatic forces in adjacent cells. The surface-near electrostatic forces are minimized by applying an external bias to the 2nd–4th cell (Fig. 7a), to the 3rd and 4th cell (Fig. 7b), and to the 1st, 3rd and 4th cell (Fig. 7c).
[1] J. Niu, G. Jin, Protein Cell 2 (2011) 445–455. [2] J.-Y. Yoon, Introduction to Biosensors: From Electrical Circuits to Immunosensors, Springer, New York, 2013. [3] D. Caicedo, A. Pandharipande, IEEE Sensors Journal 12 (2012) 849. [4] N. Balke, S. Jesse, A.N. Morozovska, E. Eliseev, D.W. Chung, Y. Kim, L. Adamczyk, R.E. García, N. Dudney, S.V. Kalinin, Nature Nanotechnology 5 (2010) 749–754. [5] P.Y. Chiou, A.T. Ohta, M.C. Wu, Nature 436 (2005) 370. [6] A. Nesterov, K. König, T. Felgenhauer, V. Lindenstruth, U. Trunk, S. Fernandez, M. Hausmann, F.R. Bischoff, F. Breitling, V. Stadler, Review of Scientific Instruments 79 (2008) 35106. [7] U. Zerweck, C. Loppacher, T. Otto, S. Grafström, L.M. Eng, Physical Review B 71 (2005) 125424. [8] Th. Glatzel, D.F. Marron, Th. Schedel-Niedrig, S. Sadewasser, M.C. Lux-Steiner, Applied Physics Letters 81 (2002) 2017. [9] Y. Rosenwaks, R. Shikler, Th. Glatzel, S. Sadewasser, Physical Review B 70 (2004) 085320. [10] Th. Glatzel, S. Sadewasser, R. Shikler, Y. Rosenwaks, M.C. Lux-Steiner, Materials Science and Engineering B 102 (2003) 138. [11] N. Duhayon, P. Eyben, M. Fouchier, T. Clarysse, W. Vandervorst, D. Alvarez, S. Schoemann, M. Ciappa, M. Stangoni, W. Fichtner, P. Formanek, M. Kittler, V. Raineri, F. Giannazzo, D. Goghero, Y. Rosenwaks, R. Shikler, S. Saraf, S. Sadewasser, N. Barreau, T. Glatzel, M. Verheijen, S.A.M. Mentink, M. von Sprekelsen, T. Maltezopoulos, R. Wiesendanger, L. Hellemans, Journal of Vacuum Science and Technology B 22 (2004) 385. [12] A.K. Henning, T. Hochwitz, J. Slinkman, J. Never, S. Hoffmann, P. Kaszuba, C. Daghlian, Journal of Applied Physics 77 (1995) 1888. [13] C. Baumgart, M. Helm, H. Schmidt, Physical Review B 80 (2009) 085305. [14] M. Ligowski, D. Moraru, M. Anwar, T. Mizuno, R. Jablonski, M. Tabe, Applied Physics Letters 93 (2008) 142101. [15] B.-Y. Tsui, C.-M. Hsieh, P.-C. Su, S.-D. Tzeng, S. Gwo, Japanese Journal of Applied Physics 47 (2008) 4448. [16] M. Tanimoto, O. Vatel, Journal of Vacuum Science and Technology B 14 (1996) 1547. [17] A. Doukkali, S. Ledain, C. Guasch, J. Bonnet, Applied Surface Science 235 (2004) 507. [18] G.H. Buh, H.J. Chung, C.K. Kim, J.H. Yi, L.T. Yoon, Y. Kuk, Applied Physics Letters 77 (2000) 106. [19] E.J. Spadafora, R. Demadrille, B. Ratier, B. Grévin, Nano Letters 10 (2010) 3337. [20] F. Mohn, L. Gross, N. Moll, G. Meyer, Nature Nanotechnology 7 (2012) 227.
Contents lists available at SciVerse ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Kelvin probe force microscopy for characterizing doped semiconductors for future sensor applications in nano- and biotechnology Heidemarie Schmidt a,∗ , Stefan Habicht b,c , Sebastian Feste b,c , Anne-Dorothea Müller d , Oliver G. Schmidt a,e a Chemnitz University of Technology, Department of Materials for Nanoelectronics, Faculty of Electrical Engineering and Information Technology, 09126 Chemnitz, Germany b Forschungszentrum Jülich, Peter Grünberg Institute 9 (PGI-9-IT), 52425 Jülich, Germany c JARA-FIT, Fundamentals of Future Information Technology, Germany d Anfatec Instruments AG, Melanchthonstr. 28, 08606 Oelsnitz, Germany e Institute for Integrative Nanosciences, IFW Dresden, Helmholtzstraße 20, 01069 Dresden, Germany
a r t i c l e
i n f o
Article history: Received 20 October 2012 Received in revised form 25 March 2013 Accepted 17 April 2013 Available online 2 May 2013 Keywords: Kelvin probe force microscopy Electrical biosensors Doped semiconductors Biomaterials Immobilization Transport
a b s t r a c t Kelvin probe force microscopy (KPFM) is one of the most promising non-contact electrical nanometrology techniques to characterize doped semiconductors. By applying a recently introduced explanation of measured KPFM signals, we show the applicability of KPFM to determine and control surface-near electrostatic forces in planar doped silicon and in doped silicon nanostructures. Surface-near electrostatic forces may be used for the immobilization of nano- and biomaterials in future sensor applications in nano- and biotechnology. Additionally, the influence of the electrostatic potential distribution in doped semiconductor nanostructures, e.g. in horizontal Si nanowires, and its influence on the surface-near electrostatic forces are discussed. It is explained how drift and diffusion of injected electrons and holes in intrinsic electric fields influence the detected KPFM signal. For example KPFM is successfully employed to locate p+ p and n+ p junctions along B-doped and As-doped p-Si nanowires, respectively. As an outlook the physical immobilization and the transport of biomaterials above arrays of separately addressable doped semiconductor cells will be discussed. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Biosensors possess an unprecedented potential for the immobilization, modification, transport, and detection of a wide range of analytes in health care, food industry, and environmental monitoring [1]. The interaction between analytes and biosensors can be driven by physical adsorption, chemical covalent immobilization, and by biological immobilization. The chemical covalent and the biological immobilization are considered to be practical and convenient, but are difficult to be controlled and sometimes deactivated. The most widely used biosensors are based on the functionalization of substrates so that only specific analytes bind to the functionalized substrate similar to a lock-and-key mechanism [2]. In this paper we focus on the detection, control, and application of surfacenear electrostatic forces above doped semiconductors for future sensor applications in nano- and biotechnology. Electrically polarizable nano- and biomaterials may differ with respect to their size,
∗ Corresponding author. Tel.: +49 371 531 32481. E-mail address: [email protected] (H. Schmidt). 0169-4332/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apsusc.2013.04.080
shape, electrical polarizability, mass and their tendency to align in electric gradient fields. Therefore, the proposed electrical sensors may be used to immobilize and transport electrically polarizable nano- and biomaterials. Only recently a newly developed magnetic biosensor has been presented [3]. This biosensor benefits from the fact that biological fluids are inherently nonmagnetic and thus provide a low magnetic background. Magnetic detection can be done sensitively even in complex fluids such as whole blood, saliva, or tissue extracts. Furthermore, magnetic actuation by an external field can be used to accelerate the speed of reactions and provide force discrimination by removing non-specifically bound labels for increased selectivity. Typically, the range of surface-near magnetic stray fields above magnetic biosensors is comparatively as large as the range of surface-near electrostatic forces above doped semiconductors with an ultrathin insulating surface layer. A problem which cannot be solved completely is that the vast majority of biomaterials is not magnetizable and must be labelled for use with magnetic biosensors. It has already been shown that electric gradient forces may be used to move single ions above a solid if an electrode is attached to the back side of the solid and if a positionable metallic
H. Schmidt et al. / Applied Surface Science 281 (2013) 24–29
Fig. 1. Schematic illustration of a Kelvin probe force microscopy cantilever above a planar doped semiconductor with a thin oxide layer (grey–blue atomic layer) showing occupied surface states at the interface between the oxide layer and the semiconductor (animated in red) and the same number of unscreened ionized immobile dopant atoms (animated in light-blue). Picture: Sander Münster, 3DKosmos.
conductive tip of an atomic force microscopy is scanned over the solid [4]. As shown by Chiou et al. [5] electrically polarizable nanoand biomaterials may be moved outside a solid carrier between the surface of the carrier and a large-area front side electrode on the m length scale by dielectrophoresis. There the front side electrode is mounted at a distance from the surface of the carrier and does not touch the solid carrier. Note that dielectrophoresis does not use surface-near electrostatic forces, but the electrical gradient established between two electrodes. Nesterov et al. [6] describe an arrangement with a array of electrodes on a chip which allows the generation of a specific pattern of gradient forces when large voltages of up to 100 V are applied between at least two electrodes of the array. The electrical field lines are mainly formed in parallel to the micro-chip. Due to the strong electric field at the edges of the electrodes, it is expected that nano-sized biomaterials are destroyed at the edges of the electrodes. In contrast, the surface-near electrostatic forces above doped semiconductors can be controlled by a small voltage not larger than the bandgap of the semiconductor. For example, in order to control surface-near electrostatic forces above planar doped silicon with a backside electrode, a maximum voltage of 1.1 V has to be applied. In this paper we present a solid carrier without front electrode for the locally controlled immobilization and transport of electrically polarizable nano- and biomaterials using surface-near electrostatic forces (Section 2). As shown in Sections 2–4 Kelvin probe force microscopy (KPFM) is one of the most promising non-contact electrical nanometrology techniques to determine surface-near electrostatic forces above planar doped silicon and above doped horizontal silicon nanowires. As an outlook the physical immobilization and transport of biomaterials on a rectangular array of separately addressable doped semiconductor cells will be discussed in Section 5. 2. Surface-near electrostatic forces The starting point for the development of new solid carriers for the immobilization and transport of nano- and biomaterials is a doped semiconductor (Fig. 1). Optional on the semiconductor surface, a thin insulating layer can be deposited. The occupied surface states at the interface between the oxide layer and the semiconductor (animated in red) and the unscreened ionized immobile dopant atoms (animated in light-blue) form an asymmetric electric
25
dipole with the electric field gradient normal to the semiconductor surface. As explained in the following the surface-near electrostatic forces may be detected by KPFM measurements because the electric field gradient influences the oscillation at the operation frequency fac of the conductive cantilever. This operation frequency is used as the KPFM feedback signal (left, Fig. 1). By applying the appropriate KPFM bias mobile majority charge carriers are injected into the semiconductor surface (animated in orange) and screen the unscreened ionized immobile dopant atoms (centre, Fig. 1). Without the electric field gradient normal to the semiconductor surface, the asymmetric electric dipole and the oscillation of the conductive cantilever at the operation frequency fac is not influenced by surface-near electrostatic forces (right, Fig. 1). To date, KPFM is a standard electrical nanometrology technique applied in various research fields, e.g. to investigate the interface dipole layer formed between a metal surface and alkali chloride thin films [7], surface defects in chalcopyrite solar cell devices [8] and other semiconductors [9,10], dopant profiles in semiconductors [11–13], especially with atomic resolution [14], and doping junctions in semiconductors [13,15–18]. KPFM has also been applied in the field of biotechnology research comprising organic solar cells as well as biomolecules and their interaction. Due to its non-destructive character and high lateral resolution, i.e. 2 nm achieved by means of UHV KPFM reported by Spadafora et al. [19], KPFM qualifies in particular for the investigation of nanostructures. Recently, low temperature UHV KPFM measurements have been performed with a CO-terminated tip on a naphtalocyanine molecule and the charge distribution within a single-molecule charge-transfer complex has been imaged [20]. In this work, a Level-AFM from Anfatec Instruments AG is employed for detecting surface-near electrostatic forces above doped semiconductors by KPFM measurements. It is demonstrated, that the transport of majority charge carriers to the measurement position during the KPFM measurement is crucial for the correct interpretation of the recorded KPFM bias. This knowledge allows a quantitative correlation of the probed lateral KPFM bias variation with the dopant distribution in planar doped semiconductors (Section 3) and in doped semiconductor nanowires (Section 4).
3. Planar doped semiconductors In contrast to state of the art solid carriers for nano- and biomaterials [5] that make use of chemical covalent or biological immobilization, surface-near electrostatic forces of the carrier presented in Fig. 2 are not affected by the environment close to the surface. That is because those surface-near electrostatic forces are chemically and biologically isolated from the environment. The direction of the surface-near electric gradient field is indicated by the direction of the arrows and depends on the charge of occupied surface states at the interface between the oxide layer and the semiconductor. The strength of the surface-near electric gradient field depends on the equal number of occupied surface states and unscreened ionized immobile dopant atoms. Close arrows indicate a strong electric gradient field (Fig. 2). The proposed new carrier is a doped semiconductor with surface-near electrostatic fields and utilizes electrical polarizability of nano- and biomaterials for immobilization and transport of electrically polarizable nano- and biomaterials. The electrical polarizability is a measure of the displacement of a positive relative to a negative charge and for the formation of electric dipoles. For more complex nano- and biomaterials also electric multipoles may form. Electric forces are long-range and strongest on the nanometer-length scale. Without external electric fields the orientation of electrically polarizable nano- and biomaterials is determined by dipole–dipole interactions. Electric dipoles or multipoles act as a torque in a uniform
26
H. Schmidt et al. / Applied Surface Science 281 (2013) 24–29
Fig. 2. Electrically polarizable particles on the surface above (a) an n-type and (b) a p-type semiconductor and an insulating layer (grey). The direction and strength of the electric gradient field are indicated by arrows. A partial negative and positive charge can be found on the surface of the blue and red particle, respectively.
external electric field and orient the electric dipole in the direction of the electric field. In a non-uniform electric field the electrically polarizable nano- and biomaterials experience an electric force which draws them into regions of higher electric field strength and may be used to immobilize them. The dynamics of the electric forces in uniform and non-uniform electric fields determines whether the inertia of immobilized nano- and biomaterials can be overcome and whether the nano- and biomaterials may be transported. In the following we assume that the nano- and biomaterials are electrically polarizable particles with a partial negative charge (−) and a partial positive charge (+). The partial charge which is indicated with the larger of both symbols mainly sits on the particle surface. Therefore, under equilibrium conditions particles with different electrical polarizability can align with different strength and orientation near the surface of the carrier material. In general, a large density of surface states in doped semiconductors leads to a large amount of adsorbed, electrically polarizable particles. On the other hand the species of dopants (donors/acceptors) and the dopant concentration in the locally doped semiconductor determines the direction and strength of surface-near electrostatic forces, respectively. Typically charge densities ranging between 1015 and 1021 cm−3 can be realized by doping and the density of localized charges in surface states typically ranges between 1010 and 1014 cm−2 . According to Coulomb’s law the forces typically decrease with decreasing distance from the carrier surface. Therefore, the thickness of the insulating layer also determines the strength of the surface-near electrostatic forces. Fig. 2 shows the use of a locally doped n-type semiconductor, n+ , n (Fig. 2a) and a locally doped p-type semiconductor, p+ , p (Fig. 2b) with an insulating layer and with surface-near electrostatic forces of opposite sign. If a backside electrode is fixed to this carrier (not shown), the surface-near electrostatic forces can be minimized by applying the corresponding Kelvin bias [13]. Furthermore, in the surface-near region of the n-type semiconductor, G unscreened ionized donors and of the p-type semiconductor, G unscreened acceptors are formed. The G occupied interface states and the G unscreened acceptors and donors form an asymmetric electrostatic dipole. The strength of the surface-near electrostatic forces above the n-type and p-type semiconductor increases with decreasing donor concentration ND and acceptor concentration NA , respectively. The properties of the interface between semiconductor and insulating layer with respect to surface charge per area and time
Fig. 3. Cross-sectional view of a silicon-on-insulator (SOI) structure including top Si layer, 145 nm thick buried oxide and p-Si substrate. The 88 nm thick Si top layer contains the Si pad and horizontal Si nanowires fabricated by means of electron beam lithography, implantation, and annealing. Courtesy of C. Baumgart.
constant of interface states can be controlled by physical, chemical, and thermal treatment of the semiconductor surface prior to the deposition of the insulating layer on top. 4. Doped semiconductor nanowires as future carriers 4.1. Preparation Si nanowires (NWs) were fabricated on (001) silicon-oninsulator (SOI) with a top Si layer thickness of 88 nm and a buried oxide (BOX) of 145 nm thickness. A cross-sectional view of the SOI structure is presented in Fig. 3. The Si substrate is p-type conducting with a small acceptor concentration of approximately 1 × 1015 cm−3 . Owing to the fabrication process, the top Si layer features a low p-type background doping of less than 1 × 1015 cm−3 . Note, that the top Si layer is covered by a native oxide layer of around 1 nm thickness. Thus surface-near electrostatic forces are chemically and biologically isolated from the environment. To provide good electrical contact for KPFM, a 200 nm thick Al layer is deposited on the Si top layer and electrically connected to the biased sample back contact. The length of the Si pad amounts to 10 m. Nanowires with widths ranging from 10 nm to 2 m were patterned by means of electron-beam lithography (EBL) and reactive ion etching. The use of EBL and top-down processing provides several advantages over chemical vapour deposition (CVD) methods, in particular an enhanced control of the location on the sample, and thus controlled alignment. As a consequence, all NWs are very uniform in length and diameter. This makes lithographically fabricated NWs more precise and reproducible which simplifies their integration into a device architecture. Transmission electron microscopy measurements indicate that the NWs have a slightly trapezoidal shape after processing. After transferring the structures into the substrate, a photo-resist implantation mask was defined by EBL in order to protect certain segments of the patterned Si top layer from implantation. For each dopant type, implantation masks have been employed to prepare completely implanted and unimplanted NWs, respectively. The implantation of the structure was carried out by means of ion implantation of As and B at an energy of
H. Schmidt et al. / Applied Surface Science 281 (2013) 24–29
27
10 keV and 2.5 keV and a dose of 2 × 1015 cm−2 and 1 × 1015 cm−2 , respectively. Implantion conditions were chosen such that a sufficiently thick crystalline seed layer remained after the implantation in order to obtain full recrystallization of the top Si layer after annealing. Before activation of dopants, the implantation mask was removed wet chemically, followed by a rapid thermal annealing for 5 s at 1000 ◦ C. Athena simulations of planar SOI structures reveal that for the applied implantation and annealing conditions both, n-type and p-type samples, feature a box-like dopant distribution with comparable concentration of activated dopants. However, the implantation depth differs remarkably for As and B implantation. The simulated acceptor concentration after B implantation amounts to 1 × 1020 cm−3 with an implantation depth of 50 nm. After As implantation the simulated donor concentration amounts to 5 × 1020 cm−3 with an implantation depth of 30 nm. Note, that the 88 nm thick top Si layer is implanted only in the near-surface region while deeper regions remain unimplanted. From the Athena simulations an acceptor concentration of 1 × 1016 cm−3 and a donor concentration of 1 × 1014 cm−3 is deduced in the top Si layer at a depth of 88 nm after B and As implantation, respectively. Note, that KPFM is sensitive to the near-surface asymmetric electric dipole in semiconductors. Thus, no conclusion about the depth-dependent dopant distribution can be drawn from KPFM measurements. In a final preparation step, 200 nm Al was deposited as metallization via a lift-off process. The Al layer enables applying an electrical contact to the top Si layer containing pad and NWs which is required for the KPFM measurements. To short-circuit the insulating BOX layer, the Al layer is connected to the sample back contact by means of silver conductive paint. The deposited Al layer is illustrated schematically in the cross-sectional view of the investigated SOI structure in Fig. 3.
4.2. Detection and control of surface-near electrostatic forces The results of the KPFM measurements across the sample with a B-implanted Si pad and the unimplanted NWs with a low background doping of less than 1015 cm−3 are presented in Fig. 4. The simultaneously probed surface topography and KPFM bias are given in Figs. 4a and b, respectively. The investigated section line across the pad and along the NW is marked as a light blue dashed line in the KPFM bias image and plotted in Fig. 4c. A lateral KPFM bias variation of 590 mV is probed between the B-implanted Si pad and the unimplanted Si NW. To understand this result, the transport of majority charge carriers to the respective measurement positions above the pad and the NW has to be discussed. The B-implanted sample exhibits a horizontal p+ p doping junction at the pad-NW interface due to the fact that the unimplanted NWs are lightly p-type conducting. When probing above the highly p-type conducting Si pad, majority charge carriers, i.e. holes, are injected via the Al contact. The acceptor concentration in the pad amounts to 1 × 1020 cm−3 and is correlated with the KPFM bias by the energy difference Ev − EF (p) = 0 meV. The acceptor concentration in the unimplanted NWs is smaller than 1 × 1015 cm−3 due to the SOI preparation process. When probing above an unimplanted NW, holes are injected via the Al contact and have to be transported through the horizontal doping junction at the padNW interface to reach the measurement position. Due to the fact that the Si pad is highly p-type conducting, the horizontal doping junction works in forward direction regarding transport of holes. Therefore, holes can be accumulated at the measurement position by applying the KPFM bias related to the energy difference of Ev − EF (p) = −560 meV. No additional bias is required for transport through the horizontal p+ p junction. Note, that additional vertical drift due to the depth-dependent dopant distribution only occurs in the implanted Si pad but is not expected in the unimplanted NW. In the KPFM bias section line in Fig. 4c it can be seen that
Fig. 4. KPFM results of the B-implanted Si pad and the unimplanted Si NW. (a) Surface topography, (b) KPFM bias with marked investigated section line (light blue), and (c) KPFM bias section line (averaged over 10 scan lines) compared to the calculated energy difference required for accumulation of holes in the Si NW. Courtesy of C. Baumgart.
the KPFM bias increases by around 50 mV above the Si pad with increasing distance to the Al contact. Above the unimplanted NW the KPFM bias value is reasonably constant and independent on the distance to the Al contact. In conclusion, in the B-implanted sample with the unimplanted NWs a p+ p doping junction occurs, through which majority charge carriers have to be transported. This doping junction works in forward direction regarding holes and the KPFM bias probed above the horizontal Si NW remains correlated with the energy difference required for accumulation of holes. Therefore, the lateral KPFM bias variation probed between the B-implanted pad and unimplanted NWs reflects the energy difference required for respective accumulation of majority charge carriers, i.e. the energy difference between EV − EF (p) = 0 meV and EV − EF (p) = −560 meV. On the contrary to this, for the As-implanted sample, a horizontal n+ p junction has to be considered at the pad-NW interface. In Fig. 5 the results of the KPFM measurements across the As-implanted Si pad and an unimplanted horizontal Si NW are presented. In Fig. 5a and b, the simultaneously probed surface topography and KPFM bias are illustrated, respectively. The light blue dashed line in the KPFM bias image marks the investigated section line along the nanowire, while the corresponding data are plotted in Fig. 5c. Remainders of the photoresist are observed, i.e. at the edge between Si pad and SiO2 (Fig. 5a). This contamination locally distorts the KPFM bias probed across the n+ p junction between the pad and
28
H. Schmidt et al. / Applied Surface Science 281 (2013) 24–29
Fig. 6. Cross-sectional view of an array carrier with four, separately addressable cells 1–4 with the same direction of surface-near electrostatic forces. Here only the charged surface states, the insulating surface layer (grey) and a particle with a partial negative charge (−) and a partial positive charge (+) in the particle centre and on the particle surface, respectively, are shown. Note that the doped semiconductor and the structured bottom electrode are not shown.
Fig. 5. KPFM results of the As-implanted Si pad and the unimplanted Si NW. (a) Surface topography, (b) KPFM bias with marked investigated section line (light blue), and (c) KPFM bias section line (averaged over 5 scan lines) compared to the calculated energy difference required for inversion of the Si pad and accumulation of holes in the Si NW. Courtesy of C. Baumgart.
the NW at the lateral position of 0 m (Fig. 5c). Despite this local distortion, the total lateral KPFM bias variation between pad and NW can be deduced from the plotted section line and amounts to around 1300 mV. For understanding this large KPFM bias variation, the horizontal n+ p junction at the pad-NW interface and its influence on the transport of injected majority charge carriers has to be discussed. When probing across the n+ -Si pad, majority charge carriers, i.e. electrons, are injected via the Al contact. The donor concentration of 5 × 1020 cm−3 is compared with a KPFM bias related to the calculated energy difference EC − EF (n) = 0 meV. When probing across the NW, majority charge carriers, i.e. holes, need to be injected via the Al contact and accumulated at the measurement position. Note, that now the n+ p junction between the n+ -type Si pad and the p-Si NW is not working in forward direction regarding the injection of holes. Thus, to enable transport of holes through the n+ -type Si pad, it needs to be inverted. This is achieved by applying a KPFM bias of −560 meV. Injected holes can now be transported through the inverted Si pad. For additional accumulation of holes in the Si NW, a KPFM bias related to the energy difference EV − EF (p) = −560 meV is required. In total, due to padinversion and hole accumulation in the NW, a lateral KPFM bias
variation of 1120 mV is theoretically expected between the n+ -Si pad and the p-Si NW. In conclusion, in the As-implanted sample with the unimplanted NWs a n+ p doping junction occurs, through which majority charge carriers have to be transported. The probed lateral KPFM bias variation between the As implanted Si pad and the unimplanted NWs reflects the energy necessary to invert the Si pad, i.e. −560 meV, and the KPFM bias required for accumulation of holes in the Si NW, i.e. −560 meV. These findings will have a beneficial impact on the design of doped silicon nanowires and of other doped semiconductor nanostructures for the immobilization and the transport of electrically polarizable biomaterials. 5. Summary and outlook We have presented a new design of a solid carrier with surface-near electrostatic forces for the immobilization and for the transport of electrically polarizable nano- and biomaterials. The carrier can be fabricated from locally doped planar semiconductors or from doped semiconductor nanostructures and is compatible with standard micro-electronics. One advantage is the use of a thin insulating, chemically inert and biocompatible surface layer. Furthermore, the presented carrier allows for recycling. A main advantage is the controllability of time-dependent, locally varying surface-near electrostatic forces by means of a bias applied to a conductive electrode attached to the carrier. By structuring the electrode and the differently implanted regions of the carrier material, nano- and biomaterials of different mass and electrical polarizability can be transported above separately addressable doped semiconductor cells of the array carrier by using the slipand stick mechanism. Preferentially translational transport (Fig. 6) is expected above an array with the same direction of surface-near electrostatic forces in adjacent cells of the carrier. The surface-near electrostatic forces are minimized by applying an external bias to
H. Schmidt et al. / Applied Surface Science 281 (2013) 24–29
29
Acknowledgments H.S. and A.-D. Müller acknowledge funding from Sächsische Aufbaubank (SAB59948/245). H.S. also thanks for financial support from the Deutsche Forschungsgemeinschaft (DFG SCHM1663/4) and C.B. for KPFM measurements and fruitful discussions.
References
Fig. 7. Cross-sectional view of an array carrier with four, separately addressable cells 1–4 with opposite direction of surface-near electrostatic forces in adjacent cells. Here only the charged surface states, the insulating surface layer (grey) and a particle with partial negative charge (−) and a partial positive charge (+) in the particle centre and on the particle surface, respectively, are shown.
the 2nd–4th cell (Fig. 6a), to the 3rd and 4th cell (Fig. 6b) and to the 1st, 3rd and 4th cell (Fig. 6c). Preferentially rotational transport of electrically poalrizable nano- and biomaterials (Fig. 7) is expected above an array carrier with opposite direction of surface-near electrostatic forces in adjacent cells. The surface-near electrostatic forces are minimized by applying an external bias to the 2nd–4th cell (Fig. 7a), to the 3rd and 4th cell (Fig. 7b), and to the 1st, 3rd and 4th cell (Fig. 7c).
[1] J. Niu, G. Jin, Protein Cell 2 (2011) 445–455. [2] J.-Y. Yoon, Introduction to Biosensors: From Electrical Circuits to Immunosensors, Springer, New York, 2013. [3] D. Caicedo, A. Pandharipande, IEEE Sensors Journal 12 (2012) 849. [4] N. Balke, S. Jesse, A.N. Morozovska, E. Eliseev, D.W. Chung, Y. Kim, L. Adamczyk, R.E. García, N. Dudney, S.V. Kalinin, Nature Nanotechnology 5 (2010) 749–754. [5] P.Y. Chiou, A.T. Ohta, M.C. Wu, Nature 436 (2005) 370. [6] A. Nesterov, K. König, T. Felgenhauer, V. Lindenstruth, U. Trunk, S. Fernandez, M. Hausmann, F.R. Bischoff, F. Breitling, V. Stadler, Review of Scientific Instruments 79 (2008) 35106. [7] U. Zerweck, C. Loppacher, T. Otto, S. Grafström, L.M. Eng, Physical Review B 71 (2005) 125424. [8] Th. Glatzel, D.F. Marron, Th. Schedel-Niedrig, S. Sadewasser, M.C. Lux-Steiner, Applied Physics Letters 81 (2002) 2017. [9] Y. Rosenwaks, R. Shikler, Th. Glatzel, S. Sadewasser, Physical Review B 70 (2004) 085320. [10] Th. Glatzel, S. Sadewasser, R. Shikler, Y. Rosenwaks, M.C. Lux-Steiner, Materials Science and Engineering B 102 (2003) 138. [11] N. Duhayon, P. Eyben, M. Fouchier, T. Clarysse, W. Vandervorst, D. Alvarez, S. Schoemann, M. Ciappa, M. Stangoni, W. Fichtner, P. Formanek, M. Kittler, V. Raineri, F. Giannazzo, D. Goghero, Y. Rosenwaks, R. Shikler, S. Saraf, S. Sadewasser, N. Barreau, T. Glatzel, M. Verheijen, S.A.M. Mentink, M. von Sprekelsen, T. Maltezopoulos, R. Wiesendanger, L. Hellemans, Journal of Vacuum Science and Technology B 22 (2004) 385. [12] A.K. Henning, T. Hochwitz, J. Slinkman, J. Never, S. Hoffmann, P. Kaszuba, C. Daghlian, Journal of Applied Physics 77 (1995) 1888. [13] C. Baumgart, M. Helm, H. Schmidt, Physical Review B 80 (2009) 085305. [14] M. Ligowski, D. Moraru, M. Anwar, T. Mizuno, R. Jablonski, M. Tabe, Applied Physics Letters 93 (2008) 142101. [15] B.-Y. Tsui, C.-M. Hsieh, P.-C. Su, S.-D. Tzeng, S. Gwo, Japanese Journal of Applied Physics 47 (2008) 4448. [16] M. Tanimoto, O. Vatel, Journal of Vacuum Science and Technology B 14 (1996) 1547. [17] A. Doukkali, S. Ledain, C. Guasch, J. Bonnet, Applied Surface Science 235 (2004) 507. [18] G.H. Buh, H.J. Chung, C.K. Kim, J.H. Yi, L.T. Yoon, Y. Kuk, Applied Physics Letters 77 (2000) 106. [19] E.J. Spadafora, R. Demadrille, B. Ratier, B. Grévin, Nano Letters 10 (2010) 3337. [20] F. Mohn, L. Gross, N. Moll, G. Meyer, Nature Nanotechnology 7 (2012) 227.