記住我
Here, we characterize the C–Si bonding buffer layer and free-standing epitaxial graphene on 4H-SiC by low-energy electron-emission spectroscopy (LEED), investigate the thermal and bias-dependent in situ Raman shift of epitaxial graphene, and explore bias-tuning of the Schottky barrier between graphene and 4H-SiC. Then, we fabricate high-detectivity graphene–SiC–graphene (GSG) UV photodetectors (PDs) from 200 to 400 nm. Furthermore, the ladder shape energy levels and electronic trapping and de-trapping effects at the SG interface were revealed and analyzed that IQS may come from the random distribution of C–Si bonding, exposed dangling bonds, and silicon vacancy defects on the SiC surface.
A 4-in. semi-insulated 4H-SiC (0001) (5 × 1015 cm−3) was heated to 1600–1800 °C at 2 × 10−3 Pa for 30 min in a graphite frame ultra-high vacuum furnace to the graphene layer. Raman characterization and LEED have been subsequently utilized to study the quality of the samples. The graphene interdigital electrodes were fabricated by SF6 gas reactive ion etching (RIE), and its Ti/Al (50 nm/100 nm) pads were prepared by electron beam evaporation sputtering. The optoelectronic characteristics of the device were investigated by a self-assembled photoelectric measurement system equipped with a Keithley 4200 SCS, a xenon light source, and a monochromator.
Figure 1(a) depicts the Raman spectrum of epitaxial graphene on 4H-SiC, along with asymmetry 2D-peak at 2742 cm−1. The asymmetric 2D-peak can be fitted by two Lorentz splitting, indicating two layers of graphene cover the buffer layer25–2725. W. Yang, F. Zhang, Z. Liu, and Z. Wu, “ Effects of annealing on the performance of 4H-SiC metal–semiconductor–metal ultraviolet photodetectors,” Mater. Sci. Semicond. Process. 11, 59–62 (2008). https://doi.org/10.1016/j.mssp.2008.11.00126. D. S. Lee, C. Riedl, B. Krauss, K. von Klitzing, U. Starke, and J. H. Smet, “ Raman spectra of epitaxial graphene on SiC and of epitaxial graphene transferred to SiO2,” Nano Lett. 8, 4320–4325 (2008). https://doi.org/10.1021/nl802156w27. A. C. Ferrari, J. C. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K. S. Novoselov, and S. Roth, “ Raman spectrum of graphene and graphene layers,” Phys. Rev. Lett. 97, 187401 (2006). https://doi.org/10.1103/PhysRevLett.97.187401 as well as the low ratio of D-peak means high-quality graphene. Compared to the exfoliated graphene, G-peaks shift from 1580 cm−1 to high frequencies 1593 cm−1 due to the residue Si -C bonds introduced strain.2828. T. Mohiuddin, A. Lombardo, R. Nair, A. Bonetti, G. Savini, R. Jalil, N. Bonini, D. Basko, C. Galiotis, and N. Marzari, “ Uniaxial strain in graphene by Raman spectroscopy: G peak splitting, Grüneisen parameters, and sample orientation,” Phys. Rev. B 79, 205433 (2009). https://doi.org/10.1103/PhysRevB.79.205433 There is a G*-peak at 2450 cm−1, which comes from the strain, attributing to the Si–C and Si dangling bonds.2929. D. Mafra, G. Samsonidze, L. Malard, D. C. Elias, J. Brant, F. Plentz, E. S. Alves, and M. A. Pimenta, “ Determination of LA and TO phonon dispersion relations of graphene near the Dirac point by double resonance Raman scattering,” Phys. Rev. B 76, 233407 (2007). https://doi.org/10.1103/PhysRevB.76.233407 Inset of Fig. 1(a) exhibits the LEED of graphene/SiC(0001) junction with the (63×63) R300 reconstructions of SiC surface, marking the lattice vectors of graphene and SiC with G1 and G2 and S1 and S2, respectively.30,3130. K. Emtsev, F. Speck, T. Seyller, L. Ley, and J. D. Riley, “ Interaction, growth, and ordering of epitaxial graphene on SiC surfaces: A comparative photoelectron spectroscopy study,” Phys. Rev. B 77, 155303 (2008). https://doi.org/10.1103/PhysRevB.77.15530331. S. Fiori, Y. Murata, S. Veronesi, A. Rossi, C. Coletti, and S. Heun, “ Li-intercalated graphene on SiC (0001): An STM study,” Phys. Rev. B 96, 125429 (2017). https://doi.org/10.1103/PhysRevB.96.125429 Figure 1(b) shows the cross section view of the graphene/SiC atomic stack. The buffer layer has the same lattice constant as the top layers and is partly chemically bonded to the SiC substrate. As well as the amount of Si dangling bonds also forms some IQS. Figure 1(c) depicts the morphology of the GSG PDs integrated with the graphene electrodes fabricated by etching 15 nm SiC mesa, as details shown in Fig. 1(d).To study the electronic properties of pristine graphene/4H-SiC(0001) heterojunctions, we turn to investigate the temperature-dependent I–V characteristics of the SG junction. Figure 2(a) illustrates the I–V characteristics of the SG heterojunction at temperatures ranging from 303 to 463 K. The rectification characteristics and symmetrically distributed I–V curve well confirm the uniform Schottky nature of the SG junction. The I–V characteristic suggests that the electronic transport of the SG junction is dominated by the thermionic emission, which is expressed by I=A0A*T2exp−∅BkT exp qVnkT−1,(1)where I denotes current density, A0 is the junction effective area (0.05 mm2), A* denotes effective Richardson's constant, T is the absolute temperature, V denotes reverse bias, n denotes ideal factor, ΔØB is SBH, q denotes electron charge (q = 1.6 × 1019 C), and k is Boltzmann's constant.The SBH of the junction can be obtained from the slope of the linear fit of the Richardson plot, as shown in the inset in Fig. 2(b), and the extracted SBH is 0.86 eV at bias 6 V, even higher than a p-doped graphene/n-type semiconductor Schottky junction (≈0.8 eV).3232. A. Wirth-Lima, P. Alves-Sousa, and W. Bezerra-Fraga, “ n-graphene/p-silicon-based Schottky junction solar cell, with very high power conversion efficiency,” SN Appl. Sci. 2, 905–913 (2020). https://doi.org/10.1007/s42452-020-2056-1 These extracted reverse bias-dependent SBHs appear much higher than in early works due to the IQS at the SG junction.24,3324. C. Sun, X. Chen, R. Hong, X. Li, X. Xu, X. Chen, J. Cai, X.-A. Zhang, W. Cai, and Z. Wu, “ Enhancing the photoelectrical performance of graphene/4H-SiC/graphene detector by tuning a Schottky barrier by bias,” Appl. Phys. Lett. 117, 071102 (2020). https://doi.org/10.1063/5.001256633. M. Dub, P. Sai, A. Przewoka, A. Krajewska, M. Sakowicz, P. Prystawko, J. Kacperski, I. Pasternak, G. Cywiński, and D. But, “ Graphene as a Schottky barrier contact to AlGaN/GaN heterostructures,” Materials 13, 4140 (2020). https://doi.org/10.3390/ma13184140 These IQS introduce a surface potential that can subsequently increase the injection barrier for 4H-SiC [the work function (φ) is 3.39 eV]3434. H.-K. Kim, S. I. Kim, S. Kim, N.-S. Lee, H.-K. Shin, and C. W. Lee, “ Relation between work function and structural properties of triangular defects in 4H-SiC epitaxial layer: Kelvin probe force microscopic and spectroscopic analyses,” Nanoscale 12, 8216–8229 (2020). https://doi.org/10.1039/C9NR10126H with low-work function graphene (φ=4.42 eV). These extracted reverse bias-dependent SBHs remain constant around 0.91 eV from 0 V and then decrease along with bias above 3.8 V. The reduction of the barrier appears in a ladder shape, which is mainly due to the diversity of IQS from dangling bonds and C and Si vacancies.3535. V. Afanasev, M. Bassler, G. Pensl, and M. Schulz, “ Intrinsic SiC/SiO2 interface states,” Phys. Status Solidi A 162, 321–337 (1997). https://doi.org/10.1002/1521-396X(199707)162:1<321::AID-PSSA321>3.0.CO;2-F The downward ladder shape SBHs mean that these trapped carriers transit at IQS and are released by the electric field, doping the graphene, leading to the SBH reduction. Then, the IQS charges again at a higher energy level with the bias increase, and another graphene doping process is repeated, resulting in a ladder SBH curve. This process repeats several times in a gradient decrease bias gap, corresponding to the multiple steps of the barrier.As depicted in Fig. 2(c), the random distribution of residual C–Si bonding between the buffer layer and SiC substrate breaks the integrity of the electron clouds of π-bonding among the sp2 hybridized carbon atoms in the buffer layer, leading to symmetry breaking and IQS. Additionally, the surface Si dangling bonds as well as C–Si bonds lead to a diverse energy level of IQS from deep energy level “I” to shallow energy level “V,” which mediates the barrier and traps the carriers.3636. S.-H. Tsai, S. Lei, X. Zhu, S.-P. Tsai, G. Yin, X. Che, P. Deng, J. Ng, X. Zhang, and W.-H. Lin, “ Interfacial states and Fano–Feshbach resonance in graphene–silicon vertical junction,” Nano Lett. 19, 6765–6771 (2019). https://doi.org/10.1021/acs.nanolett.9b01658 These states can subsequently trap carriers and hinder their transport, and a trap can capture and restrain carriers temporarily until it is released by an external stimulus such as an electric field, thermal energy, or photons.3737. H. F. Haneef, A. M. Zeidell, and O. D. Jurchescu, “ Charge carrier traps in organic semiconductors: A review on the underlying physics and impact on electronic devices,” J. Mater. Chem. C 8, 759–787 (2020). https://doi.org/10.1039/C9TC05695E As depicted by the red arrow in Fig. 2(e), at high temperatures, the G-peak shifts due to the greater stress on the lattice of graphene2828. T. Mohiuddin, A. Lombardo, R. Nair, A. Bonetti, G. Savini, R. Jalil, N. Bonini, D. Basko, C. Galiotis, and N. Marzari, “ Uniaxial strain in graphene by Raman spectroscopy: G peak splitting, Grüneisen parameters, and sample orientation,” Phys. Rev. B 79, 205433 (2009). https://doi.org/10.1103/PhysRevB.79.205433 and due to the difference in the thermal conductivity between SiC and graphene,4,384. A. A. Balandin, “ Thermal properties of graphene and nanostructured carbon materials,” Nat. Mater. 10, 569–581 (2011). https://doi.org/10.1038/nmat306438. B. H. Morkoc, S. Strite, G. Gao, M. Lin, B. Sverdlov, and M. Burns, “ Large-band-gap SiC, III-V nitride, and II-VI ZnSe-based semiconductor device technologies,” J. Appl. Phys. 76, 1363–1398 (1994). https://doi.org/10.1063/1.358463 attributing to the chemical bonds between graphene and SiC.The in situ Raman fingerprints of graphene with reverse bias are shown in Fig. 2(d) where the slight G-peaks shifts happened to correspond to the ramping of SBHs in bias 3–5, 8–9, and 12–13 V marked by pink arrows. These shifts indicate the charge doping of graphene and the upward shift of the Fermi level. The Fermi level shift relative to the Dirac point can be approximated as where n0 denotes the carrier concentration (positive for holes and negative for carriers), ℏ denotes the Planck constant, and νF denotes the Fermi velocity of graphene (νF = 1.1 × 106 cm/s). At zero bias voltage, the Fermi energy EF0 of the epitaxial graphene is calculated with Hall data n0 = 5 × 1012 cm–2 to be in the range of 0.27–0.28 eV above the Dirac point.Assume that the junction is an ideal Schottky junction where surface-state effects are neglected. For the epitaxial graphene, the expressions in Eq. (2) include extrinsic doping and the thermal equilibrium contact with the SiC. The shift of the Fermi energy level can be rewritten as EF=−ℏvFπn0−ϵsϵ0NDVbi+VR2e,(3)where VR denotes the reverse bias voltage. The charge transfers associated with the Schottky barrier read as n0−ninduced, ninduced=ϵsϵ0ND(Vbi+VR)/2e. As a typical semiconductor, we use parameter values of Vbi ∼ 0.6 V, εs ∼ 10, ε0 = 8.84 × 10−14 F/cm2, and the doping concentration ND (n-type, 2 × 106–2 × 107) 2 × 107. The displacement of the Fermi level of graphene with diverse bias in an ideal Schottky junction can be obtained from the following equation: ΔΦBVR=EFVR1−EFVR2,(4)where VR1 and VR2 denote different biases. The Fermi level shift responds to the increasing bias in the VR1 ∼ VR2 , and the region from region I to region V can be deduced from Eq. (4), 38, 20, 18, 16, and 9 meV, respectively. These theoretically calculated values are close to the experimental results, which prove that the IQS traps formed at the interface between the graphene and SiC.Figure 3(a) depicts the photocurrent and dark current of the GSG PDs. The dark current (−13 A) is 10−4 times our earlier work in Ref. 2424. C. Sun, X. Chen, R. Hong, X. Li, X. Xu, X. Chen, J. Cai, X.-A. Zhang, W. Cai, and Z. Wu, “ Enhancing the photoelectrical performance of graphene/4H-SiC/graphene detector by tuning a Schottky barrier by bias,” Appl. Phys. Lett. 117, 071102 (2020). https://doi.org/10.1063/5.0012566, characteristics with 103 on/off ratio at 315 nm and 7.3 × 10−8 W UV illumination. We consider this remarkable on/off ratio upgrades mainly due to deep energy level IQS trapping and related high SBHs that suppress dark current. In addition, the dark current corresponds to an exponential function, as the fitted curve in Fig. 3(a). Figure 3(b) shows the responsivity of the GSG PDs with the illumination ranges from 200 to 400 nm, while Fig. 3(c) depicts the corresponding detectivity with 103 of the UV–visible rejection ratios. The high Schottky barrier of the junction filters the transition of long wavelengths photogenerated carriers from graphene. The responsivity peaks appear around 260 nm (4.77 eV) from 2 to 15 V, corresponding to the graphene absorption spectrum.3939. Y.-L. Xu, H.-X. Li, C.-B. Zhou, X.-S. Xiao, Z.-C. Bai, Z.-P. Zhang, and S.-J. Qin, “ The ultraviolet absorption of graphene in the Tamm state,” Optik 219, 165015 (2020). https://doi.org/10.1016/j.ijleo.2020.165015 Notably, the peak responsivity and detectivity of the device shifted as reverse bias increased, which corresponds to the photoexcitation and interfacial state-mediated excitation. At reverse bias The penetration depth of the photon increases with the wavelength so that more photogenerated carriers will come from the response of SiC when the wavelength is longer than 275 nm. As shown in Fig. 3(b) and inset of Fig. 3(c), the responsivity and detectivity present peaks at 300 nm above 30 V. For the wavelengths shorter than 260 nm, photogenerated carriers can be easily recombined by various SiC surface defects such as silicon vacancy or intrinsic defects, etc. Meantime, the majority of photogenerated carriers appear at a deeper position in SiC for longer wavelengths (>300 nm) drifting longer distances to graphene electrodes. Due to the recombination of the surface silicon vacancies and body intrinsic defects, the carriers generated by photons (4.14 eV and 300 nm) can just overcome the Schottky barrier and these defects by a high electrical field. That is the reason why the dominant photocurrent comes from SiC, and the responsivity and detectivity show peaks at 300 nm at 30 V.Figure 4 depicts the band energy diagram of SiC and graphene, where ΦB (V) denotes the SBH and Vb is the deviation of the Fermi level between the graphene and 4H-SiC. Without illumination, IQS impeded the carriers doping in graphene, and SBHs remained constant below 3.8 V, as shown in Fig. 4(a). When the bias exceeds 3.8 V, the carriers were stimulated by the electric field, and the graphene was doped simultaneously. Meantime, the Fermi level was raised, and the SBHs were reduced. The increasing electrical field stimulus releases the trapped carriers, which induce the ladder shape of the SBH curve with the increase in reverse bias, as shown in Fig. 2(b). Carriers trapped in deep energy levels can transit to higher energy levels by applied bias and thermal emission, which further drives up Fermi levels and decreases the barriers.Under UV illumination, as shown in Fig. 4(c), the majority of carriers come from the high-energy transition in graphene, especially around the incident wavelength of 260 nm (4.77 eV). The energy of these photons has adequate energy to drive electron transition from deep energy levels at the interface although the bias is less than 15 V. However, the dominant photoelectric response comes from the graphene, so the responsivity is not high due to the surface recombination for the shorter wavelength (15 V), the electric field drives up the Fermi level, so that the photons from longer wavelength (>260 nm) can stimulate trapped carriers to overwhelm SBH from IQS, as shown in Fig. 4(d). Meanwhile, the longer-wavelength photons have deeper penetration depth so that the dominant photoelectric response comes from SiC. Photon-generated carriers from SiC gradually overwhelm the SBH from graphene layers, and the peak responsivity shifts to a longer wavelength [300 nm, as shown in Fig. 3(d)]. Therefore, the responsivity of the GSG UV PDs can be tuned through the deep energy levels by bias.The detectivity of the device was calculated to be 1.34 × 1012 Jones by the D*=RA/(2qIdark), where A denotes the optically active area of the PD. This detector has a much lower dark current and higher detectivity than previous work, as shown in Table I.
TABLE I. Comparison of GSG with other PDs. SBP: Schottky barrier photodiode.
Materials/substrateDark current (A)Responsivity (A/W)Detectivity (Jones)ReferenceEpitaxial Graphene (EG)/SiC/EG10−130.011.3 × 1012This workEG/SiC/EG10−9409.0 × 10112424. C. Sun, X. Chen, R. Hong, X. Li, X. Xu, X. Chen, J. Cai, X.-A. Zhang, W. Cai, and Z. Wu, “ Enhancing the photoelectrical performance of graphene/4H-SiC/graphene detector by tuning a Schottky barrier by bias,” Appl. Phys. Lett. 117, 071102 (2020). https://doi.org/10.1063/5.0012566EG/SiC/EG10−100.0092.6 × 10102020. E. Kus, D. Özkendir, V. Fırat, and C. Çelebi, “ Epitaxial graphene contact electrode for silicon carbide based ultraviolet photodetector,” J. Phys. D: Appl. Phys. 48, 095104 (2015). https://doi.org/10.1088/0022-3727/48/9/095104EG/SiC, SBP10−32.184.2 × 1091919. J. Yang, L. Guo, Y. Guo, W. Hu, and Z. Zhang, “ Epitaxial graphene/SiC Schottky ultraviolet photodiode with orders of magnitude adjustability in responsivity and response speed,” Appl. Phys. Lett. 112, 103501 (2018). https://doi.org/10.1063/1.5019435EG/GaN, PIN (P-type semiconductor/Intrinsic layer/N-type semiconductor diode photodetector)10−103601.2 × 10121717. K. Xu, C. Xu, Y. Xie, J. Deng, Y. Zhu, W. Guo, M. Xun, K. B. Teo, H. Chen, and J. Sun, “ Graphene GaN-based Schottky ultraviolet detectors,” IEEE Trans. Electron Devices 62, 2802–2808 (2015). https://doi.org/10.1109/TED.2015.2453399In summary, high-quality graphene is prepared with n-type semi-insulating SiC by thermal epitaxial. Responsivity-tunable UV photodetectors are fabricated and demonstrated by IQS at the interface between 4H-SiC/graphene. The 4H-SiC/graphene Schottky junction shows a 0.91 eV Schottky barrier due to the carriers trapping in the IQS. The SBHs show a ladder-shaped energy trend with the increase in the reverse bias. The responsivity peak of the device shifts from 260 to 300 nm with the increase in the bias due to the IQS assist under different energy levels. The GSG UV PDs show ultra-low dark current with high SBHs and a high detectivity of 1.34 × 1012 Jones. This work achieves responsivity-tunable UV photodetectors, according to bias modulation of the energy levels at the graphene/SiC interface.
This research was supported by the National Natural Science Foundation of China (No. 62274137), the Natural Science Foundation of Fujian Province of China for Distinguished Young Scholars (No. 2020J06002), the Natural Science Foundation of Jiangxi Province of China for Distinguished Young Scholars (No. S2021QNZD2L0013), the Science and Technology Project of Fujian Province of China (No. 2020I0001), the Science and Technology Key Projects of Xiamen (No. 3502ZCQ20191001), the Laboratory Open Fund of Beijing Smart-chip Microelectronics Technology Co., Ltd. (No. SGSC0000KJQT2207192), the Fundamental Research Funds for the Central Universities (No. 20720220026), the Shenzhen Science and Technology Program (No. JSGG20201102155800003), and the Jiangxi Provincial Natural Science Foundation (No. 20212ACB212005).
Conflict of Interest
The authors have no conflicts to disclose.
Author Contributions
Baihong Zhu and Cunzhi Sun contributed equally to this work.
Baihong Zhu: Writing – original draft (equal). Xiaping Chen: Resources (equal); Supervision (equal). Jiafa Cai: Methodology (equal); Resources (equal); Supervision (equal). Songyan Chen: Formal analysis (equal); Methodology (equal). Zheng Yun Wu: Methodology (equal); Software (equal). Deyi Fu: Methodology (equal); Resources (equal); Supervision (equal). Shaolong He: Methodology (equal); Resources (equal); Supervision (equal). Weiwei Cai: Methodology (equal); Resources (equal); Software (equal). Feng Zhang: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Software (equal); Supervision (equal); Writing – review & editing (equal). Cunzhi Sun: Investigation (equal); Methodology (equal); Supervision (equal); Validation (equal); Writing – review & editing (equal). Jiadong Chen: Formal analysis (equal). Zihao Li: Investigation (equal); Methodology (equal). Shiming Huang: Resources (equal). Shaoxiong Wu: Resources (equal); Software (equal); Supervision (equal). Dingqu Lin: Resources (equal); Supervision (equal). Yu Lin: Investigation (equal); Methodology (equal); Resources (equal). Rongdun Hong: Methodology (equal); Resources (equal); Supervision (equal).
The data that support the findings of this study are available from the corresponding authors upon reasonable request.
REFERENCES
1. N. M. Peres, “ Colloquium: The transport properties of graphene: An introduction,” Rev. Mod. Phys. 82, 2673 (2010). https://doi.org/10.1103/RevModPhys.82.2673, Google ScholarCrossref, ISI2. K. F. Mak, L. Ju, F. Wang, and T. F. Heinz, “ Optical spectroscopy of graphene: From the far infrared to the ultraviolet,” Solid State Commun. 152, 1341–1349 (2012). https://doi.org/10.1016/j.ssc.2012.04.064, Google ScholarCrossref3. K. I. Bolotin, K. J. Sikes, Z. Jiang, M. Klima, G. Fudenberg, J. Hone, P. Kim, and H. L. Stormer, “ Ultrahigh electron mobility in suspended graphene,” Solid State Commun. 146, 351–355 (2008). https://doi.org/10.1016/j.ssc.2008.02.024, Google ScholarCrossref, ISI4. A. A. Balandin, “ Thermal properties of graphene and nanostructured carbon materials,” Nat. Mater. 10, 569–581 (2011). https://doi.org/10.1038/nmat3064, Google ScholarCrossref, ISI5. Z. Ni, W. Chen, X. Fan, J. Kuo, T. Yu, A. Wee, and Z. Shen, “ Raman spectroscopy of epitaxial graphene on a SiC substrate,” Phys. Rev. B 77, 115416 (2008). https://doi.org/10.1103/PhysRevB.77.115416, Google ScholarCrossref6. S. Tang, W. Wu, X. Xie, X. Li, and J. Gu, “ Band gap opening of bilayer graphen
留言 (0)