- Email: [email protected]
Honeycomb silicon: a review of silicene
Sci. Bull. DOI 10.1007/s11434-015-0880-2
www.scibull.com www.springer.com/scp
Review
Materials Science
Honeycomb silicon: a review of silicene Jincheng Zhuang • Xun Xu • Haifeng Feng Zhi Li • Xiaolin Wang • Yi Du
•
Received: 13 July 2015 / Accepted: 14 August 2015 Ó Science China Press and Springer-Verlag Berlin Heidelberg 2015
Abstract Silicene, a new allotrope of silicon in a twodimensional honeycomb structure, has attracted intensive research interest due to its novel physical and chemical properties. Unlike carbon atoms in graphene, silicon atoms prefer to adopt sp2/sp3-hybridized state in silicene, enhancing chemical activity on the surface and allowing tunable electronic states by chemical functionalization. The silicene monolayers epitaxially grown on Ag(111) surfaces demonstrate various reconstructions with different electronic structures. In this article, the structure, phonon modes, electronic properties, and chemical properties of silicene are reviewed based on theoretical and experimental works in recent years. Keywords Silicene Electronic states Chemical functionalization
1 Introduction The discovery of graphene in 2004 by Novoselov et al. [1] demonstrated that, for the first time, stable and singleatom-thick two-dimensional (2D) materials could be exfoliated from van der Waals solids, and these materials possess unique and fascinating properties due to their exotic electronic structures [1–6]. Since then, graphene has been one of the most extensively studied materials due to the wealth of anomalous physical and chemical phenomena that occur when charge and heat transport is confined to a plane. The rapid and recent advances in this carbon-based
J. Zhuang X. Xu H. Feng Z. Li X. Wang Y. Du (&) Institute for Superconducting and Electronic Materials (ISEM), University of Wollongong, Wollongong, NSW 2525, Australia e-mail: [email protected]
2D material have raised teasing questions on exploring new 2D materials with distinct and exotic properties. Such possibilities have promoted a completely new research field of 2D crystals. There are many 2D materials that have been explored, and they exhibit rich spectrum of properties. For example, hexagonal boron nitride (h-BN) with the closest structure to graphene is an insulator [7], while monolayers MoS2 and WS2 are semiconductors with direct band gap [8, 9]. The abundant variety of properties in 2D materials suggests the candidate potential for the device engineering and application in sensing, photonics and energy storing, etc. Great efforts have been made to incorporate 2D atomic layers into devices to exhibit exceptional performance. It has been extremely difficult so far, however, to develop reliable and durable applications, especially in electronics, based on 2D materials due to their incompatibility with current semiconductor-based electronic techniques. Moreover, it is extremely hard to obtain single crystal with large size for characterization and device fabrication. Searching for the novel 2D materials is therefore highly desirable, especially for next-generation low-cost super-performance electronic devices. Silicene is similar to graphene in that it is single atom thick, and it has the same characteristic honeycomb structure. It is a big challenge, however, to synthesize silicene, despite the similarity of electronic configurations between carbon atoms and silicon atoms. Nevertheless, the sp2 hybridization is more stable than sp3 hybridization for carbon, but situation is reversed for silicon. Thus, silicon atoms are energetically not favorable to spontaneously form silicene, and this limits conventional chemical or physical methods to prepare silicene. Despite the difficulty in fabrication, many theoretical calculations have predicted an exciting and rich physics in silicene [10–12]. For instance, the Dirac fermion state was demonstrated in a
123
Sci. Bull.
linear dispersion band structure close to Fermi level by density functional theory (DFT) calculations [11]. The quantum spin Hall effect (QSHE) and quantum anomalous Hall effect (QAHE) were also predicted in silicene [13– 15]. The electronic properties of the silicene could be modulated by effect of substrate [16], defect [17, 18], interlayer coupling [19], and metal interaction [20]. These theoretical works have paved the way of exploiting the anisotropic transport behavior and experimentally investigating exotic properties of silicene [21–31]. Moreover, silicene is derived from silicon—the precise substance that silicon chips are made of. As such, it is inherently compatible with current chip-production processes, which is advantageous in contrast to graphene. There is no doubt that silicene could be a material of the future. On the energy front, silicene is expected to make solar power more practical and efficient by improving the storage capacity and output speed of solar cells [6]. Silicene may also yield advances in medical sensors (which monitor heart rates and blood flow), sharpen X-ray imaging, and improve DNA sequencing [3–6]. In addition, it will make health-monitoring devices faster, cheaper, and even disposable. It can create more efficient batteries, paper-thin cell phones, and flexible screens.
2 2D structure of silicene Silicene, silicon-based graphene-like sheets, has been proposed recently by first-principle total-energy calculations [32]. However, different from the planar structure of graphene with sp2 hybridization network, silicene is a lowbuckled structure due to mix of sp2 and sp3 hybridization. Due to the competition between sp2 hybridization and sp3 hybridization, rich phases of silicene have been reported grown on varies substrates, including Ag(111) [2, 10, 13, 28, 33, 34], Ag(110) [35], Au(110) [36], Ir(111) [3], and ZrB2-covered Si(111) [4]. Here, we focus on the structure of silicene phases grown on Ag(111). 2.1 3 9 3 silicene 3 9 3 silicene is 4 9 4 reconstruction with respect to the Ag(111) substrate or 3 9 3 reconstruction with respect to 1 9 1 silicene [10]. Large-scale 3 9 3 silicene films can be easily prepared by keeping substrate temperature around 420 K during deposition. In 2012, three groups reported detailed structure models of 3 9 3 silicene by comparing scanning tunnel microscope (STM) results with theoretical calculation [2, 13, 20]. Figure 1a shows the phase exhibiting honeycomb structures (labeled H). The other phase consisting of close-packed protrusions (labeled T) is shown in Fig. 1b. These two phases coexist with each
123
Fig. 1 (Color online) STM images and structure models of H phase silicene (a) and T phase silicene (b). Reprinted with permission from Ref. [2]. Copyright (2012) American Chemical Society
other, indicating similar formation energy and stability. However, phase T tends to form at a lower Si coverage and lower temperature compared to phase H. At higher Si coverage, phase T will disappear and be replaced by phase H. Thus, phase H is regarded as the stable phase of 3 9 3 silicene, and phase T is a precursor of phase H. Based on the structure in STM observation and first-principle calculation, a structure model of a low-buckled honeycomb structure with missing of a hexagonal silicon rings at the corner in each 3 9 3 cell is proposed, as shown in structure model in Fig. 1b. Although similar STM results have been represented by Liu et al. [20] and Vogt et al. [13], different structure models have been proposed, where the hole in STM image is attributed to six Si atoms buckled upward. The recent work of hydrogen absorption on 3 9 3 silicene reported by Qiu et al. [37] is helpful to clarify this controversy. It is found that absorption of hydrogen changes buckling structure of 3 9 3 silicene by detailed discussion on STM results and theoretical calculation; they proposed a structure model with intact honeycomb structure. 2.2
pffiffiffi pffiffiffi 7 7 silicene
With increasing substrate p temperature ffiffiffi pffiffiffi to 480 K, a new phase of silicene named as 7 7 silicene emerges [2]. This phase usually manifests itself as a defective moire´ pattern with a period about 3.8 nm, as shown in Fig. 2. Figure 2b shows that this moire´ pattern consists of hexagonal rings at bright part and defective part between them. By first-principle they construct a pffiffiffi calculation, pffiffiffi structure model for 7 7 silicene (Fig. 2d, f). They attribute the bright parts of moire´ pattern to the position deviation of silicon atoms from that of Ag(111). Thus, the strong interaction between silicene and Ag(111) substrate will make hexagonal rings stable and complete. In other parts of moire´ pattern, because of larger deviation of position between silicon atoms and that of substrate, the interaction breaks these hexagonal rings.
Sci. Bull.
and sixfold symmetry of the surface structure. Recently, Shirai et al. [39] argued that substrate pAg ffiffiffi atoms pffiffiffi may segregate on silicene surface forming a 3 3 superpffiffiffi pffiffiffi structure which is similar with 3 3 silicene from tensor low-energy electron-diffraction (LEED) pffiffiffi panalysis. ffiffiffi The similarity p between the proposed Ag 3 3 superffiffiffi pffiffiffi structure and 3 3 silicene is also reported by very recent STM worksp[40, ffiffiffi p41]. ffiffiffi As a result, these works question whether 3 3 silicene is a true layered structure. The detailed is p performed pffiffiffi discussion pffiffiffi ffiffiffi pffiffiffi in Sect. 6. Excepting for 3 3 silicene, 7 7 silicene and pffiffiffi pffiffiffi 3 3 silicene, various superstructures with different buckling patterns and periodicity are also reported [42]. But systematic experimental and theoretical investigations are lacking for these phases. Even for the three phases discussed above, there are more controversies than common views. Detailed studies on structure of silicene phases are needed.
3 Phonon modes in silicene
Fig. 2 (Color online) a A derivative STM image (200 nm 9 200 nm, Vtip = 1.43 V) of 0.9 ML (ML, monolayer) silicon atoms deposited on Ag(111) surface at substrate temperature of 480 K; b high-resolution STM image (15 nm 9 15 nm, Vtip = -1.0 V) showing the atomic structure of moire´ patterns. The bright areas exhibit complete honeycomb rings with a period of 1.0 nm, while other areas are defective and disordered. The angle between the orientation of the hexagonal rings and the direction of moire´ patterns is 30°; c dI/dV spectra taken at the moire´ pattern phase, in which a peak at 0.3 and pffiffiffi V p ffiffiffi a shoulder at 0.9 V are observed; d calculated model of 7 7 superstructure of silicene; e, f experimental and simulated STM images (1.0 eV above Fermi energy) showing the similar structure features and unit cell of lattice. Reprinted with permission from Ref. [2]. Copyright (2012) American Chemical Society
2.3
pffiffiffi pffiffiffi 3 3 silicene
Further increasing substrate temperatureptoffiffiffi 500pffiffiKffi leads to a most stable silicene phase named as 3 3 silicene. This phase is the most extensively investigated phase of silicene. Meanwhile, it may be the most controversy phase. Wu and co-workers [2, 21] proposed a structural model based on a honeycomb lattice with a unit cell (UC) that consists of three layers of silicon atoms as shown in Fig. 3. This model is suspected by Arafune et al. [38], who proposed a bilayer structure judging from the layer thickness
Raman spectroscopy is widely used in structural characterization, electron–phonon coupling (EPC) probe, and the phonon dynamics investigation in 2D Dirac fermions system [43]. In silicene, the long-wavelength optical E2g phonon mode at the C point of the Brillouin zone (BZ), which corresponds to the relative displacement of nonequivalent neighbor silicon atoms [14], is of particular interest. Any perturbations due to this buckled structure will effectively induce the direct electronic transitions across the Dirac point, that is, E2g phonons couple to lowenergy excitations. The vibrational properties of silicene are firstly investigated by means of DFT calculations [44]. By considering the small buckling of silicene and favorable energy, the non-resonance Raman spectra of free-standing silicene are presented by a prime peak (E2g) located at around 570 cm-1. It is also predicted that the Raman peaks attributed to the defects at the edges distribute at low frequencies than E2g peak. However, it has been an experimental challenge to verify vibrational properties in this silicon-based 2D Dirac fermion system because the monolayer silicene is unstable under ambient conditions [45]. To prevent the oxidation of silicene, co-deposition of the Al and O2 to form Al2O3 capping layer in ultra-high vacuum (UHV) ambient is intended before the ex situ Raman measurement, then resulting in an intact Al2O3/silicene/Ag heterostructure [46, 47]. Combined with DFT calculation, the Raman spectra of different silicene superstructures display that E2g peak locates at around 520 cm-1, much smaller than the value of simulated free-standing silicene [44]. Furthermore, Raman spectrum is also investigated depending on excitation
123
Sci. Bull.
pffiffiffi pffiffiffi Fig. 3 (Color online) a Top view and b side view of the lattice geometry of the low-buckled silicene structure (AB) and 3 3 superstructure (AB-A), respectively. Note that in the superstructure, a Si atom with p planar (2/3, 2/3) is pulled downward; c the structural phase ffiffiffi pcoordinate ffiffiffi transition diagram of silicene depending on the lattice constants of 3 3 superstructure; d a larger schematic model illuminating the pffiffiffi pffiffiffi honeycomb structure of 3 3 reconstructed silicene. Reprinted with permission form Ref. [21]. Copyright (2014) by the American Physical Society
energy to probe the resonant behavior [47]. The results indicate that the monolayer silicene is multi-hybridized by sp2 structure and sp3 structure. Nevertheless, the oxidation cannot be excluded in this ex situ measurement. The broad shoulder at lower wavenumber (450–510 cm-1) next to the silicene signature E2g peak, which is ascribed to bucklinginduced vibrational modes in [47], is actually associated with the Si–O bonds due to Si sp3 hybridization [34]. Due to the low oxygen adsorption energy on the Si surface with dangling bonds, the monolayer silicene is extremely sensitive to the oxygen. The structure of monolayer silicene would be completely destroyed in the condition of exposure to 600 Langmuir (L) oxygen. Therefore, in situ Raman measurement under UHV is expected to be performed to identify silicene in different phases and reveal the details of the phonon modes as well as their relationship to the silicene electronic properties. The in situ Raman scattering studies of phonon modes in epitaxial silicene with different reconstructions on a Ag(111) surface are performed to eliminate the effect of oxidization [48]. Combined with STM, it is implied that the characteristic Raman peaks of silicene are sensitive to the electronic structures, making it possible to unambiguously distinguish between silicene phases in a fast and non-
123
destructive way. Compared to previous research,pZhuang ffiffiffi pffiffiffi et al. [48] firstly reports the Raman spectrum of 3 3 phase as a function of coverage. The first-order asymmetric peak, interpreted as the zone-center E2g vibrational mode, of all silicene phases locates at around 530 cm-1, as shown in Fig. 4. There are two major differences between the Raman results of different layers. Firstly, no pffiffiffidefect pffiffiffi peak could be observed in the Raman results of 3 3 film, which is consistent with the STM pffiffiffi results pffiffiffi that scarce boundaries could be found in the 3 3 silicenepfilm. ffiffiffi Secondly, the much stronger E2g mode is observed in 3 p ffiffiffi 3 silicene than that of monolayer silicene. The E2g vibrational mode is generated by the bond stretching of all sp2 silicon atoms, and is the fingerprint of honeycomb pffiffiffi pffiffiffi lattice [43, 48]. Therefore, the sp2 strength pffiffiffi in pffiffi3ffi 3 silicene film is much higher than that in 7 7/3 9 3 silicene film. It should be noted that experimental frequency of E2g mode is smaller than that of free-standing silicene [44]. There are two factors which could modulate the E2g mode in 2D materials. One is the strain effect, reflected by the ˚ ) compared to freestretched Si–Si bond length (*2.35 A ˚ ) [44, 47, 48]. The difference in standing silicene (2.28 A bond length could induce 5 % in-plane tension, softening
Sci. Bull.
pffiffiffiffiffi pffiffiffiffiffi pffiffiffi pffiffiffi Fig. 4 (Color layer; bffiffiRaman spectra of 3 3 silicene pffiffiffiffiffionline) pffiffiffiffiffia Raman spectra of Ag(111) substrate, Si(111), and 13 13/4 9 4 bufferp ffiffiffi p ffi grown onffiffiffi 13 13/4 9 4 buffer layer with different coverage, SL denotes thepcoverage p ffiffiffi p ffiffiffiffiffi pffiffiffiffiffiof the 3 3 silicene layer; c Raman spectra of 3 3 silicene layer (coverage of 0.3 ML) grown on Ag(111) surface and 13 13/4 9 4 buffer layer, respectively. Reprinted with permission from Ref. [48]. Copyright (2015) by the American Physical Society
the E2g phonon frequency with a value of 50 cm-1. The other factor is the carrier doping. In fact, electron or hole doping in silicene can result in a hardening of the mode frequency in silicene [14], e.g., E2g mode will shift to higher frequency due to charge doping, as shown in Fig. 5. Angle-resolved photoemissionpffiffiffi spectroscopy (ARPES) pffiffiffi verifies that the Dirac point of 3 3 silicene grown on silver substrate is located at *0.33 eV below the Fermi surface due to electron doping from the Ag(111) substrate. E2g vibration at the zone center couples with Dirac fermions at the zone boundary, which is allowed by silicene lattice symmetry. Since the Fermi energy (EF) of silicene is lifted by electron doping, E2g mode in silicene is hardened. The E2g frequency upshifting by about 10 cm-1 agrees well with previous simulation results [14]. Since EPC strength depends primary on the phonon frequencies in 2D materials [50, 51], the shift of E2g peak in silicene induced by both strain effect and carrier doping effect could modulate the EPC strength. In general, EPC can be characterized by a dimensionless parameter c proportional to x-2 [49–51]. This relation originates from the zero-point oscillation amplitude induced by large deformation and the energy reconstruction in the perturbation theory. Both factors correspond exactly to the strain effect presenting in 2D materials. Therefore, the value of x reflects the EPC strength. According to E2g peak shift from 570 cm-1 (FS) to 520 cm-1 (strained silicene), the EPC in silicene can be enhanced up to 20 %. The enhancement of EPC is of particular interest because it gives rise to superconductivity in Bardeen–Cooper–Schrieffer (BCS) superconductors. The Raman results consist with recent observations on the existence of a superconducting gap in silicene layers [52]. Detecting the relationship between EPC and Dirac fermions through Raman scattering will
provide a direct route to investigate the exotic property in buckled two-dimensional honeycomb materials. Five distinct Raman peaks at low range pffiffiffi wave-number pffiffiffi (220–470 cm-1) are observed in 3 3 silicene [48]. The intensities of these peaks can be scaled well with each other and show a strong dependence on the coverage. The special phonon modes correspond to 2D electron gas scattering at the edge sites in the silicene [48]. The edgeinduced Raman peaks reflect the unique buckled characteristic in silicene. However, the detailed relationship of each peak with different edges, e.g., armchair edge and zigzag edge, is still unknown; tip-enhanced Raman spectroscopy is expected to advance insights into the edge effects on phonon modes in this low-buckled 2D material. In fact, Raman spectroscopy has been adopted to monitor the air ability of encapsulated epitaxial silicene [53, 54], and is crucial for the confirmation of the integrity of silicene in the field-effect transistor. Seven days after transfer, the Raman spectrum remains the same as that of freshly growth silicene, implying that the pristine silicene is preserved [53]. The Raman spectroscopy allows unambiguous, high-throughput, nondestructive identification of epitaxial silicene.
4 Electronic structures in silicene Silicene is predicted to share the same novel electronic properties as graphene [32, 55–57]. The electronic p- and p*-bands derived from the Si 3pz orbital disperse linearly to cross around the Dirac point, and charge carriers behave like massless Dirac fermions [20]. However, due to sp2– sp3-hybridized atomic arrangement [13], monolayer silicene shows a strong interfacial coupling with substrate. In
123
Sci. Bull.
Fig. 5 (Color online) a Raman spectrum of E2g peak for different coverage samples. The E2g mode frequency of FS silicene is plotted as Ref. for 0.3 SL sample and 0.8 SL sample. The dashed lines are used to mark the position of E2g mode; c ARPES [44]; b fitted resultspof ffiffiffi E2g pffiffipeak ffi results of epitaxial 3 3 silicene layer. The left part shows that two faint linear dispersed bands crossed at BZ center C point. The right part displaying constant-energy cuts of the spectral function at different binding energies verifies that both bands origin from a Dirac cone structure; d E2g mode frequency as a function pffiffiffi pffiffiof ffi coverage. Both (d) and the inset of (d) are sketched to illustrate the strain and doping effects on Raman peak position of E2g mode in 3 3 silicene, in which the strain effect softens the E2g mode, while charge doping will upshift the E2g mode. Reprinted with permission from Ref. [48]. Copyright (2015) by the American Physical Society
this case, the electrons are delocalized into the substrate, which could induce the absence of Landau levels and lose its Dirac fermion characteristics [58]. The ARPES results display that no Dirac cones could be observed in the K points of monolayer silicene BZ, instead, a dominant hybrid surface metallic band is generated by combined effect of the monolayer silicene and Ag(111) sp-band [19, 59–61]. Therefore, the electronic structure of monolayer silicene has been intensively modulated by the substrate, resultingpfrom ffiffiffi pthe ffiffiffi buckling-induced distortion. The 3 3 silicene is in indirect contact with the silver substrate, which could induce the low-buckled pffiffiffi structure and a weak interfacial coupling between 3 p ffiffiffi 3p silicene ffiffiffi pffiffiand ffi the substrate. Thus, the electronic structure of 3 3 silicene is expected to diverse from that of monolayer silicene, and to present the Dirac fermion characteristics. The pronounced quasiparticle interferences (QPI) patterns were detected in the scanning tunneling
123
spectroscopy (STS) measurement [19]. The linear band dispersion is deduced indicating the existence of pffiffiffi from pffiffiQPI, ffi Dirac fermion in 3 3 silicene [19]. The direct signature of Dirac fermion in this phase has been observed, in 0 which the Dirac cones appear on the K pand of ffiffiffi Kpffiffipoints ffi 1 9 1 silicene BZ as well as C points of 3 3 silicene BZ, as shown in Fig. 6 [63, 64]. The Dirac point is below the Fermi surface due to the electron doping from Ag(111) substrate [48]. The calculated Fermi velocity from ARPES results at room temperature is as high as 0.3 9 106 m s-1, still a bit smaller than that of graphene [63, 65], but is high enough for potential silicene-based application. The electronic structure, investigated by both STS and ARPES measurements, implies two critical discoveries in silicene system. Firstly, the electronic band of single layer silicene is drastically modulated by Si–Ag coupling. Secondly, the pDirac ffiffiffi pffiffiffifermion characteristics have been observed in 3 3 silicene, implying the weak coupling
Sci. Bull.
pffiffiffi pffiffiffi Fig. 6 (Color online) Dirac fermionics of multilayer ( 3 3) R30° silicene islands. a Scheme of Brillouin zones; b K- and V-shape linear p and p* silicene bands at C0; c waterfall line profiles of the dispersion displayed in b as a guide for the eyes. Reprinted with permission for Ref. [62]. Copyright (2013), AIP Publishing LLC
of Si atoms in different layers. Since core for the application of 2D materials is the presence of Dirac fermion properties [66], the electronic band investigations indicate that silicene is highly promising for the fabrication of innovative device with direct compatibility with silicon micro- and nano-technologies.
5 Chemical properties of silicene 5.1 Hydrogenation Hydrogenation was found as a useful chemical method to modulate the electronic property of graphene, leading to a band-gap opening [67–69]. In contrast to the graphene case, where hydrogen tends to form clusters, hydrogenated silicene exhibits a perfectly long-range ordered structure. Combining with first-principle calculations, it has been determined that there are seven hydrogen atoms in one (3 9 3) UC and that the buckling configuration of Si atoms in silicene spontaneously rearranges upon hydrogenation [37]. Figure 7 shows the typical changes induced by the hydrogenation of silicene-(3 9 3). Upon exposure of 900 L hydrogen at room temperature, a perfectly ordered
Fig. 7 (Color online) a STM image of a hydrogenated silicene(3 9 3) surface showing an ordered (3 9 3) structure; b enlargement of the hydrogenated (3 9 3) phase. The white rhombus marks an apparent unit cell of the structure. There are six bright protrusions in one HUC and one protrusion in the other HUC; c STM image showing the comparison between the position of apparent UCs of clean and hydrogenated silicene-(3 9 3); d the clean silicene-(3 9 3) surface is fully recovered after annealing the surface at 450 K. Reprinted with permission from Ref. [37]. Copyright (2015) by the American Physical Society
structure with the same (3 9 3) periodicity can be observed (Fig. 7a). Further increasing the hydrogen dosage does not induce further changes, indicating that the hydrogen adsorption is saturated. A high-resolution image of the hydrogenated structure manifests two in equivalent hydrogenated unit cell (HUCs), one with six bright spots while the other has only one bright spot in the middle (Fig. 7b). The period of the nearest bright ˚ , corresponding to the lattice constant spots is about 3.8 A of silicene-(1 9 1). Figure 7b displays clean silicene(3 9 3) is found in the left part of the image, whereas the right part is hydrogenated. The two sets of (3 9 3) UCs are shifted along Si–Si bond direction by a distance of one Si–Si bond length, resulting from the change in buckling configuration after hydrogenation. Moreover, by annealing the sample to a moderate temperature, about 450 K, dehydrogenation occurs and a clean silicene surface is recovered (Fig. 7d). The work provides a clear and fundamental picture for silicene hydrogenation. And such a uniformly ordered, reversible hydrogenation is useful to modify the electronic properties of silicene for potential applications [37].
123
Sci. Bull.
5.2 Oxidization Oxidation of silicene–silicene oxide is expected to be one of the major steps toward effective engineering silicene’s band structure, and toward an excellent introduction of oxygenated functional groups into Si network. Recent works have revealed the mechanisms of oxygen interaction with silicene sheets during oxidation process. It is found that different phases of silicene layers have significant impact on the oxidation process. In oxidation process, the oxygen adatom is identified on bridge site resulting in the configuration of double-atombonding overbridging O atoms (Od). These oxygen adatoms overbridge neighboring Si atoms leading to Si–O–Si bonds d in silicene oxide. Figure 8a–c suggests pffiffiffiffiffi that pffiffiffiffiffiO is major configuration pffiffiffi pffiffiffi in partially oxidized 13 13; 4 9 4 and 2 3 9 2 3 silicene layers (with respect to 1 9 1 Ag(111)) [34]. Although Od presents in all silicene structures, heights of oxygen adatoms residing on silicene layers are different as demonstrated by STM images in Fig. 8d. Oxygen adatoms
pffiffiffiffiffi pffiffiffiffiffi pffiffiffi pffiffiffi on 13 13 and 2 3 9 2 3 silicene layers appear to be ˚. higher than these adsorbed on 4 9 4 silicene by about 1 A The distances between nearest-neighboring TL Si atoms are pffiffiffiffiffi pffiffiffiffiffi pffiffiffi pffiffiffi ˚ for 13 13 and 2 3 9 2 3 silicene, 5.46 and 3.67 A respectively. These distances are longer than twice of typical ˚ ). Si–O bond lengths in bulk SiO2 (varies from 1.58 to 1.62 A Therefore, both top-layer (TL) and bottom-layer (BL) Si atoms involve in silicon–oxygen bonds as Si(BL)–Od– Si(TL). Oxygen adatoms prefer to reside beside TL Si rather than BL Si as shown in STM results. By contrast, distance between nearest-neighboring TL Si in 4 9 4 silicene is ˚ , indicating different buckling to the other two 2.51 A superstructures. DFT calculation indicates that TL Si atoms in 4 9 4 silicene can decrease height forming BL Si atoms by oxidation, in order to minimize total energy. Therefore, Si(BL)–Od–Si(BL) is also a possible configuration for overbridging oxygen adatoms. As a result, the height of oxygen adatoms on 4 9 4 silicene is smallest in three buckled superstructures. It should be noted that Od is an energetically favoured configuration for oxygen adatoms on
pffiffiffiffiffi pffiffiffiffiffi pffiffiffi pffiffiffi Fig. 8 (Color online) STM and STS results of oxidized silicene in 13 13 (a), 4 9 4 (b) and 2 3 9 2 3 (c) structures. The oxygen adatoms prefer to reside on TL Si atoms at initial oxidation; d line profiles of oxygen adatoms on silicene corresponding to the lines in STM images in a–c, respectively; e–g DFT simulation (top and offfiffiffiffiffi atomic structures for an oxygen adatoms on Ag(111) supported silicene pffiffiffi side views) pffiffiffi p pffiffiffiffiffi monolayers in different superstructures. e 4 9 4, f 2 3 9 2 3, g 13 13: Reprinted with permission from Ref. [34]. Copyright (2014) American Chemical Society
123
Sci. Bull.
all the three silicene structures even with the oxygen dose up to 60 L [34]. Silicene monolayers grown on Ag(111) surfaces demonstrate a band gap that is tunable by oxygen adatoms from semimetallic to semiconducting type. With the use of lowtemperature scanning tunneling microscopy, the work demonstrates the feasibility of tuning the band gap of silicene with oxygen adatoms, which, in turn, expands the base of available two-dimensional electronic materials for devices with properties that are hardly achieved with graphene oxide [34]. ARPES studies revealed the detailed modulation of oxidation on silicene’s electronic structure. The Shockley surface state (SSS) and the typical bulk sp-band of Ag(111) vanished upon the deposition of silicon atoms. The more detailed illustration for the silicene’s ARPES results could be found in Ref. [34]. In oxidized silicene, the signal of metallic hybridized surface band (HSB) is faint, combined with resurgence of SSS band, indicating that the hybridization between 4 9 4 silicene and substrate is broken after oxidization, and silicon atoms in 4 9 4 silicene have a high oxygen reactivity than that of Ag(111) substrate. Such a high oxygen reactivity could be treated as a protection layer for Ag(111) because that SSS states are revived after oxidization. 5.3 Environmental gas adsorption on silicene nanoribbon Because sensors are typically exposed to environmental gases such as N2, O2, CO2, and H2O, the effects of these gases on silicene nanoribbon’s conductance were studied by Osborn and Farajian [70]. Their calculations show that N2 interacts with silicene via physisorption with an energy of 0.42 eV/N2 and is most energetically stable on the edge of the nanoribbon. This N2 adsorption does not lead to significant deformation in the nanoribbon. O2 on the other hand interacts strongly with the pristine silicene nanoribbon with an adsorption energy of 2.96 eV/O2. Upon relaxation, the O2 molecule dissociates in favor of individual Si–O bonds. This large adsorption energy seems to indicate that pristine silicene nanoribbons would easily oxidize under ambient conditions. The effects of environmental gases on the conductance of the nanoribbon were calculated before and after the adsorption of O2 and N2 molecules [70]. The conduction curves depicted in the top panel of Fig. 9 confirm the inert behavior of N2, showing conductance nearly identical to that of the pristine nanoribbon [70]. For oxygen, we see that conductance is significantly reduced while the 0.09 eV band gap is preserved. These results indicate that although the sensing capability of silicene nanoribbons may diminish in an oxygen-containing atmosphere, the capability is preserved in a nitrogen-containing
Fig. 9 (Color online) Quantum conductance modulation resulting from environmental gas molecules adsorptions on nanoribbon. Reprinted with permission from Ref. [70]. Copyright (2014) Springer
atmosphere. The effects of CO2 and H2O adsorption are also investigated. For H2O adsorption, the minimum energy configuration results from water splitting [71] and subsequent attachment of H and OH at the edge (with a binding energy (BE) of 1.62 eV) while CO2 adsorbs via physisorption (BE *0.46 eV). Similar to the case of oxygen adsorption, owing to the destructive effect on the nanoribbon structure, water molecules should also be removed from the environment for proper sensor functionality. Overall, these results indicate that long silicene nanoribbons could provide a unique nano-sensor capable of single-molecule resolution.
6 Argument on existence of
pffiffiffi pffiffiffi 3 3 silicene
Metal/semiconductor structures are of interest in several fields, especially in electronics. Ag–Si system has been heavily studied for several decades for its interesting atomic arrangement, surface electronic states, and the related physic properties. pffiffiffi pffiffiffi Prior to the intense interest in silicene, Ag 3 3 R30° reconstruction with a honeycomb structure on Si(111) has been widely investigated. The honeycomb-chained triangle (HCT) model and inequivalent triangle (IET) model were used to describe the atomic structure of the
123
Sci. Bull.
Fig. 10 (Color online) STM image at RT with a Vbias = -0.6 V, I = 1.0 nA and b Vbias = -0.3 V and I = 0.3 nA; c HCT model; d IET model. Reprinted with permission from Ref. [72]. Copyright (2006) by the American Physical Society
pffiffiffi pffiffiffi Si(111)-( 3 3)-Ag surface, as shown in Fig. 10a, b [72]. At the beginning, it was believed that HCT structure observed at higher temperature (higher than 67 K) is an averaged structural result between IET? and IET- structures at low temperature by thermal Zhang pffiffiffi fluctuations. pffiffiffi et al. [72] found that Si(111)-( 3 3)-Ag with IET structure can be observed at both RT and LT by STM. It suggests that the HCT–IET phase transition does not exist, which may be caused by the conditions of the tip [73]. It is noticeable pffiffiffi that pffiffiffirecent work by Mannix et al. [40] claimed that 3 3 silicene is structurally and electronically identical to HCT on Si(111) at the atomicscale through characterization. p ffiffiffi pffiffiSTM ffi pffiffiffi As pffiffiffishown in Fig. 11, both 3 3 silicene and 3 3 Ag are honeycomb structures with a similar constant lattice of 0.65 nm and a height of monolayer of 0.31 though this p work pffiffiffi nm. pffiffiEven ffi ffiffiffi cannot demonstrate that 3 3 silicene is the 3 pffiffiffi 3 Ag reconstruction on the surface of Si(111), debates about the real atomic structure of silicene have been raised. Further pffiffiffi works pffiffiffi about atomic structure of silicene, especially 3 9 3 phase, are urgently needed to explain those novel physic and chemical properties observed on silicene.
pffiffiffi pffiffiffi Fig. 11 (Color online) a Growth schematic for Si on Ag(111) and Ag on Si(111). b STM topography of the 3 3 phase for Si on Ag(111) (scale bar pffiffiffi p ffiffiffi = 50 nm, Vbias = -1.0 V, I = 400 pA); c line profile revealing a 0.31 nm step height. d Atomic-scale STM topography of the 3 pffiffiffi p3ffiffiffi phase for Si on Ag(111); e line profile (scale bar = 1 nm, Vbias = 0.3 V, I = 1.0 nA). f Atomic-scale STM topography of the lateral atomic 3 3 phase for Ag on Si(111); g line profile (scale bar = 1 nm, Vbias = -1.0 V, I = 100 pA) revealing an indistinguishable pffiffiffi pffiffiffi periodicity compared to the Vbias = -1.0 V phase for Si on Ag(111) shown in part e. h STM topography of the 3 3 phase for pffiffiffiAg pon ffiffiffi Si(111); i line profile (scale bar = 50 nm, Vbias = 0.85 V, I = 1.4 nA), showing identical step heights and similar island shapes to the 3 3 phase for Si on Ag(111) shown in part a. Reprinted with permission from Ref. [40]. Copyright (2014) American Chemical Society
123
Sci. Bull.
7 Summary The research on silicene has promoted a spectacular step since epitaxial silicene layers were first fabricated in 2012. This novel material has evoked intensive interest of scientists in the wide regions. Although this is a fairly new 2D material, recent outcomes on insight into silicene demonstrated encouraging features shedding a light on the electronics industry and giving a promising route toward applications of electronic device. Acknowledgments This work was supported by the Australian Research Council (ARC) through Discovery Project (DP 140102581), LIEF Grants (LE100100081 and LE110100099). Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Novoselov KS, Geim AK, Morozov SV et al (2004) Electric field effect in atomically thin carbon films. Science 306:666–669 2. Feng BJ, Ding ZJ, Meng S et al (2012) Evidence of silicene in honeycomb structure of silicon on Ag(111). Nano Lett 12:3507–3511 3. Meng L, Wang YL, Zhang LZ et al (2013) Buckled silicene formation on Ir(111). Nano Lett 13:685–690 4. Fleurence A, Friedlein R, Ozaki T et al (2012) Experimental evidence for epitaxial silicene on diboride thin films. Phys Rev Lett 108:245501 5. Gao JF, Zhao JJ (2012) Initial geometries, interaction mechanism and high stability of silicene on Ag(111) surface. Sci Rep 2:861 6. Kara A, Enriquez H, Seitsonen AP et al (2012) A review on silicene—new candidate for electrons. Surf Sci Rep 67:1–18 7. Watanabe K, Takashi T, Kanda H (2004) Direct-bandgap properties and evidence for ultraviolet lasing of hexagonal boron nitride single crystal. Nat Mater 3:404–409 8. Seifert G, Terrones H, Terrones M et al (2000) Structure and electronic properties of MoS2 nanotubes. Phys Rev Lett 85:146–149 9. Braga D, Lezama IG, Berger H et al (2012) Quantitative determination of the band gap of WS2 with ambipolar ionic liquidgated transistors. Nano Lett 12:5218–5223 10. Lalmi B, Oughaddou H, Enriquez H et al (2010) Epitaxial growth of a silicene sheet. Appl Phys Lett 97:223109 11. O’Hare A, Kusmartsev FV, Kugel KI (2012) A stable ‘‘flat’’ form of two-dimensional crystals: could graphene, silicene, germanene be minigap semiconductors? Nano Lett 12:1045–1052 12. Resta A, Leoni T, Barth C et al (2013) Atomic structures of silicene layers grown on Ag(111): scanning tunnelling microscopy and noncontact atomic force microscopy observations. Sci Rep 3:2399 13. Vogt P, De Padova P, Quaresima C et al (2012) Silicene: compelling experimental evidence for graphenelike two-dimensional silicon. Phys Rev Lett 108:155501 14. Yan JA, Stein R, Schaefer DM et al (2013) Electron–phonon coupling in two-dimensional silicene and germanene. Phys Rev B 88:121403 15. Chen L, Li H, Feng BJ et al (2013) Spontaneous symmetry breaking and dynamic phase transition in monolayer silicene. Phys Rev Lett 110:085504
16. Liu HS, Gao JF, Zhao JJ (2013) Silicene on substrates: a way to preserve or tune its electronic properties. J Phys Chem C117:10353–10359 17. Gao JF, Zhang JF, Liu HS (2013) Structures, mobilities, electronic and magnetic properties of point defects in silicene. Nanoscale 5:9785–9792 18. Feng BJ, Li WB, Qiu JL et al (2015) Variable coupling strength of silicene on Ag(111). Chin Phys Lett 32:037302 19. Feng Y, Feng BJ, Xie ZJ et al (2014) Observation of a flat band in silicene. Chin Phys Lett 31:127303 20. Liu HS, Han NN, Zhao JJ (2014) Band gap opening in bilayer silicene by alkali metal interaction. J Phys Condens Matter 26:475303 21. Chen L, Liu CC, Feng BJ et al (2012) Evidence for Dirac fermions in a honeycomb lattice based on silicon. Phys Rev Lett 109:056804 22. Liu CC, Feng WX, Yao YG (2011) Quantum spin hall effect in silicene and two-dimensional germanium. Phys Rev Lett 107:076802 23. Quhe R, Fei RX, Liu QH et al (2012) Tunable and sizable band gap in silicene by surface adsorption. Sci Rep 2:853 24. Ezawa M (2012) Valley-polarized metals and quantum anomalous hall effect in silicene. Phys Rev Lett 109:055502 25. Ezawa M (2013) Photoinduced topological phase transition and a single Dirac-cone state in silicene. Phys Rev Lett 110:026603 26. Chiappe D, Grazianetti C, Tallarida G et al (2012) Local electronic properties of corrugated silicene phases. Adv Mater 24:5088–5093 27. Enriquez H, Vizzini S, Kara A et al (2012) Silicene structures on silver surfaces. J Phys Condens Matter 24:314211 28. Jamgotchian H, Colignon Y, Hamzaoui N et al (2012) Growth of silicene layers on Ag(111): unexpected effect of the substrate temperature. J Phys Condens Matter 24:172001 29. Ferrari AC, Basko DM (2013) Raman spectroscopy as a versatile tool for studying the properties of graphene. Nat Nanotechnol 8:235–246 30. Liu Z, Suenaga K, Harris PJF et al (2009) Open and closed edges of graphene layers. Phys Rev Lett 102:015501 31. Li WX, Stampfl C, Scheffler M (2002) Oxygen adsorption on Ag(111): a density-functional theory investigation. Phys Rev B 65:075407 32. Takeda K, Shiraishi K (1994) Theoretical possibility of stage corrugation in Si and Ge analogs of graphite. Phys Rev B 50:14916–14922 33. Lin CL, Arafune R, Kawahara K (2012) Structure of silicene grown on Ag(111). Appl Phys Express 5:45802 34. Du Y, Zhuang JC, Liu HS et al (2014) Tuning the band gap in silicene by oxidation. ACS Nano 8:10019–10025 35. Aufray B, Kara A, Vizzini S et al (2010) Graphene-like silicon nanoribbons on Ag(110): a possible formation of silicene. Appl Phys Lett 96:183102 36. Tchalala MR, Enriquez H, Mayne AJ et al (2013) Formation of one-dimensional self-assembled silicon nanoribbons on Au(110)(2 9 1). Appl Phys Lett 102:083107 37. Qiu JL, Fu HX, Xu Y et al (2015) Ordered and reversible hydrogenation of silicene. Phys Rev Lett 114:126101 38. Arafune R, Lin CL, Kawahara K et al (2013) Structural transition of silicene on Ag(111). Surf Sci 608:297–300 39. Shirai T, Shirasawa T, Hirahara T et al (2014) Structure determination of multilayer silicene grown on Ag(111) films by electron diffraction: evidence for Ag segregation at the surface. Phys Rev B 89:241403 40. Mannix AJ, Kiraly B, Fisher BL et al (2014) Silicon growth at the two-dimensional limit on Ag(111). ACS Nano 8:7538–7547
123
Sci. Bull. 41. Feng JG, Wagner SR, Zhang PP (2015) Interfacial coupling and electronic structure of two-dimensional silicon growth on the Ag(111) surface at high temperature. Sci Rep 5:10310 42. Liu ZL, Wang MX, Xu JP et al (2014) Various atomic structures of monolayer silicene fabricated on Ag(111). New J Phys 16:075006 43. Kopnin NB, Sonin EB (2008) BSC superconductivity of Dirac electrons in graphene layers. Phys Rev Lett 100:246808 44. Scalise E, Houssa M, Pourtois G et al (2013) Vibrational properties of silicene and germanene. Nano Res 6:19–28 45. De Padova P, Ottaviani C, Quaresima C et al (2014) 24 h stability of thick multilayer silicene in air. 2D Mater 1:021003 46. Molle A, Grazianetti C, Chiappe D et al (2013) Hindering the oxidation of silicene with non-reactive encapsulation. Adv Funct Mater 23:4340–4344 47. Cinquata E, Scalise E, Chiappe D et al (2013) Getting through the nature of silicene: an sp2–sp3 two-dimensional silicon nanosheet. J Phys Chem C 117:16719–16724 48. Zhuang JC, Xu X, Du Y et al (2015) Investigation of electron– phonon coupling in epitaxial silicene by in situ Raman spectroscopy. Phys Rev B 91:161409 49. Ferrari AC, Meyer JC, Scardaci V et al (2006) Raman spectrum of graphene and graphene layers. Phys Rev Lett 97:187401 50. McMillan W (1968) Transition temperature of strong-coupled superconductors. Phys Rev 167:331 51. Si C, Liu Z, Duan WH et al (2013) First-principles calculations on the effect of doping and biaxial tensile strain on electron– phonon coupling in graphene. Phys Rev Lett 111:196802 52. Chen L, Feng BJ, Wu KH (2013) Observation of a possible superconducting gap in silicene on Ag(111) surface. Appl Phys Lett 102:081602 53. Tao L, Cinquata E, Chippe D et al (2015) Silicene field-effect transistors operating at room temperature. Nat Nanotechnol 10:227–231 54. Le Lay G (2015) 2D materials: silicene transistors. Nat Nanotechnol 10:202–203 55. Guzma´n-Verri GG, Lew Yan Voon LC (2007) Electronic structure of silicon-based nanostructures. Phys Rev B 76:075131 56. Cahangirov S, Topsakal M, Aktu¨rk E et al (2009) Two- and onedimensional hneycomb structures of silicon and germanium. Phys Rev Lett 102:236804 57. Lebe`gue S, Eriksson O (2009) Electronic structure of two-dimensional crystals from ab initio theory. Phys Rev B 79:115409
123
58. Lin CL, Arafune R, Kawahara K et al (2013) Substrate-induced symmetry breaking in silicene. Phys Rev Lett 110:076801 59. Tsoutsou D, Xenogiannopoulou E, Golias E et al (2013) Evidence for hybrid surface metallic band in (4 9 4) silicene on Ag(111). Appl Phys Lett 103:231604 60. Xu X, Zhuang JC, Du Y et al (2014) Effects of oxygen adsorption on the surface state of epitaxial silicene on Ag(111). Sci Rep 4:7543 61. Avila J, De Padova P, Cho S et al (2013) Presence of gapped silicene-derived band in the prototypical (3 9 3) silicene phase on silver (111) substrates. J Phys Condens Matter 25:262001 62. De Padova P, Vogt P, Resta A et al (2013) Evidence of Dirac fermions in multilayer silicene. Appl Phys Lett 102:163106 63. De Padova P, Avila J, Resta A et al (2013) The quasiparticle band dispersion in epitaxial multilayer silicene. J Phys Condens Matter 25:382202 64. Rutter GM, Crain JN, Guisinger NP et al (2007) Scattering and interference in epitaxial graphene. Science 317:219–222 65. Feng BJ, Li H, Liu CC et al (2013) Observation of Dirac cone warping and chirality effects in silicene. ACS Nano 7:9049–9054 66. Sofo JO, Chaudhari AS, Barber GD (2007) Graphane: a twodimensional hydrocarbon. Phys Rev B 75:153401 67. Elias DC, Nair RR, Mohiuddin TMG et al (2009) Control of graphene’s properties by reversible hydrogenation: evidence for graphene. Science 323:610–613 68. Balog R, Jørgensen B, Nilsson L et al (2010) Bandgap opening in graphene induced by patterned hydrogen adsorption. Nat Mater 9:315–319 69. Andreev T, Barke I, Ho¨vel H (2004) Adsorbed rare-gas layers on Au(111): shift of the Shockley surface state studied with ultraviolet photoelectron spectroscopy and scanning tunnelling spectroscopy. Phys Rev B 70:205426 70. Osborn TH, Farajian AA (2014) Silicene nanoribbons as carbon monoxide nanosensors with molecular resolution. Nano Res 7:945–952 71. Konecˇny´ R, Doren DJ (1997) Adsorption of water on Si(100)(2 9 1)—a study with density-functional theory. J Chem Phys 106:2426–2435 72. Zhang HM, Gustafsson JB, LSO (2006) Surface atomic pffiffiJohansson ffi pffiffiffi structure of Ag/Si(111)-( 3 3). Phys Rev B 74:201304 pffiffiffi pffiffiffi 73. Sato N, Nagao T, Hasegawa S (1999) Si(111)-( 3 3)-Ag surface at low temperatures: symmetry breaking and surface twin boundaries. Surf Sci 442:65–73
www.scibull.com www.springer.com/scp
Review
Materials Science
Honeycomb silicon: a review of silicene Jincheng Zhuang • Xun Xu • Haifeng Feng Zhi Li • Xiaolin Wang • Yi Du
•
Received: 13 July 2015 / Accepted: 14 August 2015 Ó Science China Press and Springer-Verlag Berlin Heidelberg 2015
Abstract Silicene, a new allotrope of silicon in a twodimensional honeycomb structure, has attracted intensive research interest due to its novel physical and chemical properties. Unlike carbon atoms in graphene, silicon atoms prefer to adopt sp2/sp3-hybridized state in silicene, enhancing chemical activity on the surface and allowing tunable electronic states by chemical functionalization. The silicene monolayers epitaxially grown on Ag(111) surfaces demonstrate various reconstructions with different electronic structures. In this article, the structure, phonon modes, electronic properties, and chemical properties of silicene are reviewed based on theoretical and experimental works in recent years. Keywords Silicene Electronic states Chemical functionalization
1 Introduction The discovery of graphene in 2004 by Novoselov et al. [1] demonstrated that, for the first time, stable and singleatom-thick two-dimensional (2D) materials could be exfoliated from van der Waals solids, and these materials possess unique and fascinating properties due to their exotic electronic structures [1–6]. Since then, graphene has been one of the most extensively studied materials due to the wealth of anomalous physical and chemical phenomena that occur when charge and heat transport is confined to a plane. The rapid and recent advances in this carbon-based
J. Zhuang X. Xu H. Feng Z. Li X. Wang Y. Du (&) Institute for Superconducting and Electronic Materials (ISEM), University of Wollongong, Wollongong, NSW 2525, Australia e-mail: [email protected]
2D material have raised teasing questions on exploring new 2D materials with distinct and exotic properties. Such possibilities have promoted a completely new research field of 2D crystals. There are many 2D materials that have been explored, and they exhibit rich spectrum of properties. For example, hexagonal boron nitride (h-BN) with the closest structure to graphene is an insulator [7], while monolayers MoS2 and WS2 are semiconductors with direct band gap [8, 9]. The abundant variety of properties in 2D materials suggests the candidate potential for the device engineering and application in sensing, photonics and energy storing, etc. Great efforts have been made to incorporate 2D atomic layers into devices to exhibit exceptional performance. It has been extremely difficult so far, however, to develop reliable and durable applications, especially in electronics, based on 2D materials due to their incompatibility with current semiconductor-based electronic techniques. Moreover, it is extremely hard to obtain single crystal with large size for characterization and device fabrication. Searching for the novel 2D materials is therefore highly desirable, especially for next-generation low-cost super-performance electronic devices. Silicene is similar to graphene in that it is single atom thick, and it has the same characteristic honeycomb structure. It is a big challenge, however, to synthesize silicene, despite the similarity of electronic configurations between carbon atoms and silicon atoms. Nevertheless, the sp2 hybridization is more stable than sp3 hybridization for carbon, but situation is reversed for silicon. Thus, silicon atoms are energetically not favorable to spontaneously form silicene, and this limits conventional chemical or physical methods to prepare silicene. Despite the difficulty in fabrication, many theoretical calculations have predicted an exciting and rich physics in silicene [10–12]. For instance, the Dirac fermion state was demonstrated in a
123
Sci. Bull.
linear dispersion band structure close to Fermi level by density functional theory (DFT) calculations [11]. The quantum spin Hall effect (QSHE) and quantum anomalous Hall effect (QAHE) were also predicted in silicene [13– 15]. The electronic properties of the silicene could be modulated by effect of substrate [16], defect [17, 18], interlayer coupling [19], and metal interaction [20]. These theoretical works have paved the way of exploiting the anisotropic transport behavior and experimentally investigating exotic properties of silicene [21–31]. Moreover, silicene is derived from silicon—the precise substance that silicon chips are made of. As such, it is inherently compatible with current chip-production processes, which is advantageous in contrast to graphene. There is no doubt that silicene could be a material of the future. On the energy front, silicene is expected to make solar power more practical and efficient by improving the storage capacity and output speed of solar cells [6]. Silicene may also yield advances in medical sensors (which monitor heart rates and blood flow), sharpen X-ray imaging, and improve DNA sequencing [3–6]. In addition, it will make health-monitoring devices faster, cheaper, and even disposable. It can create more efficient batteries, paper-thin cell phones, and flexible screens.
2 2D structure of silicene Silicene, silicon-based graphene-like sheets, has been proposed recently by first-principle total-energy calculations [32]. However, different from the planar structure of graphene with sp2 hybridization network, silicene is a lowbuckled structure due to mix of sp2 and sp3 hybridization. Due to the competition between sp2 hybridization and sp3 hybridization, rich phases of silicene have been reported grown on varies substrates, including Ag(111) [2, 10, 13, 28, 33, 34], Ag(110) [35], Au(110) [36], Ir(111) [3], and ZrB2-covered Si(111) [4]. Here, we focus on the structure of silicene phases grown on Ag(111). 2.1 3 9 3 silicene 3 9 3 silicene is 4 9 4 reconstruction with respect to the Ag(111) substrate or 3 9 3 reconstruction with respect to 1 9 1 silicene [10]. Large-scale 3 9 3 silicene films can be easily prepared by keeping substrate temperature around 420 K during deposition. In 2012, three groups reported detailed structure models of 3 9 3 silicene by comparing scanning tunnel microscope (STM) results with theoretical calculation [2, 13, 20]. Figure 1a shows the phase exhibiting honeycomb structures (labeled H). The other phase consisting of close-packed protrusions (labeled T) is shown in Fig. 1b. These two phases coexist with each
123
Fig. 1 (Color online) STM images and structure models of H phase silicene (a) and T phase silicene (b). Reprinted with permission from Ref. [2]. Copyright (2012) American Chemical Society
other, indicating similar formation energy and stability. However, phase T tends to form at a lower Si coverage and lower temperature compared to phase H. At higher Si coverage, phase T will disappear and be replaced by phase H. Thus, phase H is regarded as the stable phase of 3 9 3 silicene, and phase T is a precursor of phase H. Based on the structure in STM observation and first-principle calculation, a structure model of a low-buckled honeycomb structure with missing of a hexagonal silicon rings at the corner in each 3 9 3 cell is proposed, as shown in structure model in Fig. 1b. Although similar STM results have been represented by Liu et al. [20] and Vogt et al. [13], different structure models have been proposed, where the hole in STM image is attributed to six Si atoms buckled upward. The recent work of hydrogen absorption on 3 9 3 silicene reported by Qiu et al. [37] is helpful to clarify this controversy. It is found that absorption of hydrogen changes buckling structure of 3 9 3 silicene by detailed discussion on STM results and theoretical calculation; they proposed a structure model with intact honeycomb structure. 2.2
pffiffiffi pffiffiffi 7 7 silicene
With increasing substrate p temperature ffiffiffi pffiffiffi to 480 K, a new phase of silicene named as 7 7 silicene emerges [2]. This phase usually manifests itself as a defective moire´ pattern with a period about 3.8 nm, as shown in Fig. 2. Figure 2b shows that this moire´ pattern consists of hexagonal rings at bright part and defective part between them. By first-principle they construct a pffiffiffi calculation, pffiffiffi structure model for 7 7 silicene (Fig. 2d, f). They attribute the bright parts of moire´ pattern to the position deviation of silicon atoms from that of Ag(111). Thus, the strong interaction between silicene and Ag(111) substrate will make hexagonal rings stable and complete. In other parts of moire´ pattern, because of larger deviation of position between silicon atoms and that of substrate, the interaction breaks these hexagonal rings.
Sci. Bull.
and sixfold symmetry of the surface structure. Recently, Shirai et al. [39] argued that substrate pAg ffiffiffi atoms pffiffiffi may segregate on silicene surface forming a 3 3 superpffiffiffi pffiffiffi structure which is similar with 3 3 silicene from tensor low-energy electron-diffraction (LEED) pffiffiffi panalysis. ffiffiffi The similarity p between the proposed Ag 3 3 superffiffiffi pffiffiffi structure and 3 3 silicene is also reported by very recent STM worksp[40, ffiffiffi p41]. ffiffiffi As a result, these works question whether 3 3 silicene is a true layered structure. The detailed is p performed pffiffiffi discussion pffiffiffi ffiffiffi pffiffiffi in Sect. 6. Excepting for 3 3 silicene, 7 7 silicene and pffiffiffi pffiffiffi 3 3 silicene, various superstructures with different buckling patterns and periodicity are also reported [42]. But systematic experimental and theoretical investigations are lacking for these phases. Even for the three phases discussed above, there are more controversies than common views. Detailed studies on structure of silicene phases are needed.
3 Phonon modes in silicene
Fig. 2 (Color online) a A derivative STM image (200 nm 9 200 nm, Vtip = 1.43 V) of 0.9 ML (ML, monolayer) silicon atoms deposited on Ag(111) surface at substrate temperature of 480 K; b high-resolution STM image (15 nm 9 15 nm, Vtip = -1.0 V) showing the atomic structure of moire´ patterns. The bright areas exhibit complete honeycomb rings with a period of 1.0 nm, while other areas are defective and disordered. The angle between the orientation of the hexagonal rings and the direction of moire´ patterns is 30°; c dI/dV spectra taken at the moire´ pattern phase, in which a peak at 0.3 and pffiffiffi V p ffiffiffi a shoulder at 0.9 V are observed; d calculated model of 7 7 superstructure of silicene; e, f experimental and simulated STM images (1.0 eV above Fermi energy) showing the similar structure features and unit cell of lattice. Reprinted with permission from Ref. [2]. Copyright (2012) American Chemical Society
2.3
pffiffiffi pffiffiffi 3 3 silicene
Further increasing substrate temperatureptoffiffiffi 500pffiffiKffi leads to a most stable silicene phase named as 3 3 silicene. This phase is the most extensively investigated phase of silicene. Meanwhile, it may be the most controversy phase. Wu and co-workers [2, 21] proposed a structural model based on a honeycomb lattice with a unit cell (UC) that consists of three layers of silicon atoms as shown in Fig. 3. This model is suspected by Arafune et al. [38], who proposed a bilayer structure judging from the layer thickness
Raman spectroscopy is widely used in structural characterization, electron–phonon coupling (EPC) probe, and the phonon dynamics investigation in 2D Dirac fermions system [43]. In silicene, the long-wavelength optical E2g phonon mode at the C point of the Brillouin zone (BZ), which corresponds to the relative displacement of nonequivalent neighbor silicon atoms [14], is of particular interest. Any perturbations due to this buckled structure will effectively induce the direct electronic transitions across the Dirac point, that is, E2g phonons couple to lowenergy excitations. The vibrational properties of silicene are firstly investigated by means of DFT calculations [44]. By considering the small buckling of silicene and favorable energy, the non-resonance Raman spectra of free-standing silicene are presented by a prime peak (E2g) located at around 570 cm-1. It is also predicted that the Raman peaks attributed to the defects at the edges distribute at low frequencies than E2g peak. However, it has been an experimental challenge to verify vibrational properties in this silicon-based 2D Dirac fermion system because the monolayer silicene is unstable under ambient conditions [45]. To prevent the oxidation of silicene, co-deposition of the Al and O2 to form Al2O3 capping layer in ultra-high vacuum (UHV) ambient is intended before the ex situ Raman measurement, then resulting in an intact Al2O3/silicene/Ag heterostructure [46, 47]. Combined with DFT calculation, the Raman spectra of different silicene superstructures display that E2g peak locates at around 520 cm-1, much smaller than the value of simulated free-standing silicene [44]. Furthermore, Raman spectrum is also investigated depending on excitation
123
Sci. Bull.
pffiffiffi pffiffiffi Fig. 3 (Color online) a Top view and b side view of the lattice geometry of the low-buckled silicene structure (AB) and 3 3 superstructure (AB-A), respectively. Note that in the superstructure, a Si atom with p planar (2/3, 2/3) is pulled downward; c the structural phase ffiffiffi pcoordinate ffiffiffi transition diagram of silicene depending on the lattice constants of 3 3 superstructure; d a larger schematic model illuminating the pffiffiffi pffiffiffi honeycomb structure of 3 3 reconstructed silicene. Reprinted with permission form Ref. [21]. Copyright (2014) by the American Physical Society
energy to probe the resonant behavior [47]. The results indicate that the monolayer silicene is multi-hybridized by sp2 structure and sp3 structure. Nevertheless, the oxidation cannot be excluded in this ex situ measurement. The broad shoulder at lower wavenumber (450–510 cm-1) next to the silicene signature E2g peak, which is ascribed to bucklinginduced vibrational modes in [47], is actually associated with the Si–O bonds due to Si sp3 hybridization [34]. Due to the low oxygen adsorption energy on the Si surface with dangling bonds, the monolayer silicene is extremely sensitive to the oxygen. The structure of monolayer silicene would be completely destroyed in the condition of exposure to 600 Langmuir (L) oxygen. Therefore, in situ Raman measurement under UHV is expected to be performed to identify silicene in different phases and reveal the details of the phonon modes as well as their relationship to the silicene electronic properties. The in situ Raman scattering studies of phonon modes in epitaxial silicene with different reconstructions on a Ag(111) surface are performed to eliminate the effect of oxidization [48]. Combined with STM, it is implied that the characteristic Raman peaks of silicene are sensitive to the electronic structures, making it possible to unambiguously distinguish between silicene phases in a fast and non-
123
destructive way. Compared to previous research,pZhuang ffiffiffi pffiffiffi et al. [48] firstly reports the Raman spectrum of 3 3 phase as a function of coverage. The first-order asymmetric peak, interpreted as the zone-center E2g vibrational mode, of all silicene phases locates at around 530 cm-1, as shown in Fig. 4. There are two major differences between the Raman results of different layers. Firstly, no pffiffiffidefect pffiffiffi peak could be observed in the Raman results of 3 3 film, which is consistent with the STM pffiffiffi results pffiffiffi that scarce boundaries could be found in the 3 3 silicenepfilm. ffiffiffi Secondly, the much stronger E2g mode is observed in 3 p ffiffiffi 3 silicene than that of monolayer silicene. The E2g vibrational mode is generated by the bond stretching of all sp2 silicon atoms, and is the fingerprint of honeycomb pffiffiffi pffiffiffi lattice [43, 48]. Therefore, the sp2 strength pffiffiffi in pffiffi3ffi 3 silicene film is much higher than that in 7 7/3 9 3 silicene film. It should be noted that experimental frequency of E2g mode is smaller than that of free-standing silicene [44]. There are two factors which could modulate the E2g mode in 2D materials. One is the strain effect, reflected by the ˚ ) compared to freestretched Si–Si bond length (*2.35 A ˚ ) [44, 47, 48]. The difference in standing silicene (2.28 A bond length could induce 5 % in-plane tension, softening
Sci. Bull.
pffiffiffiffiffi pffiffiffiffiffi pffiffiffi pffiffiffi Fig. 4 (Color layer; bffiffiRaman spectra of 3 3 silicene pffiffiffiffiffionline) pffiffiffiffiffia Raman spectra of Ag(111) substrate, Si(111), and 13 13/4 9 4 bufferp ffiffiffi p ffi grown onffiffiffi 13 13/4 9 4 buffer layer with different coverage, SL denotes thepcoverage p ffiffiffi p ffiffiffiffiffi pffiffiffiffiffiof the 3 3 silicene layer; c Raman spectra of 3 3 silicene layer (coverage of 0.3 ML) grown on Ag(111) surface and 13 13/4 9 4 buffer layer, respectively. Reprinted with permission from Ref. [48]. Copyright (2015) by the American Physical Society
the E2g phonon frequency with a value of 50 cm-1. The other factor is the carrier doping. In fact, electron or hole doping in silicene can result in a hardening of the mode frequency in silicene [14], e.g., E2g mode will shift to higher frequency due to charge doping, as shown in Fig. 5. Angle-resolved photoemissionpffiffiffi spectroscopy (ARPES) pffiffiffi verifies that the Dirac point of 3 3 silicene grown on silver substrate is located at *0.33 eV below the Fermi surface due to electron doping from the Ag(111) substrate. E2g vibration at the zone center couples with Dirac fermions at the zone boundary, which is allowed by silicene lattice symmetry. Since the Fermi energy (EF) of silicene is lifted by electron doping, E2g mode in silicene is hardened. The E2g frequency upshifting by about 10 cm-1 agrees well with previous simulation results [14]. Since EPC strength depends primary on the phonon frequencies in 2D materials [50, 51], the shift of E2g peak in silicene induced by both strain effect and carrier doping effect could modulate the EPC strength. In general, EPC can be characterized by a dimensionless parameter c proportional to x-2 [49–51]. This relation originates from the zero-point oscillation amplitude induced by large deformation and the energy reconstruction in the perturbation theory. Both factors correspond exactly to the strain effect presenting in 2D materials. Therefore, the value of x reflects the EPC strength. According to E2g peak shift from 570 cm-1 (FS) to 520 cm-1 (strained silicene), the EPC in silicene can be enhanced up to 20 %. The enhancement of EPC is of particular interest because it gives rise to superconductivity in Bardeen–Cooper–Schrieffer (BCS) superconductors. The Raman results consist with recent observations on the existence of a superconducting gap in silicene layers [52]. Detecting the relationship between EPC and Dirac fermions through Raman scattering will
provide a direct route to investigate the exotic property in buckled two-dimensional honeycomb materials. Five distinct Raman peaks at low range pffiffiffi wave-number pffiffiffi (220–470 cm-1) are observed in 3 3 silicene [48]. The intensities of these peaks can be scaled well with each other and show a strong dependence on the coverage. The special phonon modes correspond to 2D electron gas scattering at the edge sites in the silicene [48]. The edgeinduced Raman peaks reflect the unique buckled characteristic in silicene. However, the detailed relationship of each peak with different edges, e.g., armchair edge and zigzag edge, is still unknown; tip-enhanced Raman spectroscopy is expected to advance insights into the edge effects on phonon modes in this low-buckled 2D material. In fact, Raman spectroscopy has been adopted to monitor the air ability of encapsulated epitaxial silicene [53, 54], and is crucial for the confirmation of the integrity of silicene in the field-effect transistor. Seven days after transfer, the Raman spectrum remains the same as that of freshly growth silicene, implying that the pristine silicene is preserved [53]. The Raman spectroscopy allows unambiguous, high-throughput, nondestructive identification of epitaxial silicene.
4 Electronic structures in silicene Silicene is predicted to share the same novel electronic properties as graphene [32, 55–57]. The electronic p- and p*-bands derived from the Si 3pz orbital disperse linearly to cross around the Dirac point, and charge carriers behave like massless Dirac fermions [20]. However, due to sp2– sp3-hybridized atomic arrangement [13], monolayer silicene shows a strong interfacial coupling with substrate. In
123
Sci. Bull.
Fig. 5 (Color online) a Raman spectrum of E2g peak for different coverage samples. The E2g mode frequency of FS silicene is plotted as Ref. for 0.3 SL sample and 0.8 SL sample. The dashed lines are used to mark the position of E2g mode; c ARPES [44]; b fitted resultspof ffiffiffi E2g pffiffipeak ffi results of epitaxial 3 3 silicene layer. The left part shows that two faint linear dispersed bands crossed at BZ center C point. The right part displaying constant-energy cuts of the spectral function at different binding energies verifies that both bands origin from a Dirac cone structure; d E2g mode frequency as a function pffiffiffi pffiffiof ffi coverage. Both (d) and the inset of (d) are sketched to illustrate the strain and doping effects on Raman peak position of E2g mode in 3 3 silicene, in which the strain effect softens the E2g mode, while charge doping will upshift the E2g mode. Reprinted with permission from Ref. [48]. Copyright (2015) by the American Physical Society
this case, the electrons are delocalized into the substrate, which could induce the absence of Landau levels and lose its Dirac fermion characteristics [58]. The ARPES results display that no Dirac cones could be observed in the K points of monolayer silicene BZ, instead, a dominant hybrid surface metallic band is generated by combined effect of the monolayer silicene and Ag(111) sp-band [19, 59–61]. Therefore, the electronic structure of monolayer silicene has been intensively modulated by the substrate, resultingpfrom ffiffiffi pthe ffiffiffi buckling-induced distortion. The 3 3 silicene is in indirect contact with the silver substrate, which could induce the low-buckled pffiffiffi structure and a weak interfacial coupling between 3 p ffiffiffi 3p silicene ffiffiffi pffiffiand ffi the substrate. Thus, the electronic structure of 3 3 silicene is expected to diverse from that of monolayer silicene, and to present the Dirac fermion characteristics. The pronounced quasiparticle interferences (QPI) patterns were detected in the scanning tunneling
123
spectroscopy (STS) measurement [19]. The linear band dispersion is deduced indicating the existence of pffiffiffi from pffiffiQPI, ffi Dirac fermion in 3 3 silicene [19]. The direct signature of Dirac fermion in this phase has been observed, in 0 which the Dirac cones appear on the K pand of ffiffiffi Kpffiffipoints ffi 1 9 1 silicene BZ as well as C points of 3 3 silicene BZ, as shown in Fig. 6 [63, 64]. The Dirac point is below the Fermi surface due to the electron doping from Ag(111) substrate [48]. The calculated Fermi velocity from ARPES results at room temperature is as high as 0.3 9 106 m s-1, still a bit smaller than that of graphene [63, 65], but is high enough for potential silicene-based application. The electronic structure, investigated by both STS and ARPES measurements, implies two critical discoveries in silicene system. Firstly, the electronic band of single layer silicene is drastically modulated by Si–Ag coupling. Secondly, the pDirac ffiffiffi pffiffiffifermion characteristics have been observed in 3 3 silicene, implying the weak coupling
Sci. Bull.
pffiffiffi pffiffiffi Fig. 6 (Color online) Dirac fermionics of multilayer ( 3 3) R30° silicene islands. a Scheme of Brillouin zones; b K- and V-shape linear p and p* silicene bands at C0; c waterfall line profiles of the dispersion displayed in b as a guide for the eyes. Reprinted with permission for Ref. [62]. Copyright (2013), AIP Publishing LLC
of Si atoms in different layers. Since core for the application of 2D materials is the presence of Dirac fermion properties [66], the electronic band investigations indicate that silicene is highly promising for the fabrication of innovative device with direct compatibility with silicon micro- and nano-technologies.
5 Chemical properties of silicene 5.1 Hydrogenation Hydrogenation was found as a useful chemical method to modulate the electronic property of graphene, leading to a band-gap opening [67–69]. In contrast to the graphene case, where hydrogen tends to form clusters, hydrogenated silicene exhibits a perfectly long-range ordered structure. Combining with first-principle calculations, it has been determined that there are seven hydrogen atoms in one (3 9 3) UC and that the buckling configuration of Si atoms in silicene spontaneously rearranges upon hydrogenation [37]. Figure 7 shows the typical changes induced by the hydrogenation of silicene-(3 9 3). Upon exposure of 900 L hydrogen at room temperature, a perfectly ordered
Fig. 7 (Color online) a STM image of a hydrogenated silicene(3 9 3) surface showing an ordered (3 9 3) structure; b enlargement of the hydrogenated (3 9 3) phase. The white rhombus marks an apparent unit cell of the structure. There are six bright protrusions in one HUC and one protrusion in the other HUC; c STM image showing the comparison between the position of apparent UCs of clean and hydrogenated silicene-(3 9 3); d the clean silicene-(3 9 3) surface is fully recovered after annealing the surface at 450 K. Reprinted with permission from Ref. [37]. Copyright (2015) by the American Physical Society
structure with the same (3 9 3) periodicity can be observed (Fig. 7a). Further increasing the hydrogen dosage does not induce further changes, indicating that the hydrogen adsorption is saturated. A high-resolution image of the hydrogenated structure manifests two in equivalent hydrogenated unit cell (HUCs), one with six bright spots while the other has only one bright spot in the middle (Fig. 7b). The period of the nearest bright ˚ , corresponding to the lattice constant spots is about 3.8 A of silicene-(1 9 1). Figure 7b displays clean silicene(3 9 3) is found in the left part of the image, whereas the right part is hydrogenated. The two sets of (3 9 3) UCs are shifted along Si–Si bond direction by a distance of one Si–Si bond length, resulting from the change in buckling configuration after hydrogenation. Moreover, by annealing the sample to a moderate temperature, about 450 K, dehydrogenation occurs and a clean silicene surface is recovered (Fig. 7d). The work provides a clear and fundamental picture for silicene hydrogenation. And such a uniformly ordered, reversible hydrogenation is useful to modify the electronic properties of silicene for potential applications [37].
123
Sci. Bull.
5.2 Oxidization Oxidation of silicene–silicene oxide is expected to be one of the major steps toward effective engineering silicene’s band structure, and toward an excellent introduction of oxygenated functional groups into Si network. Recent works have revealed the mechanisms of oxygen interaction with silicene sheets during oxidation process. It is found that different phases of silicene layers have significant impact on the oxidation process. In oxidation process, the oxygen adatom is identified on bridge site resulting in the configuration of double-atombonding overbridging O atoms (Od). These oxygen adatoms overbridge neighboring Si atoms leading to Si–O–Si bonds d in silicene oxide. Figure 8a–c suggests pffiffiffiffiffi that pffiffiffiffiffiO is major configuration pffiffiffi pffiffiffi in partially oxidized 13 13; 4 9 4 and 2 3 9 2 3 silicene layers (with respect to 1 9 1 Ag(111)) [34]. Although Od presents in all silicene structures, heights of oxygen adatoms residing on silicene layers are different as demonstrated by STM images in Fig. 8d. Oxygen adatoms
pffiffiffiffiffi pffiffiffiffiffi pffiffiffi pffiffiffi on 13 13 and 2 3 9 2 3 silicene layers appear to be ˚. higher than these adsorbed on 4 9 4 silicene by about 1 A The distances between nearest-neighboring TL Si atoms are pffiffiffiffiffi pffiffiffiffiffi pffiffiffi pffiffiffi ˚ for 13 13 and 2 3 9 2 3 silicene, 5.46 and 3.67 A respectively. These distances are longer than twice of typical ˚ ). Si–O bond lengths in bulk SiO2 (varies from 1.58 to 1.62 A Therefore, both top-layer (TL) and bottom-layer (BL) Si atoms involve in silicon–oxygen bonds as Si(BL)–Od– Si(TL). Oxygen adatoms prefer to reside beside TL Si rather than BL Si as shown in STM results. By contrast, distance between nearest-neighboring TL Si in 4 9 4 silicene is ˚ , indicating different buckling to the other two 2.51 A superstructures. DFT calculation indicates that TL Si atoms in 4 9 4 silicene can decrease height forming BL Si atoms by oxidation, in order to minimize total energy. Therefore, Si(BL)–Od–Si(BL) is also a possible configuration for overbridging oxygen adatoms. As a result, the height of oxygen adatoms on 4 9 4 silicene is smallest in three buckled superstructures. It should be noted that Od is an energetically favoured configuration for oxygen adatoms on
pffiffiffiffiffi pffiffiffiffiffi pffiffiffi pffiffiffi Fig. 8 (Color online) STM and STS results of oxidized silicene in 13 13 (a), 4 9 4 (b) and 2 3 9 2 3 (c) structures. The oxygen adatoms prefer to reside on TL Si atoms at initial oxidation; d line profiles of oxygen adatoms on silicene corresponding to the lines in STM images in a–c, respectively; e–g DFT simulation (top and offfiffiffiffiffi atomic structures for an oxygen adatoms on Ag(111) supported silicene pffiffiffi side views) pffiffiffi p pffiffiffiffiffi monolayers in different superstructures. e 4 9 4, f 2 3 9 2 3, g 13 13: Reprinted with permission from Ref. [34]. Copyright (2014) American Chemical Society
123
Sci. Bull.
all the three silicene structures even with the oxygen dose up to 60 L [34]. Silicene monolayers grown on Ag(111) surfaces demonstrate a band gap that is tunable by oxygen adatoms from semimetallic to semiconducting type. With the use of lowtemperature scanning tunneling microscopy, the work demonstrates the feasibility of tuning the band gap of silicene with oxygen adatoms, which, in turn, expands the base of available two-dimensional electronic materials for devices with properties that are hardly achieved with graphene oxide [34]. ARPES studies revealed the detailed modulation of oxidation on silicene’s electronic structure. The Shockley surface state (SSS) and the typical bulk sp-band of Ag(111) vanished upon the deposition of silicon atoms. The more detailed illustration for the silicene’s ARPES results could be found in Ref. [34]. In oxidized silicene, the signal of metallic hybridized surface band (HSB) is faint, combined with resurgence of SSS band, indicating that the hybridization between 4 9 4 silicene and substrate is broken after oxidization, and silicon atoms in 4 9 4 silicene have a high oxygen reactivity than that of Ag(111) substrate. Such a high oxygen reactivity could be treated as a protection layer for Ag(111) because that SSS states are revived after oxidization. 5.3 Environmental gas adsorption on silicene nanoribbon Because sensors are typically exposed to environmental gases such as N2, O2, CO2, and H2O, the effects of these gases on silicene nanoribbon’s conductance were studied by Osborn and Farajian [70]. Their calculations show that N2 interacts with silicene via physisorption with an energy of 0.42 eV/N2 and is most energetically stable on the edge of the nanoribbon. This N2 adsorption does not lead to significant deformation in the nanoribbon. O2 on the other hand interacts strongly with the pristine silicene nanoribbon with an adsorption energy of 2.96 eV/O2. Upon relaxation, the O2 molecule dissociates in favor of individual Si–O bonds. This large adsorption energy seems to indicate that pristine silicene nanoribbons would easily oxidize under ambient conditions. The effects of environmental gases on the conductance of the nanoribbon were calculated before and after the adsorption of O2 and N2 molecules [70]. The conduction curves depicted in the top panel of Fig. 9 confirm the inert behavior of N2, showing conductance nearly identical to that of the pristine nanoribbon [70]. For oxygen, we see that conductance is significantly reduced while the 0.09 eV band gap is preserved. These results indicate that although the sensing capability of silicene nanoribbons may diminish in an oxygen-containing atmosphere, the capability is preserved in a nitrogen-containing
Fig. 9 (Color online) Quantum conductance modulation resulting from environmental gas molecules adsorptions on nanoribbon. Reprinted with permission from Ref. [70]. Copyright (2014) Springer
atmosphere. The effects of CO2 and H2O adsorption are also investigated. For H2O adsorption, the minimum energy configuration results from water splitting [71] and subsequent attachment of H and OH at the edge (with a binding energy (BE) of 1.62 eV) while CO2 adsorbs via physisorption (BE *0.46 eV). Similar to the case of oxygen adsorption, owing to the destructive effect on the nanoribbon structure, water molecules should also be removed from the environment for proper sensor functionality. Overall, these results indicate that long silicene nanoribbons could provide a unique nano-sensor capable of single-molecule resolution.
6 Argument on existence of
pffiffiffi pffiffiffi 3 3 silicene
Metal/semiconductor structures are of interest in several fields, especially in electronics. Ag–Si system has been heavily studied for several decades for its interesting atomic arrangement, surface electronic states, and the related physic properties. pffiffiffi pffiffiffi Prior to the intense interest in silicene, Ag 3 3 R30° reconstruction with a honeycomb structure on Si(111) has been widely investigated. The honeycomb-chained triangle (HCT) model and inequivalent triangle (IET) model were used to describe the atomic structure of the
123
Sci. Bull.
Fig. 10 (Color online) STM image at RT with a Vbias = -0.6 V, I = 1.0 nA and b Vbias = -0.3 V and I = 0.3 nA; c HCT model; d IET model. Reprinted with permission from Ref. [72]. Copyright (2006) by the American Physical Society
pffiffiffi pffiffiffi Si(111)-( 3 3)-Ag surface, as shown in Fig. 10a, b [72]. At the beginning, it was believed that HCT structure observed at higher temperature (higher than 67 K) is an averaged structural result between IET? and IET- structures at low temperature by thermal Zhang pffiffiffi fluctuations. pffiffiffi et al. [72] found that Si(111)-( 3 3)-Ag with IET structure can be observed at both RT and LT by STM. It suggests that the HCT–IET phase transition does not exist, which may be caused by the conditions of the tip [73]. It is noticeable pffiffiffi that pffiffiffirecent work by Mannix et al. [40] claimed that 3 3 silicene is structurally and electronically identical to HCT on Si(111) at the atomicscale through characterization. p ffiffiffi pffiffiSTM ffi pffiffiffi As pffiffiffishown in Fig. 11, both 3 3 silicene and 3 3 Ag are honeycomb structures with a similar constant lattice of 0.65 nm and a height of monolayer of 0.31 though this p work pffiffiffi nm. pffiffiEven ffi ffiffiffi cannot demonstrate that 3 3 silicene is the 3 pffiffiffi 3 Ag reconstruction on the surface of Si(111), debates about the real atomic structure of silicene have been raised. Further pffiffiffi works pffiffiffi about atomic structure of silicene, especially 3 9 3 phase, are urgently needed to explain those novel physic and chemical properties observed on silicene.
pffiffiffi pffiffiffi Fig. 11 (Color online) a Growth schematic for Si on Ag(111) and Ag on Si(111). b STM topography of the 3 3 phase for Si on Ag(111) (scale bar pffiffiffi p ffiffiffi = 50 nm, Vbias = -1.0 V, I = 400 pA); c line profile revealing a 0.31 nm step height. d Atomic-scale STM topography of the 3 pffiffiffi p3ffiffiffi phase for Si on Ag(111); e line profile (scale bar = 1 nm, Vbias = 0.3 V, I = 1.0 nA). f Atomic-scale STM topography of the lateral atomic 3 3 phase for Ag on Si(111); g line profile (scale bar = 1 nm, Vbias = -1.0 V, I = 100 pA) revealing an indistinguishable pffiffiffi pffiffiffi periodicity compared to the Vbias = -1.0 V phase for Si on Ag(111) shown in part e. h STM topography of the 3 3 phase for pffiffiffiAg pon ffiffiffi Si(111); i line profile (scale bar = 50 nm, Vbias = 0.85 V, I = 1.4 nA), showing identical step heights and similar island shapes to the 3 3 phase for Si on Ag(111) shown in part a. Reprinted with permission from Ref. [40]. Copyright (2014) American Chemical Society
123
Sci. Bull.
7 Summary The research on silicene has promoted a spectacular step since epitaxial silicene layers were first fabricated in 2012. This novel material has evoked intensive interest of scientists in the wide regions. Although this is a fairly new 2D material, recent outcomes on insight into silicene demonstrated encouraging features shedding a light on the electronics industry and giving a promising route toward applications of electronic device. Acknowledgments This work was supported by the Australian Research Council (ARC) through Discovery Project (DP 140102581), LIEF Grants (LE100100081 and LE110100099). Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Novoselov KS, Geim AK, Morozov SV et al (2004) Electric field effect in atomically thin carbon films. Science 306:666–669 2. Feng BJ, Ding ZJ, Meng S et al (2012) Evidence of silicene in honeycomb structure of silicon on Ag(111). Nano Lett 12:3507–3511 3. Meng L, Wang YL, Zhang LZ et al (2013) Buckled silicene formation on Ir(111). Nano Lett 13:685–690 4. Fleurence A, Friedlein R, Ozaki T et al (2012) Experimental evidence for epitaxial silicene on diboride thin films. Phys Rev Lett 108:245501 5. Gao JF, Zhao JJ (2012) Initial geometries, interaction mechanism and high stability of silicene on Ag(111) surface. Sci Rep 2:861 6. Kara A, Enriquez H, Seitsonen AP et al (2012) A review on silicene—new candidate for electrons. Surf Sci Rep 67:1–18 7. Watanabe K, Takashi T, Kanda H (2004) Direct-bandgap properties and evidence for ultraviolet lasing of hexagonal boron nitride single crystal. Nat Mater 3:404–409 8. Seifert G, Terrones H, Terrones M et al (2000) Structure and electronic properties of MoS2 nanotubes. Phys Rev Lett 85:146–149 9. Braga D, Lezama IG, Berger H et al (2012) Quantitative determination of the band gap of WS2 with ambipolar ionic liquidgated transistors. Nano Lett 12:5218–5223 10. Lalmi B, Oughaddou H, Enriquez H et al (2010) Epitaxial growth of a silicene sheet. Appl Phys Lett 97:223109 11. O’Hare A, Kusmartsev FV, Kugel KI (2012) A stable ‘‘flat’’ form of two-dimensional crystals: could graphene, silicene, germanene be minigap semiconductors? Nano Lett 12:1045–1052 12. Resta A, Leoni T, Barth C et al (2013) Atomic structures of silicene layers grown on Ag(111): scanning tunnelling microscopy and noncontact atomic force microscopy observations. Sci Rep 3:2399 13. Vogt P, De Padova P, Quaresima C et al (2012) Silicene: compelling experimental evidence for graphenelike two-dimensional silicon. Phys Rev Lett 108:155501 14. Yan JA, Stein R, Schaefer DM et al (2013) Electron–phonon coupling in two-dimensional silicene and germanene. Phys Rev B 88:121403 15. Chen L, Li H, Feng BJ et al (2013) Spontaneous symmetry breaking and dynamic phase transition in monolayer silicene. Phys Rev Lett 110:085504
16. Liu HS, Gao JF, Zhao JJ (2013) Silicene on substrates: a way to preserve or tune its electronic properties. J Phys Chem C117:10353–10359 17. Gao JF, Zhang JF, Liu HS (2013) Structures, mobilities, electronic and magnetic properties of point defects in silicene. Nanoscale 5:9785–9792 18. Feng BJ, Li WB, Qiu JL et al (2015) Variable coupling strength of silicene on Ag(111). Chin Phys Lett 32:037302 19. Feng Y, Feng BJ, Xie ZJ et al (2014) Observation of a flat band in silicene. Chin Phys Lett 31:127303 20. Liu HS, Han NN, Zhao JJ (2014) Band gap opening in bilayer silicene by alkali metal interaction. J Phys Condens Matter 26:475303 21. Chen L, Liu CC, Feng BJ et al (2012) Evidence for Dirac fermions in a honeycomb lattice based on silicon. Phys Rev Lett 109:056804 22. Liu CC, Feng WX, Yao YG (2011) Quantum spin hall effect in silicene and two-dimensional germanium. Phys Rev Lett 107:076802 23. Quhe R, Fei RX, Liu QH et al (2012) Tunable and sizable band gap in silicene by surface adsorption. Sci Rep 2:853 24. Ezawa M (2012) Valley-polarized metals and quantum anomalous hall effect in silicene. Phys Rev Lett 109:055502 25. Ezawa M (2013) Photoinduced topological phase transition and a single Dirac-cone state in silicene. Phys Rev Lett 110:026603 26. Chiappe D, Grazianetti C, Tallarida G et al (2012) Local electronic properties of corrugated silicene phases. Adv Mater 24:5088–5093 27. Enriquez H, Vizzini S, Kara A et al (2012) Silicene structures on silver surfaces. J Phys Condens Matter 24:314211 28. Jamgotchian H, Colignon Y, Hamzaoui N et al (2012) Growth of silicene layers on Ag(111): unexpected effect of the substrate temperature. J Phys Condens Matter 24:172001 29. Ferrari AC, Basko DM (2013) Raman spectroscopy as a versatile tool for studying the properties of graphene. Nat Nanotechnol 8:235–246 30. Liu Z, Suenaga K, Harris PJF et al (2009) Open and closed edges of graphene layers. Phys Rev Lett 102:015501 31. Li WX, Stampfl C, Scheffler M (2002) Oxygen adsorption on Ag(111): a density-functional theory investigation. Phys Rev B 65:075407 32. Takeda K, Shiraishi K (1994) Theoretical possibility of stage corrugation in Si and Ge analogs of graphite. Phys Rev B 50:14916–14922 33. Lin CL, Arafune R, Kawahara K (2012) Structure of silicene grown on Ag(111). Appl Phys Express 5:45802 34. Du Y, Zhuang JC, Liu HS et al (2014) Tuning the band gap in silicene by oxidation. ACS Nano 8:10019–10025 35. Aufray B, Kara A, Vizzini S et al (2010) Graphene-like silicon nanoribbons on Ag(110): a possible formation of silicene. Appl Phys Lett 96:183102 36. Tchalala MR, Enriquez H, Mayne AJ et al (2013) Formation of one-dimensional self-assembled silicon nanoribbons on Au(110)(2 9 1). Appl Phys Lett 102:083107 37. Qiu JL, Fu HX, Xu Y et al (2015) Ordered and reversible hydrogenation of silicene. Phys Rev Lett 114:126101 38. Arafune R, Lin CL, Kawahara K et al (2013) Structural transition of silicene on Ag(111). Surf Sci 608:297–300 39. Shirai T, Shirasawa T, Hirahara T et al (2014) Structure determination of multilayer silicene grown on Ag(111) films by electron diffraction: evidence for Ag segregation at the surface. Phys Rev B 89:241403 40. Mannix AJ, Kiraly B, Fisher BL et al (2014) Silicon growth at the two-dimensional limit on Ag(111). ACS Nano 8:7538–7547
123
Sci. Bull. 41. Feng JG, Wagner SR, Zhang PP (2015) Interfacial coupling and electronic structure of two-dimensional silicon growth on the Ag(111) surface at high temperature. Sci Rep 5:10310 42. Liu ZL, Wang MX, Xu JP et al (2014) Various atomic structures of monolayer silicene fabricated on Ag(111). New J Phys 16:075006 43. Kopnin NB, Sonin EB (2008) BSC superconductivity of Dirac electrons in graphene layers. Phys Rev Lett 100:246808 44. Scalise E, Houssa M, Pourtois G et al (2013) Vibrational properties of silicene and germanene. Nano Res 6:19–28 45. De Padova P, Ottaviani C, Quaresima C et al (2014) 24 h stability of thick multilayer silicene in air. 2D Mater 1:021003 46. Molle A, Grazianetti C, Chiappe D et al (2013) Hindering the oxidation of silicene with non-reactive encapsulation. Adv Funct Mater 23:4340–4344 47. Cinquata E, Scalise E, Chiappe D et al (2013) Getting through the nature of silicene: an sp2–sp3 two-dimensional silicon nanosheet. J Phys Chem C 117:16719–16724 48. Zhuang JC, Xu X, Du Y et al (2015) Investigation of electron– phonon coupling in epitaxial silicene by in situ Raman spectroscopy. Phys Rev B 91:161409 49. Ferrari AC, Meyer JC, Scardaci V et al (2006) Raman spectrum of graphene and graphene layers. Phys Rev Lett 97:187401 50. McMillan W (1968) Transition temperature of strong-coupled superconductors. Phys Rev 167:331 51. Si C, Liu Z, Duan WH et al (2013) First-principles calculations on the effect of doping and biaxial tensile strain on electron– phonon coupling in graphene. Phys Rev Lett 111:196802 52. Chen L, Feng BJ, Wu KH (2013) Observation of a possible superconducting gap in silicene on Ag(111) surface. Appl Phys Lett 102:081602 53. Tao L, Cinquata E, Chippe D et al (2015) Silicene field-effect transistors operating at room temperature. Nat Nanotechnol 10:227–231 54. Le Lay G (2015) 2D materials: silicene transistors. Nat Nanotechnol 10:202–203 55. Guzma´n-Verri GG, Lew Yan Voon LC (2007) Electronic structure of silicon-based nanostructures. Phys Rev B 76:075131 56. Cahangirov S, Topsakal M, Aktu¨rk E et al (2009) Two- and onedimensional hneycomb structures of silicon and germanium. Phys Rev Lett 102:236804 57. Lebe`gue S, Eriksson O (2009) Electronic structure of two-dimensional crystals from ab initio theory. Phys Rev B 79:115409
123
58. Lin CL, Arafune R, Kawahara K et al (2013) Substrate-induced symmetry breaking in silicene. Phys Rev Lett 110:076801 59. Tsoutsou D, Xenogiannopoulou E, Golias E et al (2013) Evidence for hybrid surface metallic band in (4 9 4) silicene on Ag(111). Appl Phys Lett 103:231604 60. Xu X, Zhuang JC, Du Y et al (2014) Effects of oxygen adsorption on the surface state of epitaxial silicene on Ag(111). Sci Rep 4:7543 61. Avila J, De Padova P, Cho S et al (2013) Presence of gapped silicene-derived band in the prototypical (3 9 3) silicene phase on silver (111) substrates. J Phys Condens Matter 25:262001 62. De Padova P, Vogt P, Resta A et al (2013) Evidence of Dirac fermions in multilayer silicene. Appl Phys Lett 102:163106 63. De Padova P, Avila J, Resta A et al (2013) The quasiparticle band dispersion in epitaxial multilayer silicene. J Phys Condens Matter 25:382202 64. Rutter GM, Crain JN, Guisinger NP et al (2007) Scattering and interference in epitaxial graphene. Science 317:219–222 65. Feng BJ, Li H, Liu CC et al (2013) Observation of Dirac cone warping and chirality effects in silicene. ACS Nano 7:9049–9054 66. Sofo JO, Chaudhari AS, Barber GD (2007) Graphane: a twodimensional hydrocarbon. Phys Rev B 75:153401 67. Elias DC, Nair RR, Mohiuddin TMG et al (2009) Control of graphene’s properties by reversible hydrogenation: evidence for graphene. Science 323:610–613 68. Balog R, Jørgensen B, Nilsson L et al (2010) Bandgap opening in graphene induced by patterned hydrogen adsorption. Nat Mater 9:315–319 69. Andreev T, Barke I, Ho¨vel H (2004) Adsorbed rare-gas layers on Au(111): shift of the Shockley surface state studied with ultraviolet photoelectron spectroscopy and scanning tunnelling spectroscopy. Phys Rev B 70:205426 70. Osborn TH, Farajian AA (2014) Silicene nanoribbons as carbon monoxide nanosensors with molecular resolution. Nano Res 7:945–952 71. Konecˇny´ R, Doren DJ (1997) Adsorption of water on Si(100)(2 9 1)—a study with density-functional theory. J Chem Phys 106:2426–2435 72. Zhang HM, Gustafsson JB, LSO (2006) Surface atomic pffiffiJohansson ffi pffiffiffi structure of Ag/Si(111)-( 3 3). Phys Rev B 74:201304 pffiffiffi pffiffiffi 73. Sato N, Nagao T, Hasegawa S (1999) Si(111)-( 3 3)-Ag surface at low temperatures: symmetry breaking and surface twin boundaries. Surf Sci 442:65–73