Letter pubs.acs.org/NanoLett
Carrier Mobility Enhancement of Tensile Strained Si and SiGe Nanowires via Surface Defect Engineering J. W. Ma,† W. J. Lee,† J. M. Bae,† K. S. Jeong,† S. H. Oh,† J. H. Kim,† S.-H. Kim,‡ J.-H. Seo,‡ J.-P. Ahn,‡ H. Kim,§ and M.-H. Cho*,† †
Institute of Physics and Applied Physics, Yonsei University, Seoul 120-749, Korea Nano Analysis Center, KIST, Seoul 130-650, Korea § School of Advanced Materials Science and Engineering, Sungkyunkwan University, Suwon 440-746, Korea ‡
S Supporting Information *
ABSTRACT: Changes in the carrier mobility of tensile strained Si and SiGe nanowires (NWs) were examined using an electrical push-to-pull device (E-PTP, Hysitron). The changes were found to be closely related to the chemical structure at the surface, likely defect states. As tensile strain is increased, the resistivity of SiGe NWs deceases in a linear manner. However, the corresponding values for Si NWs increased with increasing tensile strain, which is closely related to broken bonds induced by defects at the NW surface. Broken bonds at the surface, which communicate with the defect state of Si are critically altered when Ge is incorporated in Si NW. In addition, the number of defects could be significantly decreased in Si NWs by incorporating a surface passivated Al2O3 layer, which removes broken bonds, resulting in a proportional decrease in the resistivity of Si NWs with increasing strain. Moreover, the presence of a passivation layer dramatically increases the extent of fracture strain in NWs, and a significant enhancement in mobility of about 2.6 times was observed for a tensile strain of 5.7%. KEYWORDS: Si, SiGe, nanowire, strain, surface defect, passivation
A
mong the various nanomaterials, one-dimensional (1D) semiconducting nanowires (NWs) have been investigated for use in field-effect transistor (FET) devices because of the successful implementation of a building block approach for the assembly of nanodevices and device arrays for highly integrated devices.1,2 Semiconductor NWs can have an on/off current ratio that varies by many orders of magnitude with small changes in gate voltage. They also offer the possibility of ballistic transport since electron scattering in one dimension is reduced to a considerable extent.3 In general, the performance of NW-based electronic devices often rival devices that are constructed from the best bulk and epitaxial single-crystal semiconductors. Si and Ge NWs can enhance the performance of FETs without jeopardizing device reliability. FETs based on boron and phosphorus-doped Si and Ge NWs have been subjects of active study during the past several years since these two materials offer great compatibility with industrial chip technology.4−8 The performances of nanodevices based on Si and Ge NWs are comparable to the best performance of the corresponding planar devices.9,10 Strain is important in Si technology and is applied as a means to enhance electron or hole mobility. In the case of relaxation time, mobility is affected by the effective mass me,h and deformation potential Ea © 2015 American Chemical Society
μe,h =
e⟨τ ⟩ 1 ∝ 2 me,h Ea me,h
Therefore, a smaller effective mass or deformation potential can be directly related to higher currents and superior device performance.11 Moreover, lower effective mass and higher velocity mean that carriers would be less sensitive to surface roughness, an important factor in NWs and other onedimensional structures.12 Strain has been used within planar Si metal-oxide-semiconductors FET (MOSFET) to enhance mobility by reducing electron intervalley scattering and redistributing carriers within conduction valleys to a lower effective mass.13 He et al. reported that the enhancement in the mobility of Si NWs can be attributed to strain that is up to a few orders of magnitude greater than that of bulk Si.14 Many researchers have reported on the electrical properties of Si and Ge NWs based on the application of strain engineering through both simulation and experimental data. The findings indicate that the conductivity or mobility of Si and Ge NWs are increased due to their modified effective mass or deformation potential. However, at the end of the Si device Received: April 26, 2015 Revised: October 12, 2015 Published: October 22, 2015 7204
DOI: 10.1021/acs.nanolett.5b01634 Nano Lett. 2015, 15, 7204−7210
Letter
Nano Letters
Figure 1. (a) SEM image and (b) HRTEM image of Si NWs. (c) HRTEM image of the SiGe NWs. Insets of (b) and (c) show the corresponding FFT images of the Si NW and SiGe NW with zone axis along [110], respectively. FFT images indicating the growth direction of NWs is [111], and the crystalline structure is single crystal and fcc phase. EDX result in (c) shows that Si/Ge atomic ratio is 25:75.
Figure 2. Experimental design for observing changes in resistivity due to tensile strain. (a) SEM image of E-PTP device. (b) Magnified image of (a) after sample preparation for the E-PTP experiment. Forming a junction between four electrodes (source high, sense high, sense low, and source low) and a NW after fixing the NW on E-PTP device, four-point probe measurement was conducted during tensile experiment. (c) TEM image of E-PTP experiment. Electrical measurements were made after inducing tensile strain in an unstrained NW. The ratio of NW diameter to length is 1/12.4 in an unstrained NW and 1/13.2 in a 3.5% strained NW. (d) Changes in resistivity of as-grown Si and SiGe NWs. The trend for the change in resistivity of NWs is contrary to each other.
roadmap, concerns continue to exist that the benefits associated with strain could be lost since, in nanoscale one-dimensional devices, the carriers are accelerated to high energies and the scattering of the carriers is more significantly affected by a
surface state with a high surface charge density and a distorted band structure.15,16 On the basis of findings reported by Cui et al., the passivation of defects at the Si surface by chemical modification can cause an approximately 20-fold enhancement 7205
DOI: 10.1021/acs.nanolett.5b01634 Nano Lett. 2015, 15, 7204−7210
Letter
Nano Letters
To investigate the electrical properties of Si and SiGe NWs as a function of tensile strain, an E-PTP device was used, as shown in Figure 2 of the SEM and HRTEM images. Si and SiGe NWs with diameters of 90−130 nm and lengths of ∼10 μm were used for the E-PTP experiments. The NWs were cleaned with a dilute HF solution before the preparation of the devices. Each end of a NW was welded onto the device by means of e-beam assisted Pt deposition. Subsequently, for a four-point probe measurement during the tensile experiment, a junction was formed between four electrodes and a NW was attached by ion-beam Pt deposition. The Pt lines between the electrodes and a NW were processed using e-beam Pt deposition, and the carbon in the Pt was then removed by vaporization through rapid thermal annealing (RTA) to conduct a Pt line (Figure S1). Ion-beam assisted Pt deposition on a Si NW resulted in the formation of Pt silicide in the Pt/Si interface, resulting in an ohmic contact (Figure S2).19 The preparations for the E-PTP experiment are outlined in Figure S3. The change in the length of the NWs could be directly measured from the series of TEM images that were obtained during the tensile experiments. Using this difference in NW length, the strain was calculated using the formula ε = (Lf − Li)/Li (Li = initial length of a NW and Lf = final length of a NW after tensile experiment).18 The resistivity (ρ) of the NWs was calculated using the results of the four-point probe measurements:
in performance in FET nanodevices in transconductance and mobility. To provide insights into changes in the performance of strained Si nanodevices, more detailed experiments using surface defect engineering in the performance of strained Si nanodevices is required. We report herein on a detailed investigation of changes in the electrical conductivity of Si and SiGe NWs as a function of tensile strain. The findings show, in the case of as-grown Si NWs, that conductivity is decreased with increasing tensile strain, whereas conductivity was increased in strained SiGe NWs. The most likely reason for this is that defects generated by an applied strain result in the degradation of electrical conduction. That is, the chemical structure at the wire surface caused by defect states induces changes in the electronic structure of the NWs. In particular, we observed that during the tensile experiments, these surface defect states induced another defect, which is closely related to O-vacancies at the interface of the oxidized Si/Si NW. By controlling surface defects through a passivation process using an Al2O3 layer, the electrical conductivity of Si NWs was proportionally increased because it permits the generated Ovacancies to be effectively suppressed. Moreover, using an electrical push-to-pull system, we were able to apply an ultra large tensile strain of over 6% to the NWs, resulting in a measurable mobility enhancement of about 2.6 times. Si and SiGe NWs were synthesized by the VLS method using an ultrahigh vacuum CVD system. A 2 nm thick Au film (as a catalyst) was deposited on a cleaned Si (111) substrate, which was then annealed at a pressure of ∼1 × 10−8 Torr for 5 min, resulting in the formation of Au−Si alloy droplets from the Au film on the Si substrate. After the formation of droplets, the NWs were synthesized by filling the main chamber with a mixture of SiH4 and GeH4 as precursors, and H2 as the carrier gas, while maintaining a fixed total pressure of 2 Torr by a feedback system using a throttle valve and a baratron gauge [the ratio of SiH4/GeH4/H2 = 20:0:200 (Si NWs), 10:10:100 (SiGe NWs)]. The process temperature was set at 400−450 °C because the phase diagrams of Au−Si and Au−Ge indicate identical eutectic temperatures and similar growth kinetics.17 High-resolution transmission electron microscopy (HRTEM) images showed that all of the NWs were grown in single crystalline forms with a growth direction of [111] with smooth sidewalls regardless of the alloy composition, as shown in Figure 1. Fast Fourier transform (FFT) images of Si and SiGe NWs show that the growth direction of the NWs was [111] and that the crystalline structure is a face-centered cubic (FCC) phase (inset, Figure 1b,c). The chemical composition of the SiGe NW had a Ge content of 0.75 and a Si content of 0.25, as evidenced by a energy dispersive X-ray (EDX) analysis (inset, Figure 1c). For reliable experimental conditions, we selected Si and SiGe NWs that had been grown vertically with similar diameters and lengths. Tensile strain experiments and fourpoint probe measurements were conducted using an electrical push-to-pull device (E-PTP, Hysitron) in HRTEM. A fourpoint probe measurement was conducted in the voltage sweep mode with a range from 0.01 to 0.03 V. Preparation of devices for use in the E-PTP experiments involved the use of a nanomanipulator (MM3A, Kleindeck) installed in a focused ion beam (FIB) system (Quanta 3D, FEI).18 The chemical structure was examined by X-ray photoelectron spectroscopy (XPS) with a monochromatic Al Kα (1486.6 eV) source. The chemical composition at the NW surface was determined by electron energy loss spectroscopy (EELS).
ρ=R
A L
where R is the result of the four-point probe measurement, A is the cross-sectional area of a NW, and L is the actual length of the NW used in the tensile experiment. In semiconductors such as Si and Ge, mechanical stress affects electronic band structure. Thus, modifying the effective electron mass results in the modulation of mobility and resistivity. Since the NWs tend to become elongated and become slightly thinner when being pulled, the change in resistance by an applied stress results from both dimensional and piezoresistivity changes. Since the observed changes in resistance are significantly larger than the dimension changes (Figure 2c), we converted the changes in resistance to a change in resistivity. Figure 2d shows the changes in the resistivity of as-grown Si NWs and as-grown SiGe NWs. As the tensile strain increased, the resistivity of strained Si NWs was increased, while that of SiGe NWs decreased linearly with increasing tensile strain. In addition, after the applied tensile was relaxed, the resistivity values of the Si NW were increased to a value near the result at the largest strain, whereas that of the SiGe NWs was reduced back to the value near the initial unstrained state. Lugstein et al. reported that the resistivity of Si(111) NW with longitudinal strain behavior along the (111) direction was decreased, largely in proportion to the applied strain. In particular, the conductivity of the 100 nm thick Si NWs could be increased by about twofold for a 3.5% elongation.20 Theoretical calculations based on the band structure of Si and SiGe NW predicted an increase in conductivity for tensile uniaxial strain resulting from an enhancement in both electron and hole mobilities.5,21 Thus, the increase in the resistivity in Si NWs indicated that other factors also affect carrier transport properties, in addition to their known piezoresistance properties. Because an increase in resistivity was only observed in the case of Si NWs, but not SiGe NWs, the factor responsible for lowering the conductivity 7206
DOI: 10.1021/acs.nanolett.5b01634 Nano Lett. 2015, 15, 7204−7210
Letter
Nano Letters
Figure 3. Changes in the mechanical and electrical properties of Si and SiGe NWs after surface passivation using a ALD-Al2O3 layer. (a) Strain− stress curves of Si and SiGe NWs before and after passivation. While the Young’s modulus of Si NWs was decreased after passivation, that of SiGe NWs was essentially unchanged. However, the fracture strain and fracture strength of both NWs were largely increased after passivation. (b) Change in resistivity of Si and SiGe NWs after passivation. The resistivity of both NWs is decreased in proportion to the increase in tensile strain. (c) Mobility enhancement of Si and SiGe NWs before and after passivation. ★ denotes the mobility enhancement in strained Si film reported in Paul’s experiments.23 It was difficult to induce a tensile strain of over 2% in experiments with planar device.
reason for the degradation in conductivity that was observed during the tensile experiment is caused by the difference in the number of defects at the Si NW surface, contrary to the case of the SiGe NW surface. Since surface defect states could be successfully controlled by the use of a passivation layer, we investigated the changes in mobility with sufficiently high strain. In the case of NWs, because the carrier concentration is low (resulting in small electric field E), the relative mobility change can be expressed using the current density equation:
appeared to be mainly related to some permanent change in the Si NWs that occurred during the tensile experiment. Mobility in FET nanodevices can be affected by surface roughness, the nature of the dopant used, and surface defects in NWs. Defect states at the NW surface have the potential to change device performance by an external field or stress during operation.14 In a previous study, we reported that defect states are formed more easily at the surface of as-grown Si NWs compared to SiGe NWs, and that they can be cured by passivation of the ALD-Al2O3 shell.18 Surface defects can be critical for carrier scattering because the charge carrier on the surface of a nano structure is the main contributor to the conduction process, resulting in a degradation in conductivity.3 In order to change the chemical structure of the surface of NWs, an Al2O3 shell was applied to the Si1−xGex NWs to prevent the formation of surface oxidation defects using an atomic layer deposition (ALD) system (Figure S4) because Al2O3 is a well-known barrier to the diffusion of externally supplied oxygen and the interfacial layer on the NW surface can be effectively removed at the initial ALD process in the Al2O3/ Si system.22 Figure 3a shows strain−stress curves for Si and SiGe NWs before and after Al2O3 passivation. The Young’s modulus (the ratio of stress to strain) of the Si NWs was largely decreased after passivation (before passivation, 182 ± 23 GPa; after passivation, 167 ± 28 GPa), whereas that of the SiGe NWs after passivation was negligible within the error range (before passivation, 112 ± 28 GPa; after passivation, 115 ± 18 GPa). These results indicate that defects can be significantly reduced through the use of a passivation process in the case of Si NWs, compared to the SiGe NWs. This difference between Si NW and SiGe NW can be attributed to the fact that the number of surface defect states generated in Si NWs is larger than that of the SiGe NWs under strain. After the deposition of a passivated Al2O3 layer, the resistivity of both Si NWs and SiGe NWs was decreased in proportion to the increase in tensile strain, as shown in Figure 3b. Changes in the resistivity of strained Si NWs after passivation indicate that one significant
Jp =
ξεμp V 2 L3
= σE =
1 E ρ
where the geometric prefactor ξ depends on the ratio R/L, ε is the dielectric constant of the semiconductor, and μp is the hole mobility.20 This relation shows that a significant change in current due to strain can only be attributed to a significant mobility change. Figure 3c provides information on the mobility enhancement of the NWs before and after passivation. The results for as-grown Si NWs show the opposite tendency against the calculated and experimental reports, whereas the trend for the others is consistent with previous reports. In particular, the value of fracture strain in the NWs was largely increased due to the presence of a passivated Al2O3 shell. Mobility enhancement becomes significant when the tensile strain exceeds 6%. Such a mobility enhancement due to ultralarge strain has not been reported yet in other experiments dealing with strained channel properties. We confirmed that the surface of the Si NWs and the SiGe NWs is covered by a thin oxide layer of SiO2−x containing Si surface defects or oxide defects.18 To investigate the nature of the defect states at a NW surface during tensile strain experiments, we carried out EELS in the scanning transmission electron microscope (STEM) mode. More than 10 spectra were taken near the surface of a NW across the diameter, and we then selected 3−5 of them to demonstrate the behavior of the surface states during the tensile experiment. To check the 7207
DOI: 10.1021/acs.nanolett.5b01634 Nano Lett. 2015, 15, 7204−7210
Letter
Nano Letters
Since significant degradation in mobility can be caused by defects at the NW surface, it is extremely important to recognize the behavior of the defects during the tensile experiments. To investigate changes in the defect states, we obtained EELS spectra of the Si NW after completion of the tensile experiment. Similar to the results for the unstrained NW, EELS spectra in the interfacial region between the NW and SiO2−x layer (regions indicated by a red cross and a green cross in Figure 5a) show that the Si surface defects (indicated by blue arrows in Figure 5b,c) and SiO2−x states (pointed with an orange arrow in Figure 5b) can be clearly observed. In particular, in comparing the O K-edge spectra of the Si NW after the tensile test with that before the tensile test, we were able to observe new states that were largely generated in the lower energy region compared to the states at conduction band edge (indicated by a yellow−green arrow, Figure 5c,d). This indicates that defect states located at the NW surface can cause the generation of additional defect states. The states observed near 523 eV are related to O atom vacancies at the Si surface.29 This result cannot be acceptable in viewing the defect generation process. Although a change in mobility can be affected by surface carrier scattering with defects at interfacial regions such as an oxide defect of SiO2−x and a surface defect of the NW, only Si defects generated at the NW surface critically affect the fracture process of a NW under conditions of tensile stress. To identify the origin of the defect generation due to tensile strain, it is necessary to have detailed information regarding the surface chemical structure of the NWs. XPS Si 2p spectra and O 1s spectra show that there are SiO2−x states as well as Si−OH states on the surface of the as-grown Si NWs, which was also confirmed by FT-IR results (Figure 6). The formation of Si−OH states at the Si surface can induce defect states related to the unstable Si−OH and dangling bonds by the following equation:
chemical states at the NW surface, we compared Si L-edge EELS spectra of an as-grown Si NW and a SiO2/Si sample, respectively (Figure S5). The EELS spectra in the core region and surface region of a Si NW show a characteristic shape for silicon and silicon oxide, respectively.24−26 However, in the case of spectra obtained for the interfacial region between the surface native oxide and the Si NW, some states located in the lower energy region (pointed with red arrow, Figure S5b) were observed, compared to the Si characteristic peak at near 101 eV, which is related to defect states generated at the Si NW surface. These states observed at the conduction band edge were caused by unstable states that are formed at the Si surface, which induced gap states within the Si band gap.25 Moreover, broadened states corresponding to SiO2−x were observed near 104 eV (indicated by a purple arrow, Figure S5b).24,26 Surface defects and SiO2−x states observed at the interfacial region were not observed in Si L-edge spectra of a chemically stable interface in the SiO2/Si sample (Figure S5c). Figure 4 shows
Si + H 2O → : Si + H−OH → ·Si−SiOH + ·Si−SiH
where : is an Si dangling bond.30 These states, particularly dangling bonds, result in a reduction in carrier mobility and an increase in carrier lifetime, which is similar to the effects from the interfacial SiO2−x states.31,32 Moreover, the application of tensile strain induces a charge accumulation in dangling bonds, which could be the origin of the generated defect.33 In density functional theory (DFT) calculations, the distortion of the Si NW structure was clearly observed when a tensile strain was applied to a Si surface containing vacancies, which induces the generation of charged vacancy states (inset of Figure 5d). The DFT results show that the defective NW surface with Si−OH states including Si dangling bonds or O-vacancies can produce broken bonds caused by vacancy formation near the defects much easier than the defect-free surface as the tensile strain is increased (Figure S6). The energy of formation of these vacancy-related broken bonds was gradually decreased with increasing tensile strain because of the increase in coulomb interactions between electron−hole pairs in broken sites of dangling bonds; that is, the vacancy-related defects increase with increasing tensile strain. Surface passivation using an Al2O3 layer can result in the removal of surface broken bonds such as dangling bonds and vacancies. This explains why defect generation was prevented after the tensile strain experiment, as shown in Figures 5d and S7. These findings, therefore, suggest that the defects located at the NW surface induce bond braking in neighboring defect sites, thus increasing tensile
Figure 4. EELS spectra of Si and SiGe NWs were observed near the surface to identify the surface chemical structure of the NWs. (a) Si Ledge and (b) O K-edge EELS spectra of the NWs before and after passivation. While changes in the surface states were clearly observed in the Si NWs, those were scarcely observed in the SiGe NWs. Red arrows indicate the surface states shown in the as-grown NWs.
EELS spectra for the surface region of the unstrained Si and SiGe NW before and after Al2O3 passivation. In Figure 4a, the defect states in the Si L-edge spectra of the as-grown Si NW (arrow region) had obviously decreased as the result of the passivation, whereas the change in the spectra of the SiGe NW before and after passivation was small. This indicates that the effect of passivation is not significant in SiGe NW because only a small number of defect states are formed on the surface of the as-grown SiGe NW, compared to a Si NW. A change in defect states can be also observed in O K-edge spectra. In Figure 4b, O K-edge spectra of an as-grown Si NW show that the states of the conduction band edge related to surface defects (indicated by an arrow) were obviously formed but were significantly decreased after passivation.27−29 However, in the as-grown SiGe NW, the defect states at the NW surface were decreased substantially, compared to that of Si NW. Moreover, the defect states were also decreased slightly after passivation. 7208
DOI: 10.1021/acs.nanolett.5b01634 Nano Lett. 2015, 15, 7204−7210
Letter
Nano Letters
Figure 5. To investigate the effect of tensile strain on the surface states of the NWs, EELS spectra for the NWs were obtained after the tensile experiments (3% strain). (a) STEM image of the as-grown Si NW. Each × denotes the spot where an EELS spectrum was collected. (b) Si L-edge and (c) O K-edge EELS spectra of the Si NW after tensile experiment. The color of the EELS spectrum is the same as that of each observed spot in (a). (d) Change in EELS spectra of the Si NW before and after the tensile experiment. Newly generated states were only observed in the as-grown Si NWs after the tensile experiment. The inset of (d) shows the structural changes at the surface including Si−OH bonds of Si NW system observed from DFT calculation. Structural deformation due to tensile strain was observed in the system with vacancy formation, but not in the system without vacancies.
Figure 6. (a) XPS Si 2p spectra and (b) XPS O 1s spectra of Si NWs before and after Al2O3 passivation. Si 2p and O 1s spectra were deconvoluted using the peak positions of Si0 at 99.3 eV, Si1+ + Si−OH at 100.0 eV, Si2+ at 101.2 eV, Si3+ at 102.1 eV, Si4+ at 103.1 eV, Al−O at 531.5 eV, Si−O at 531.7 eV, Si−O−Si at 532.5 eV, and Si−OH at 533.4 eV, respectively. After passivation, the surface defect states related to SiO2−x and SiOH were significantly decreased. (c) FT-IR spectra of the NWs before and after passivation. Si−OH peak in the Si NWs was changed to Al−O peak after the passivation process. In the SiGe NWs, the appearance of Si−OH and Ge−OH vibration modes was essentially negligible.
properties. In particular, after passivation, a large mobility enhancement of about 2.6 times was observed in the Si NWs with a tensile strain of 5.7%. Moreover, although the mobility enhancement in Si NWs due to strain was very high compared to that in SiGe NWs (Figure 3c), the degree of enhancement was 1.5 times larger in the SiGe NWs than in the Si NWs (Figure 3b). Because the mechanical property of the surface passivated NWs would approach ideal values of a perfect semiconductor crystal,18 this suggests that curing the defect
strain, which causes mobility to be degraded in proportion to increasing strain. Information regarding changes in electrical properties due to the surface states of strained NWs has been missing for many years. For ultrahigh mobility devices utilizing strained SiGe NW FET, positive modulation of the surface chemical structure is needed to decrease the defect states at the NW surface. We applied a passivation process to decrease the number of surface defects that cause the degradation in carrier transport 7209
DOI: 10.1021/acs.nanolett.5b01634 Nano Lett. 2015, 15, 7204−7210
Letter
Nano Letters
(20) Lugstein, A.; Steinmair, M.; Steiger, A.; Kosina, H.; Bertagnolli, E. Nano Lett. 2010, 10, 3204. (21) Buca, D.; Hollander, B.; Feste, S.; Lenk, St.; Trinkaus, H.; Mantl, S.; Loo, R.; Caymax, M. Appl. Phys. Lett. 2007, 90, 032108. (22) Ma, J. W.; Lee, W. J.; Cho, M.-H.; Chung, K. B.; An, C.-H.; Kim, H.; Cho, Y. J.; Moon, D. W.; Cho, H. J. J. Electrochem. Soc. 2011, 158, G79. (23) Paul, D. J. Semicond. Sci. Technol. 2004, 19, R75. (24) Batson, P. E. J. Electron Microsc. 1996, 45, 51. (25) Muller, D. A.; Sorsch, T.; Moccio, S.; Baumann, F.; EvansLutterodt, H. K.; Timp, G. Nature 1999, 399, 758. (26) Batson, P. E. J. Electron Microsc. 2000, 49, 267. (27) O’Reilly, E. P.; Robertson, J. Phys. Rev. B: Condens. Matter Mater. Phys. 1983, 27, 3780. (28) Lucovsky, G.; Miotti, L.; Bastos, K. P. J. Vac. Sci. Technol. B 2011, 29, 01AA01. (29) Lucovsky, G.; Zeller, D.; Wu, K.; Whitten, J. L. Microelectron. Eng. 2011, 88, 1537. (30) Gallet, J.-J.; Bournel, F.; Rochet, F.; Kohler, U.; Kubsky, S.; Silly, M. G.; Sirotti, F.; Pierucci, D. J. Phys. Chem. C 2011, 115, 7686. (31) Winer, K.; Hirabayashi, I.; Ley, L. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 38, 7680. (32) Stegemann, B.; Sixtensson, D.; Luβky, T.; Schoepke, A.; Didschuns, I.; Rech, B.; Schmidt, M. Nanotechnology 2008, 19, 424020. (33) Fernández-Serra, M.-V.; Adessi, Ch.; Blase, X. Nano Lett. 2006, 6, 2674.
states located at the NW surface is critical for improving performance and securing reliable strained SixGe1−x NW FET. Furthermore, if surface defects could be controlled, Si and SiGe nanodevices could be the basic component of new nanodevices for use in applications such as high efficiency nanoelectronics, nanosensors, and flexible devices.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.5b01634. Interfacial defect analysis using EELS and DFT calculation (PDF) Tensile experiment process of SiGe NW (AVI)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the Korea Research Institute of Standards and Science under the Metrology Research Center project and Yonsei-Samsung Semiconductor Research Center (YSSRC) program
■
REFERENCES
(1) Cui, Y.; Zhong, Z.; Wang, D.; Wang, W. U.; Lieber, C. M. Nano Lett. 2003, 3, 149. (2) Li, Y.; Qian, F.; Xiang, J.; Lieber, C. M. Mater. Today 2006, 9, 18. (3) Wu, X.; Kulkarni, J. S.; Collins, G.; Petkov, N.; Almécija, D.; Boland, J. J.; Erts, D.; Holmes, J. D. Chem. Mater. 2008, 20, 5954. (4) Wong, H.-S. P. Solid-State Electron. 2005, 49, 755. (5) Buin, A. K.; Verma, A.; Svizhenko, A.; Anantram, M. P. Nano Lett. 2008, 8, 760. (6) Whang, S. J.; Lee, S. J.; Yang, W. F.; Cho, B. J. Appl. Phys. Lett. 2007, 91, 072105. (7) Chaudhry, A.; Ramamurthi, V.; Fong, E.; Islam, M. S. Nano Lett. 2007, 7, 1536. (8) Ma, D. D. D.; Lee, C. S.; Lee, S. T. Appl. Phys. Lett. 2001, 79, 2468. (9) Jin, S.; Whang, D. M.; McAlpine, M. C.; Friedman, R. S.; Wu, Y.; Lieber, C. M. Nano Lett. 2004, 4, 915. (10) Zheng, G.; Lu, W.; Jin, S.; Lieber, C. M. Adv. Mater. 2004, 16, 1890. (11) Leu, P. W.; Svizhenko, A.; Cho, K. Phys. Rev. B: Condens. Matter Mater. Phys. 2008, 77, 235305. (12) Svizhenko, A.; Leu, P. W.; Cho, K. Phys. Rev. B: Condens. Matter Mater. Phys. 2007, 75, 125417. (13) Haugerud, B. M.; Bosworth, L. A.; Belford, R. E. J. Appl. Phys. 2003, 94, 4102. (14) He, R.; Yang, P. Nat. Nanotechnol. 2006, 1, 42. (15) Hoyt, J. L.; Nayfeh, H. M.; Eguchi, S.; Aberg, I.; Xia, G.; Drake, T.; Fitzgerald, E. A.; Antoniadis, D. A. IEDM Technol. Digest 2002, 23. (16) Thompson, S. E.; Parthasarathy, S. Mater. Today 2006, 9, 20. (17) Lew, K. K.; Pan, L.; Dickey, E. C.; Redwing, J. M. Adv. Mater. 2003, 15, 2073. (18) Ma, J. W.; Lee, W. J.; Bae, J. M.; Jeong, K. S.; Kang, Y. S.; Cho, M.-H.; Seo, J. H.; Ahn, J. P.; Chung, K. B.; Song, J. Y. Nano Lett. 2013, 13, 1118. (19) Larrieu, G.; Dubois, E.; Wallart, X.; Baie, X.; Katcki, J. J. Appl. Phys. 2003, 94, 7801. 7210
DOI: 10.1021/acs.nanolett.5b01634 Nano Lett. 2015, 15, 7204−7210
