Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

High‐Performance Graphene‐Dielectric Interface by UV‐Assisted Atomic Layer Deposition for Graphene Field Effect Transistor

High‐Performance Graphene‐Dielectric Interface by UV‐Assisted Atomic Layer Deposition for... IntroductionGraphene has excellent electrical and mechanical properties, such as high carrier mobility, surface‐to‐volume ratio, and flexibility.[1–3] These interesting properties of graphene make it a promising candidate material for flexible electronics and sensors.[4,5] In particular, sensors based on graphene field‐effect transistors (GFETs) exhibit high sensitivity owing to their high transconductance.[6] In the fabrication of high‐sensitivity GFET devices, the interface property between graphene and the dielectric film is one of the design factors that highly influence device performance.[7] Therefore, to manufacture such high‐sensitivity GFET‐based sensors, it is essential to deposit a thin and uniform high‐k dielectric film on the graphene surface without the degradation of graphene's superior electrical properties.Atomic layer deposition (ALD) has been reported to be an adequate method for depositing highly uniform thin films with precise thickness control on various substrates, including graphene. As a chemical vapor deposition (CVD) variant, ALD has been shown to fabricate intimate thin film‐graphene interfaces.[8,9] However, the film formation is hindered by the chemically inert nature of graphene's basal plane due to its strong sp2 carbon bonding.[10] Therefore, various functionalization techniques for graphene surfaces, such as wet chemical treatment,[11] thermal oxidation treatment,[12] and seed layering,[13] have been studied. Although these techniques functionalize the graphene surface, they may cause degradation of graphene properties owing to chemical contamination and problems such as the formation of unnecessary seed layers. Surface functionalization using plasma is another graphene surface functionalization technique that is being actively researched.[14] However, graphene may be damaged or etched out because of high‐energy ions. In our previous study, we applied an atmospheric pressure plasma treatment technique to functionalize the graphene surface without degrading its properties with low plasma energy while depositing a high‐quality ALD dielectric film on the graphene surface.[15] However, these pre‐treatment techniques necessitate additional treatment processes and graphene may be exposed to air after the pre‐treatment process for the deposition of the dielectric film. As graphene is exposed to air during manufacturing, it can be p‐doped by environmental sources such as H2O and O2 in the air.[16]UV‐assisted ALD (UV‐ALD), in which the film is exposed to UV light at every cycle during the ALD process, is an effective ALD technique that can possibly solve these problems. UV light with a relatively low energy range (<10 eV) can functionalize the graphene surface without damaging or etching. A high‐quality dielectric film with fewer contaminants can be deposited as UV energy is transferred during the deposition process, facilitating the ligand exchange reaction. Yoon et al. reported that dense stoichiometric Al2O3 thin films could be deposited via UV‐ALD on Si substrate.[17] Consequently, a thin film can be deposited by applying UV‐ALD on the graphene surface without degrading its superior electrical properties, thereby realizing a high‐quality dielectric‐graphene interface for high‐performance GFETs. Nevertheless, to the best of our knowledge, the application of UV‐ALD to thin‐film deposition on 2D materials, including graphene surfaces or graphene‐based devices, has never been demonstrated.In this study, we demonstrate the first successful application of UV‐ALD to effectively functionalize graphene surfaces and, simultaneously, fabricate high‐performance GFETs with UV‐ALD Al2O3 dielectric thin films. High‐resolution field emission scanning electron microscopy (HR‐SEM), atomic force microscopy (AFM), X‐ray diffraction (XRD), and Raman spectroscopy results showed the optimal per‐cycle UV irradiation (5 s per cycle) in the ALD process promote the density of the Al2O3 dielectric film by facilitating the ligand exchange reaction, while simultaneously improving the graphene‐dielectric interface quality. As a result, GFETs with an Al2O3 dielectric layer deposited by UV‐ALD showed a low Dirac voltage close to 0 V (i.e., 8 V) and high hole mobility of 1221 cm2 V−1 s−1 that is a 207% improvement compared to those by ALD without UV.Experimental SectionUV‐ALD ProcessA customized ALD system equipped with a UV lamp (PE300BUV, Cermax) on the chamber's lid was used, as shown in Figure 1a. The distance between the UV lamp and substrate was maintained at 6 cm with a CaF2 window to pass UV light. ALD Al2O3 film was deposited on a Si wafer (reference) and graphene using trimethylaluminum (TMA) as the precursor and deionized water as the oxygen source. Argon gas was used as the carrier gas at a flow rate of 40 sccm. TMA and H2O were self‐evaporated at room temperature and the reaction chamber was maintained at <100 °C. Each ALD cycle consisted of 1) 0.1 s TMA pulse, 2) 60 s purge, 3) 0.1 s H2O pulse, and 4) 60 s purge with UV irradiation for 0, 5, 20, and 50 s at the H2O pulse step (Figure 1b, after which the samples were named (UV0/5/20/50). The Al2O3 film was deposited with the ALD cycle number of 150 cycles on Si and graphene substrates. The UV lamp power was fixed at 230 W, corresponding to an intensity of 0.12 W cm−2, and the UV wavelength and energy ranges were mostly in the range of 250–400 nm and 3–5 eV, respectively (>95% of relative output).1FigureSchematic of a) apparatus and b) process (1 cycle) of UV‐ALD used in this study.GFET Device FabricationThe GFETs used in this study were fabricated based on multilayer CVD graphene on 285 nm SiO2/ highly p‐doped Si (Graphene Supermarkets), which was followed by annealing in Ar (200 sccm) and H2 (50 sccm) atmospheres at 360 °C for 2 h. In this step, surface contaminants, such as polymer residues, were eliminated.[18,19] After annealing, a 20 nm Au electrode was deposited by thermal evaporation. Finally, an Al2O3 dielectric layer (150 cycles) was deposited by UV‐ALD on the annealed graphene.Structural and Electrical CharacterizationRaman spectroscopy of the graphene surface was carried out using a LabRam Aramis system (Horiba Jobin Yvon, Japan) equipped with a 514 nm laser. The laser was focused onto a 1 µm spot size using a 100× microscope objective. Compositional analysis was conducted by X‐ray photoelectron spectroscopy (XPS) (K‐Alpha+, Thermo Fisher Scientific) with a monochromated Al X‐ray source gun, a spot size of 200 µm, and a step size of 0.05 eV step−1. The Al2O3 film surface morphology was analyzed using high‐resolution field‐emission scanning electron microscopy (HR‐FESEM) (SU8010, Hitachi High Technologies Corporation) and atomic force microscopy (AFM) (XE‐100, Park Systems). The crystallinity, thickness, density, and roughness of the ALD Al2O3 films were characterized by grazing‐incidence X‐ray diffraction (GIXRD) and X‐ray reflectivity (XRR) (SmartLab, Rigaku Corporation). The electrical properties of the GFETs were measured in the air using a semiconductor analyzer (4200 SCS, Keithley). The back gate voltage (Vbg) was swept from −120 to 120 V, and the drain‐source current (Ids) was measured with the drain‐source voltage (Vds) of 0.1 V.Result and DiscussionFirst, we deposited and tested UV‐ALD Al2O3 on the Si substrate to observe the effect of UV irradiation solely on the film properties. The Al2O3 film became smoother under UV irradiation during the ALD process, as shown in Figure S1 (Supporting Information). Based on XRR analysis, the density of Al2O3 film increased from 2.9 to 3.2 and 3.3 g cm−3 with UV irradiation times of 0, 5, and 20 s (UV0/5/20), respectively (Figure S2, Supporting Information; Figure 2a) with similarly small roughness of < 0.1 nm in all samples. The density of UV0 on the Si sample (ALD Al2O3 without UV) was comparable to the previously reported value (2.8–2.9 g cm−3)[20] for thermal ALD Al2O3 deposited at < 100 °C, while that of UV5/20 on the Si samples increased by about 10%. In the XPS survey analysis, all Al2O3 films on Si with and without UV irradiation were stoichiometric (Al: 36.9, 39.3, and 39.3 at%; O: 61.5, 60, and 60 at%) (Figure 2b). The carbon content in the Al2O3 films decreased upon UV irradiation from 1.6 at% (UV0) to 0.7 at% (UV5/20) (Figure 2b). This is because UV irradiation eliminates unreacted species at low temperatures and recreates functional groups during the ALD process, as reported by Yoon et al.[17] Interestingly, the high‐resolution XPS analysis of the O1s peak (Figure 2c–e) and Al2p (Figure S3, Supporting Information) shows a clear contrast between ALD Al2O3 films without and with UV irradiation. The Al2O3 peaks (near 532.2 eV) are dominant in all samples in Figure 2c–f (83/94/94% in UV0/5/20); in the meantime, the AlOOH peak (near 530.5 eV) is relatively large in the UV0 sample (17%) and decreases in the UV5/20 samples (6%). This phenomenon is ascribed to the removal of unreacted species, such as AlOOH, and enhanced ligand exchange by UV irradiation, which supplements insufficient thermal energy.[17] In summary, UV‐ALD has proven to deposit a denser and smoother stoichiometric Al2O3 thin film with less contamination.2FigureCharacterizations of UV‐ALD Al2O3 film on Si: a) The summary of density and roughness based on XRR, b) XPS spectra of Al2O3 film deposited by UV‐ALD. (UV0/5/20), c–e) high‐resolution XPS spectra of O1s peak of c) UV0, d) UV5, and e) UV20, and f) the summary of the area intensity ratio between Al2O3 and (Al2O3 + AlOOH) peaks based on the O1s peak deconvolution.After confirming the denser and smoother stoichiometric Al2O3 film deposition by UV‐ALD, we applied UV‐ALD to Al2O3 deposition on graphene. The Al2O3 film morphologies were characterized by HR‐SEM and AFM (Figures S4, Supporting Information; Figure 3a‐e). While Al2O3 is deposited predominantly on the grain boundaries of graphene owing to the high chemical reactivity of grain boundaries, i.e., line defects,[21] HR‐SEM and AFM analysis of the Al2O3 deposition on the basal plane of graphene show that the film becomes rougher (17.5 ± 0.4 nm (UV0), 18.2 ± 0.5 nm (UV5), 20.1 ± 0.6 nm (UV20), and 27.0 ± 0.43 nm (UV50)), the average grain size becomes larger (17.5 ± 0.4 nm (UV0), 18.2 ± 0.5 nm (UV5), 20.1 ± 0.6 nm (UV20), and 27.0 ± 0.43 nm (UV50)), and the area coverage decreases (97% (UV0), 96% (UV5), 85% (UV20), and 44% (UV50)) with increasing UV irradiation time (Figure 3f). Additionally, XPS analysis shows that the carbon content at the surface decreases at first from 13.5 at% (UV0) to 10.6/11.6 at% (UV5/20) due to more facile ligand exchange as was observed in the experiment on Si and increases to 41.8 at% (UV50) due to the bare graphene surface without Al2O3 layer being partially exposed (Figure 3g). Characteristic peaks of trigonal Al2O3 (113) (from the standard ICDD, PDF card no. 01‐070‐3319) and graphitic carbon (004)[22] are shown in XRD spectra, regardless of the UV irradiation time (Figure 3h). Overall, the morphology analysis indicates that the excessive (>20 s) UV irradiation in the ALD process clearly inhibited the initial nucleation of the Al2O3 film on the graphene basal plane, leading to sparsely distributed large grains with low area coverage, while the optimal (5 s) UV irradiation did not.3FigureAFM images of UV‐ALD Al2O3 film on graphene of a) UV0, b) UV5, c) UV20, and d) UV50, and e) the comparison of line profiles. f) The summary of grain size and area coverage from AFM analysis. g) XPS spectra and h) GI‐XRD spectra of UV‐ALD Al2O3 film on graphene.This trend is opposed to the result on the Si substrate; the UV irradiation in the H2O half cycle of the ALD process is known to create more OH functional groups, which readily react with the TMA molecules in the following half cycle, and increase the nucleation density, which may have occurred in the deposition on Si substrate. Nevertheless, the UV irradiation in the ALD process on graphene substrates may have also affected the desorption of the OH functional groups at the same time. It is notable that the UV energy (3–5 eV) used in the ALD process is larger than the calculated binding energies of C–OH bonding (2–3 eV)[23] or graphene‐TMA bonding (1.84 eV),[14] which are smaller than that of SiOH (5.0 eV).[24] Therefore, UV energy may be effective in breaking graphene–OH or graphene–TMA bondings, especially in excessive UV irradiation time (>20 s). Similarly, Mulyana, et al. reported that UV irradiation can effectively reduce the graphene oxide produced by UV/O3 treatment into graphene at room temperature atmospheric conditions.[25] As a result, UV20/50 on graphene seems to have a rough and non‐uniform surface due to the elimination of functional groups at the initial stage of film growth and the lower density of nucleation sites.Raman spectra of graphene with ALD Al2O3 layer deposited without (UV0) and with UV irradiation (UV5/20/50) are demonstrated (Figure 4). Raman results show Al2O3 peaks at ≈500 cm−1 and Si peak at ≈950 cm−1 indicating Al2O3 deposition on graphene/SiO2/Si substrate.[26] The G (1580 cm−1) peak is known to stem from the ideal graphitic lattice, i.e., the CC bond of graphite materials, and the D peak (1350 cm−1) represents the disordered graphitic lattice.[27] The broadening and the positive shift of the G peak as a function of UV irradiation time imply the formation of structural disorder in graphene (Figure 4b).[28] High‐resolution Raman analysis further unveils the nature of the defects on the graphene surface (Figure 4c–f). Deconvolution of the D peak identifies the bands from disordered graphitic lattice (e.g., edges (D1), surface layers (D2), polyenes/ionic impurities (D4)), and amorphous carbon (D3).[29] The D peak is hardly observed in the graphene substrate of the UV0 sample; however, as the UV irradiation time increases, the D peak increases, indicating that the symmetry of the graphene lattice is broken, and defects are formed. UV irradiation clearly shows to induce various kinds of disorders in graphitic lattice on the graphene layer as proven by the increase of all the D1 (most significant), D2, and D4 bands intensities, which is even more prevalent in UV20/50 samples. The slight increase of D3 band intensity implies that the transition from crystalline to amorphous carbon may have occurred possibly due to the accumulation of defects on graphene.[30]4Figurea) Raman spectra and b) peak position and FWHM of G peaks of graphene with UV‐ALD (UV0/5/20/50) Al2O3 layer. High‐resolution spectra of c) UV0, d) UV5, e) UV20, and f) UV50.The transfer characteristics of GFETs based on UV‐ALD Al2O3 coated graphene as an electron channel were analyzed to determine the effect of the UV irradiation time. GFETs with an Al2O3/Au/graphene structure (Figure 5a) were fabricated and annealed (360 °C, 2 h, Ar/H2) to remove polymer residues that may induce strong p‐doping and adversely affect mobility.[31] Figure 5b shows the transfer characteristics of GFETs with UV0/5/20/50 ALD Al2O3 layers (namely, UV0/5/20/50 GFETs), and Figure 5c,d shows the summarized data of Dirac voltage (VDirac) and hole mobility (µh). To begin with, GFETs with annealed bare graphene (before ALD) showed the VDirac of 43.5 ± 6.9 V and the µh of 899 ± 31 cm2 V−1 s−1 (gray lines in Figure 5c,d). UV0 GFET shows extremely high VDirac (120 V) and low µh (397 cm2 V−1 s−1) possibly owing to residual unreacted oxygen‐containing groups between graphene and Al2O3, resulting in p‐type doping of graphene; however, UV5 GFET showed even lower VDirac (8 V) and 207% increase in µh (1221 cm2 V−1 s−1) compared to UV0 GFET, which implies the removal of unreacted oxygen‐containing groups and the formation of the more intimate graphene‐Al2O3 interface.[15] As the UV irradiation time further increases, VDirac in turn increases to 20 V (UV20) and >120 V (UV50), and µh decreases to 184 cm2 V−1 s−1 (UV20) and 38 cm2 V−1 s−1 (UV50). The degradation of carrier mobility is ascribed to defect formation on graphene as shown in Figure 4c–f), where Dsum/G increased from 0.82 (UV0) to 1.15/1.83/1.84 (UV5/20/50) as UV irradiation becomes longer.[32] Also, the exposure of graphene surface to air in UV50 GFETs due to non‐continuous ALD film deposition did not fully passivate the graphene from ambient H2O and O2, resulting in p‐doping and mobility decrease of graphene.[16] The structural defects induced by excessive UV irradiation on the graphene surface may have expedited such phenomena. This study demonstrates that UV‐ALD is an effective and simple option to realize high‐quality interface between graphene and dielectric thin film with improved GFET performance compared to previous studies as shown in Table S1 (Supporting Information).5Figurea) Device schematic of Al2O3 / Au / graphene GFETs structure. b) Transfer characteristics of GFETs with UV0/5/20/50 ALD Al2O3 dielectric layers. c,d) Summary of Dirac voltage (VDirac) and hole mobility (µh) of GFETs according to UV irradiation time (gray lines represent the properties of bare graphenes before ALD).ConclusionIn conclusion, we reported the first successful application of UV‐ALD on a graphene substrate and the enhanced performance of GFETs with Al2O3 dielectric thin films deposited by UV‐ALD. During the ALD process, the optimal UV irradiation, i.e., 5 s per cycle (UV5), induced the growth of uniform film by reconstructing nucleation sites with the intimate graphene‐dielectric interface; it also replenished insufficient thermal energy for ligand exchange, leading to the deposition of high‐quality, i.e., denser, smoother, and purer, Al2O3 film even at low temperature (<100 °C). In contrast, excessive UV irradiation (UV20/50) caused nonuniform film deposition possibly owing to the debonding of ‐OH or TMA on graphene that is otherwise reaction sites in ALD half‐cycles. As a result, the GFETs with high‐quality UV‐ALD dielectric layer with optimal UV irradiation show improved performance with VDriac close to 0 V and µh of 1221 cm2 V−1 s−1, approximately 3 times increase compared to the GFET with thermal ALD dielectric layer. The successful application of low‐temperature UV‐ALD to GFETs with the high‐quality dielectric‐graphene interface shown in this study may have significant implications in realizing high‐performance graphene or other 2D materials (e.g., MoS2, h‐BN)‐based flexible electronic devices.AcknowledgementsThis research was supported by the Nano‐Convergence Foundation funded by the Ministry of Trade, Industry, and Energy (MOTIE) of Korea (no. 20000272). This work also was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2022R1A2C400120511) and by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF‐2021R1A6A1A03039981). The Raman analysis was conducted by Dr. Minkyung Choi at Raman Research Institute.Conflict of InterestThe authors declare no conflict of interest.Data Availability StatementThe data that support the findings of this study are available from the corresponding author upon reasonable request.C. Lee, X. Wei, J. W. Kysar, J. Hone, Science 2008, 321, 385.F. Schwierz, Nat. Nanotechnol. 2010, 5, 487.Y. Shao, J. Wang, H. Wu, J. Liu, I. A. Aksay, Y. Lin, Electroanalysis 2010, 22, 1027.K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J.‐H. Ahn, P. Kim, J.‐Y. Choi, B. H. Hong, Nature 2009, 457, 706.S.‐K. Lee, B. J. Kim, H. Jang, S. C. Yoon, C. Lee, B. H. Hong, J. A. Rogers, J. H. Cho, J.‐H. Ahn, Nano Lett. 2011, 11, 4642.F. Yan, M. Zhang, J. Li, Adv. Healthcare Mater. 2014, 3, 313.V. Miikkulainen, M. Vehkamäki, K. Mizohata, T. Hatanpää, Adv. Mater. Interfaces 2021, 8, 2100014.X. Wang, S. M. Tabakman, H. Dai, J. Am. Chem. Soc. 2008, 130, 8152.M. Baitimirova, R. Viter, J. Andzane, A. van der Lee, D. Voiry, I. Iatsunskyi, E. Coy, L. Mikoliunaite, S. Tumenas, K. Załęski, J. Phys. Chem. C 2016, 120, 23716.K. P. Loh, Q. Bao, P. K. Ang, J. Yang, J. Mater. Chem. 2010, 20, 2277.N. Y. Garces, V. D. Wheeler, J. K. Hite, G. G. Jernigan, J. L. Tedesco, N. Nepal, C. R. EddyJr, D. K. Gaskill, J. Appl. Phys. 2011, 109, 124304.P. Solís‐Fernández, J. I. Paredes, S. Villar‐Rodil, L. Guardia, M. J. Fernández‐Merino, G. Dobrik, L. P. Biró, A. Martínez‐Alonso, J. M. D. Tascón, J. Phys. Chem. C 2011, 115, 7956.W. C. Shin, J. H. Bong, S.‐Y. Choi, B. J. Cho, ACS Appl. Mater. Interfaces 2013, 5, 11515.R. H. J. Vervuurt, B. Karasulu, M. A. Verheijen, W. (Erwin) M. M. Kessels, A. A. Bol, Chem. Mater. 2017, 29, 2090.J. W. Shin, M. H. Kang, S. Oh, B. C. Yang, K. Seong, H.‐S. Ahn, T. H. Lee, J. An, Nanotechnology 2018, 29, 195602.A. di Bartolomeo, F. Giubileo, F. Romeo, P. Sabatino, G. Carapella, L. Iemmo, T. Schroeder, G. Lupina, Nanotechnology 2015, 26, 475202.K. H. Yoon, H. Kim, Y.‐E. K. Lee, N. K. Shrestha, M. M. Sung, RSC Adv. 2017, 7, 5601.Y.‐C. Lin, C.‐C. Lu, C.‐H. Yeh, C. Jin, K. Suenaga, P.‐W. Chiu, Nano Lett. 2012, 12, 414.W. Choi, M. A. Shehzad, S. Park, Y. Seo, RSC Adv. 2017, 7, 6943.M. D. Groner, F. H. Fabreguette, J. W. Elam, S. M. George, Chem. Mater. 2004, 16, 639.K. Kim, H.‐B.‐R. Lee, R. W. Johnson, J. T. Tanskanen, N. Liu, M.‐G. Kim, C. Pang, C. Ahn, S. F. Bent, Z. Bao, Nat. Commun. 2014, 5, 1.Z. Q. Li, C. J. Lu, Z. P. Xia, Y. Zhou, Z. Luo, Carbon 2007, 45, 1686.A. S. Dobrota, S. Gutić, A. Kalijadis, M. Baljozović, S. v Mentus, N. v Skorodumova, I. A. Pašti, RSC Adv. 2016, 6, 57910.D. J. Lucas, L. A. Curtiss, J. A. Pople, J. Chem. Phys. 1993, 99, 6697.Y. Mulyana, M. Uenuma, Y. Ishikawa, Y. Uraoka, J. Phys. Chem. C 2014, 118, 27372.T. Sudare, A. Zenzai, S. Tamura, M. Kiyama, F. Hayashi, K. Teshima, CrystEngComm 2019, 21, 7211.A. Eckmann, A. Felten, A. Mishchenko, L. Britnell, R. Krupke, K. S. Novoselov, C. Casiraghi, Nano Lett 2012, 12, 3925.C. Casiraghi, Phys. Status Solidi RRL 2009, 3, 175.A. Sadezky, H. Muckenhuber, H. Grothe, R. Niessner, U. Pöschl, Carbon 2005, 43, 1731.A. Gao, C. J. Lee, F. Bijkerk, J. Appl. Phys. 2014, 116, 054312.A. Pirkle, J. Chan, A. Venugopal, D. Hinojos, C. W. Magnuson, S. McDonnell, L. Colombo, E. M. Vogel, R. S. Ruoff, R. M. Wallace, Appl. Phys. Lett. 2011, 99, 122108.R. Beams, L. G. Cançado, L. Novotny, J Phys Condens Matter 2015, 27, 083002. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Advanced Electronic Materials Wiley

High‐Performance Graphene‐Dielectric Interface by UV‐Assisted Atomic Layer Deposition for Graphene Field Effect Transistor

Loading next page...
 
/lp/wiley/high-performance-graphene-dielectric-interface-by-uv-assisted-atomic-zCofkx7Irw

References (32)

Publisher
Wiley
Copyright
© 2023 Wiley‐VCH GmbH
eISSN
2199-160X
DOI
10.1002/aelm.202300074
Publisher site
See Article on Publisher Site

Abstract

IntroductionGraphene has excellent electrical and mechanical properties, such as high carrier mobility, surface‐to‐volume ratio, and flexibility.[1–3] These interesting properties of graphene make it a promising candidate material for flexible electronics and sensors.[4,5] In particular, sensors based on graphene field‐effect transistors (GFETs) exhibit high sensitivity owing to their high transconductance.[6] In the fabrication of high‐sensitivity GFET devices, the interface property between graphene and the dielectric film is one of the design factors that highly influence device performance.[7] Therefore, to manufacture such high‐sensitivity GFET‐based sensors, it is essential to deposit a thin and uniform high‐k dielectric film on the graphene surface without the degradation of graphene's superior electrical properties.Atomic layer deposition (ALD) has been reported to be an adequate method for depositing highly uniform thin films with precise thickness control on various substrates, including graphene. As a chemical vapor deposition (CVD) variant, ALD has been shown to fabricate intimate thin film‐graphene interfaces.[8,9] However, the film formation is hindered by the chemically inert nature of graphene's basal plane due to its strong sp2 carbon bonding.[10] Therefore, various functionalization techniques for graphene surfaces, such as wet chemical treatment,[11] thermal oxidation treatment,[12] and seed layering,[13] have been studied. Although these techniques functionalize the graphene surface, they may cause degradation of graphene properties owing to chemical contamination and problems such as the formation of unnecessary seed layers. Surface functionalization using plasma is another graphene surface functionalization technique that is being actively researched.[14] However, graphene may be damaged or etched out because of high‐energy ions. In our previous study, we applied an atmospheric pressure plasma treatment technique to functionalize the graphene surface without degrading its properties with low plasma energy while depositing a high‐quality ALD dielectric film on the graphene surface.[15] However, these pre‐treatment techniques necessitate additional treatment processes and graphene may be exposed to air after the pre‐treatment process for the deposition of the dielectric film. As graphene is exposed to air during manufacturing, it can be p‐doped by environmental sources such as H2O and O2 in the air.[16]UV‐assisted ALD (UV‐ALD), in which the film is exposed to UV light at every cycle during the ALD process, is an effective ALD technique that can possibly solve these problems. UV light with a relatively low energy range (<10 eV) can functionalize the graphene surface without damaging or etching. A high‐quality dielectric film with fewer contaminants can be deposited as UV energy is transferred during the deposition process, facilitating the ligand exchange reaction. Yoon et al. reported that dense stoichiometric Al2O3 thin films could be deposited via UV‐ALD on Si substrate.[17] Consequently, a thin film can be deposited by applying UV‐ALD on the graphene surface without degrading its superior electrical properties, thereby realizing a high‐quality dielectric‐graphene interface for high‐performance GFETs. Nevertheless, to the best of our knowledge, the application of UV‐ALD to thin‐film deposition on 2D materials, including graphene surfaces or graphene‐based devices, has never been demonstrated.In this study, we demonstrate the first successful application of UV‐ALD to effectively functionalize graphene surfaces and, simultaneously, fabricate high‐performance GFETs with UV‐ALD Al2O3 dielectric thin films. High‐resolution field emission scanning electron microscopy (HR‐SEM), atomic force microscopy (AFM), X‐ray diffraction (XRD), and Raman spectroscopy results showed the optimal per‐cycle UV irradiation (5 s per cycle) in the ALD process promote the density of the Al2O3 dielectric film by facilitating the ligand exchange reaction, while simultaneously improving the graphene‐dielectric interface quality. As a result, GFETs with an Al2O3 dielectric layer deposited by UV‐ALD showed a low Dirac voltage close to 0 V (i.e., 8 V) and high hole mobility of 1221 cm2 V−1 s−1 that is a 207% improvement compared to those by ALD without UV.Experimental SectionUV‐ALD ProcessA customized ALD system equipped with a UV lamp (PE300BUV, Cermax) on the chamber's lid was used, as shown in Figure 1a. The distance between the UV lamp and substrate was maintained at 6 cm with a CaF2 window to pass UV light. ALD Al2O3 film was deposited on a Si wafer (reference) and graphene using trimethylaluminum (TMA) as the precursor and deionized water as the oxygen source. Argon gas was used as the carrier gas at a flow rate of 40 sccm. TMA and H2O were self‐evaporated at room temperature and the reaction chamber was maintained at <100 °C. Each ALD cycle consisted of 1) 0.1 s TMA pulse, 2) 60 s purge, 3) 0.1 s H2O pulse, and 4) 60 s purge with UV irradiation for 0, 5, 20, and 50 s at the H2O pulse step (Figure 1b, after which the samples were named (UV0/5/20/50). The Al2O3 film was deposited with the ALD cycle number of 150 cycles on Si and graphene substrates. The UV lamp power was fixed at 230 W, corresponding to an intensity of 0.12 W cm−2, and the UV wavelength and energy ranges were mostly in the range of 250–400 nm and 3–5 eV, respectively (>95% of relative output).1FigureSchematic of a) apparatus and b) process (1 cycle) of UV‐ALD used in this study.GFET Device FabricationThe GFETs used in this study were fabricated based on multilayer CVD graphene on 285 nm SiO2/ highly p‐doped Si (Graphene Supermarkets), which was followed by annealing in Ar (200 sccm) and H2 (50 sccm) atmospheres at 360 °C for 2 h. In this step, surface contaminants, such as polymer residues, were eliminated.[18,19] After annealing, a 20 nm Au electrode was deposited by thermal evaporation. Finally, an Al2O3 dielectric layer (150 cycles) was deposited by UV‐ALD on the annealed graphene.Structural and Electrical CharacterizationRaman spectroscopy of the graphene surface was carried out using a LabRam Aramis system (Horiba Jobin Yvon, Japan) equipped with a 514 nm laser. The laser was focused onto a 1 µm spot size using a 100× microscope objective. Compositional analysis was conducted by X‐ray photoelectron spectroscopy (XPS) (K‐Alpha+, Thermo Fisher Scientific) with a monochromated Al X‐ray source gun, a spot size of 200 µm, and a step size of 0.05 eV step−1. The Al2O3 film surface morphology was analyzed using high‐resolution field‐emission scanning electron microscopy (HR‐FESEM) (SU8010, Hitachi High Technologies Corporation) and atomic force microscopy (AFM) (XE‐100, Park Systems). The crystallinity, thickness, density, and roughness of the ALD Al2O3 films were characterized by grazing‐incidence X‐ray diffraction (GIXRD) and X‐ray reflectivity (XRR) (SmartLab, Rigaku Corporation). The electrical properties of the GFETs were measured in the air using a semiconductor analyzer (4200 SCS, Keithley). The back gate voltage (Vbg) was swept from −120 to 120 V, and the drain‐source current (Ids) was measured with the drain‐source voltage (Vds) of 0.1 V.Result and DiscussionFirst, we deposited and tested UV‐ALD Al2O3 on the Si substrate to observe the effect of UV irradiation solely on the film properties. The Al2O3 film became smoother under UV irradiation during the ALD process, as shown in Figure S1 (Supporting Information). Based on XRR analysis, the density of Al2O3 film increased from 2.9 to 3.2 and 3.3 g cm−3 with UV irradiation times of 0, 5, and 20 s (UV0/5/20), respectively (Figure S2, Supporting Information; Figure 2a) with similarly small roughness of < 0.1 nm in all samples. The density of UV0 on the Si sample (ALD Al2O3 without UV) was comparable to the previously reported value (2.8–2.9 g cm−3)[20] for thermal ALD Al2O3 deposited at < 100 °C, while that of UV5/20 on the Si samples increased by about 10%. In the XPS survey analysis, all Al2O3 films on Si with and without UV irradiation were stoichiometric (Al: 36.9, 39.3, and 39.3 at%; O: 61.5, 60, and 60 at%) (Figure 2b). The carbon content in the Al2O3 films decreased upon UV irradiation from 1.6 at% (UV0) to 0.7 at% (UV5/20) (Figure 2b). This is because UV irradiation eliminates unreacted species at low temperatures and recreates functional groups during the ALD process, as reported by Yoon et al.[17] Interestingly, the high‐resolution XPS analysis of the O1s peak (Figure 2c–e) and Al2p (Figure S3, Supporting Information) shows a clear contrast between ALD Al2O3 films without and with UV irradiation. The Al2O3 peaks (near 532.2 eV) are dominant in all samples in Figure 2c–f (83/94/94% in UV0/5/20); in the meantime, the AlOOH peak (near 530.5 eV) is relatively large in the UV0 sample (17%) and decreases in the UV5/20 samples (6%). This phenomenon is ascribed to the removal of unreacted species, such as AlOOH, and enhanced ligand exchange by UV irradiation, which supplements insufficient thermal energy.[17] In summary, UV‐ALD has proven to deposit a denser and smoother stoichiometric Al2O3 thin film with less contamination.2FigureCharacterizations of UV‐ALD Al2O3 film on Si: a) The summary of density and roughness based on XRR, b) XPS spectra of Al2O3 film deposited by UV‐ALD. (UV0/5/20), c–e) high‐resolution XPS spectra of O1s peak of c) UV0, d) UV5, and e) UV20, and f) the summary of the area intensity ratio between Al2O3 and (Al2O3 + AlOOH) peaks based on the O1s peak deconvolution.After confirming the denser and smoother stoichiometric Al2O3 film deposition by UV‐ALD, we applied UV‐ALD to Al2O3 deposition on graphene. The Al2O3 film morphologies were characterized by HR‐SEM and AFM (Figures S4, Supporting Information; Figure 3a‐e). While Al2O3 is deposited predominantly on the grain boundaries of graphene owing to the high chemical reactivity of grain boundaries, i.e., line defects,[21] HR‐SEM and AFM analysis of the Al2O3 deposition on the basal plane of graphene show that the film becomes rougher (17.5 ± 0.4 nm (UV0), 18.2 ± 0.5 nm (UV5), 20.1 ± 0.6 nm (UV20), and 27.0 ± 0.43 nm (UV50)), the average grain size becomes larger (17.5 ± 0.4 nm (UV0), 18.2 ± 0.5 nm (UV5), 20.1 ± 0.6 nm (UV20), and 27.0 ± 0.43 nm (UV50)), and the area coverage decreases (97% (UV0), 96% (UV5), 85% (UV20), and 44% (UV50)) with increasing UV irradiation time (Figure 3f). Additionally, XPS analysis shows that the carbon content at the surface decreases at first from 13.5 at% (UV0) to 10.6/11.6 at% (UV5/20) due to more facile ligand exchange as was observed in the experiment on Si and increases to 41.8 at% (UV50) due to the bare graphene surface without Al2O3 layer being partially exposed (Figure 3g). Characteristic peaks of trigonal Al2O3 (113) (from the standard ICDD, PDF card no. 01‐070‐3319) and graphitic carbon (004)[22] are shown in XRD spectra, regardless of the UV irradiation time (Figure 3h). Overall, the morphology analysis indicates that the excessive (>20 s) UV irradiation in the ALD process clearly inhibited the initial nucleation of the Al2O3 film on the graphene basal plane, leading to sparsely distributed large grains with low area coverage, while the optimal (5 s) UV irradiation did not.3FigureAFM images of UV‐ALD Al2O3 film on graphene of a) UV0, b) UV5, c) UV20, and d) UV50, and e) the comparison of line profiles. f) The summary of grain size and area coverage from AFM analysis. g) XPS spectra and h) GI‐XRD spectra of UV‐ALD Al2O3 film on graphene.This trend is opposed to the result on the Si substrate; the UV irradiation in the H2O half cycle of the ALD process is known to create more OH functional groups, which readily react with the TMA molecules in the following half cycle, and increase the nucleation density, which may have occurred in the deposition on Si substrate. Nevertheless, the UV irradiation in the ALD process on graphene substrates may have also affected the desorption of the OH functional groups at the same time. It is notable that the UV energy (3–5 eV) used in the ALD process is larger than the calculated binding energies of C–OH bonding (2–3 eV)[23] or graphene‐TMA bonding (1.84 eV),[14] which are smaller than that of SiOH (5.0 eV).[24] Therefore, UV energy may be effective in breaking graphene–OH or graphene–TMA bondings, especially in excessive UV irradiation time (>20 s). Similarly, Mulyana, et al. reported that UV irradiation can effectively reduce the graphene oxide produced by UV/O3 treatment into graphene at room temperature atmospheric conditions.[25] As a result, UV20/50 on graphene seems to have a rough and non‐uniform surface due to the elimination of functional groups at the initial stage of film growth and the lower density of nucleation sites.Raman spectra of graphene with ALD Al2O3 layer deposited without (UV0) and with UV irradiation (UV5/20/50) are demonstrated (Figure 4). Raman results show Al2O3 peaks at ≈500 cm−1 and Si peak at ≈950 cm−1 indicating Al2O3 deposition on graphene/SiO2/Si substrate.[26] The G (1580 cm−1) peak is known to stem from the ideal graphitic lattice, i.e., the CC bond of graphite materials, and the D peak (1350 cm−1) represents the disordered graphitic lattice.[27] The broadening and the positive shift of the G peak as a function of UV irradiation time imply the formation of structural disorder in graphene (Figure 4b).[28] High‐resolution Raman analysis further unveils the nature of the defects on the graphene surface (Figure 4c–f). Deconvolution of the D peak identifies the bands from disordered graphitic lattice (e.g., edges (D1), surface layers (D2), polyenes/ionic impurities (D4)), and amorphous carbon (D3).[29] The D peak is hardly observed in the graphene substrate of the UV0 sample; however, as the UV irradiation time increases, the D peak increases, indicating that the symmetry of the graphene lattice is broken, and defects are formed. UV irradiation clearly shows to induce various kinds of disorders in graphitic lattice on the graphene layer as proven by the increase of all the D1 (most significant), D2, and D4 bands intensities, which is even more prevalent in UV20/50 samples. The slight increase of D3 band intensity implies that the transition from crystalline to amorphous carbon may have occurred possibly due to the accumulation of defects on graphene.[30]4Figurea) Raman spectra and b) peak position and FWHM of G peaks of graphene with UV‐ALD (UV0/5/20/50) Al2O3 layer. High‐resolution spectra of c) UV0, d) UV5, e) UV20, and f) UV50.The transfer characteristics of GFETs based on UV‐ALD Al2O3 coated graphene as an electron channel were analyzed to determine the effect of the UV irradiation time. GFETs with an Al2O3/Au/graphene structure (Figure 5a) were fabricated and annealed (360 °C, 2 h, Ar/H2) to remove polymer residues that may induce strong p‐doping and adversely affect mobility.[31] Figure 5b shows the transfer characteristics of GFETs with UV0/5/20/50 ALD Al2O3 layers (namely, UV0/5/20/50 GFETs), and Figure 5c,d shows the summarized data of Dirac voltage (VDirac) and hole mobility (µh). To begin with, GFETs with annealed bare graphene (before ALD) showed the VDirac of 43.5 ± 6.9 V and the µh of 899 ± 31 cm2 V−1 s−1 (gray lines in Figure 5c,d). UV0 GFET shows extremely high VDirac (120 V) and low µh (397 cm2 V−1 s−1) possibly owing to residual unreacted oxygen‐containing groups between graphene and Al2O3, resulting in p‐type doping of graphene; however, UV5 GFET showed even lower VDirac (8 V) and 207% increase in µh (1221 cm2 V−1 s−1) compared to UV0 GFET, which implies the removal of unreacted oxygen‐containing groups and the formation of the more intimate graphene‐Al2O3 interface.[15] As the UV irradiation time further increases, VDirac in turn increases to 20 V (UV20) and >120 V (UV50), and µh decreases to 184 cm2 V−1 s−1 (UV20) and 38 cm2 V−1 s−1 (UV50). The degradation of carrier mobility is ascribed to defect formation on graphene as shown in Figure 4c–f), where Dsum/G increased from 0.82 (UV0) to 1.15/1.83/1.84 (UV5/20/50) as UV irradiation becomes longer.[32] Also, the exposure of graphene surface to air in UV50 GFETs due to non‐continuous ALD film deposition did not fully passivate the graphene from ambient H2O and O2, resulting in p‐doping and mobility decrease of graphene.[16] The structural defects induced by excessive UV irradiation on the graphene surface may have expedited such phenomena. This study demonstrates that UV‐ALD is an effective and simple option to realize high‐quality interface between graphene and dielectric thin film with improved GFET performance compared to previous studies as shown in Table S1 (Supporting Information).5Figurea) Device schematic of Al2O3 / Au / graphene GFETs structure. b) Transfer characteristics of GFETs with UV0/5/20/50 ALD Al2O3 dielectric layers. c,d) Summary of Dirac voltage (VDirac) and hole mobility (µh) of GFETs according to UV irradiation time (gray lines represent the properties of bare graphenes before ALD).ConclusionIn conclusion, we reported the first successful application of UV‐ALD on a graphene substrate and the enhanced performance of GFETs with Al2O3 dielectric thin films deposited by UV‐ALD. During the ALD process, the optimal UV irradiation, i.e., 5 s per cycle (UV5), induced the growth of uniform film by reconstructing nucleation sites with the intimate graphene‐dielectric interface; it also replenished insufficient thermal energy for ligand exchange, leading to the deposition of high‐quality, i.e., denser, smoother, and purer, Al2O3 film even at low temperature (<100 °C). In contrast, excessive UV irradiation (UV20/50) caused nonuniform film deposition possibly owing to the debonding of ‐OH or TMA on graphene that is otherwise reaction sites in ALD half‐cycles. As a result, the GFETs with high‐quality UV‐ALD dielectric layer with optimal UV irradiation show improved performance with VDriac close to 0 V and µh of 1221 cm2 V−1 s−1, approximately 3 times increase compared to the GFET with thermal ALD dielectric layer. The successful application of low‐temperature UV‐ALD to GFETs with the high‐quality dielectric‐graphene interface shown in this study may have significant implications in realizing high‐performance graphene or other 2D materials (e.g., MoS2, h‐BN)‐based flexible electronic devices.AcknowledgementsThis research was supported by the Nano‐Convergence Foundation funded by the Ministry of Trade, Industry, and Energy (MOTIE) of Korea (no. 20000272). This work also was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2022R1A2C400120511) and by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF‐2021R1A6A1A03039981). The Raman analysis was conducted by Dr. Minkyung Choi at Raman Research Institute.Conflict of InterestThe authors declare no conflict of interest.Data Availability StatementThe data that support the findings of this study are available from the corresponding author upon reasonable request.C. Lee, X. Wei, J. W. Kysar, J. Hone, Science 2008, 321, 385.F. Schwierz, Nat. Nanotechnol. 2010, 5, 487.Y. Shao, J. Wang, H. Wu, J. Liu, I. A. Aksay, Y. Lin, Electroanalysis 2010, 22, 1027.K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J.‐H. Ahn, P. Kim, J.‐Y. Choi, B. H. Hong, Nature 2009, 457, 706.S.‐K. Lee, B. J. Kim, H. Jang, S. C. Yoon, C. Lee, B. H. Hong, J. A. Rogers, J. H. Cho, J.‐H. Ahn, Nano Lett. 2011, 11, 4642.F. Yan, M. Zhang, J. Li, Adv. Healthcare Mater. 2014, 3, 313.V. Miikkulainen, M. Vehkamäki, K. Mizohata, T. Hatanpää, Adv. Mater. Interfaces 2021, 8, 2100014.X. Wang, S. M. Tabakman, H. Dai, J. Am. Chem. Soc. 2008, 130, 8152.M. Baitimirova, R. Viter, J. Andzane, A. van der Lee, D. Voiry, I. Iatsunskyi, E. Coy, L. Mikoliunaite, S. Tumenas, K. Załęski, J. Phys. Chem. C 2016, 120, 23716.K. P. Loh, Q. Bao, P. K. Ang, J. Yang, J. Mater. Chem. 2010, 20, 2277.N. Y. Garces, V. D. Wheeler, J. K. Hite, G. G. Jernigan, J. L. Tedesco, N. Nepal, C. R. EddyJr, D. K. Gaskill, J. Appl. Phys. 2011, 109, 124304.P. Solís‐Fernández, J. I. Paredes, S. Villar‐Rodil, L. Guardia, M. J. Fernández‐Merino, G. Dobrik, L. P. Biró, A. Martínez‐Alonso, J. M. D. Tascón, J. Phys. Chem. C 2011, 115, 7956.W. C. Shin, J. H. Bong, S.‐Y. Choi, B. J. Cho, ACS Appl. Mater. Interfaces 2013, 5, 11515.R. H. J. Vervuurt, B. Karasulu, M. A. Verheijen, W. (Erwin) M. M. Kessels, A. A. Bol, Chem. Mater. 2017, 29, 2090.J. W. Shin, M. H. Kang, S. Oh, B. C. Yang, K. Seong, H.‐S. Ahn, T. H. Lee, J. An, Nanotechnology 2018, 29, 195602.A. di Bartolomeo, F. Giubileo, F. Romeo, P. Sabatino, G. Carapella, L. Iemmo, T. Schroeder, G. Lupina, Nanotechnology 2015, 26, 475202.K. H. Yoon, H. Kim, Y.‐E. K. Lee, N. K. Shrestha, M. M. Sung, RSC Adv. 2017, 7, 5601.Y.‐C. Lin, C.‐C. Lu, C.‐H. Yeh, C. Jin, K. Suenaga, P.‐W. Chiu, Nano Lett. 2012, 12, 414.W. Choi, M. A. Shehzad, S. Park, Y. Seo, RSC Adv. 2017, 7, 6943.M. D. Groner, F. H. Fabreguette, J. W. Elam, S. M. George, Chem. Mater. 2004, 16, 639.K. Kim, H.‐B.‐R. Lee, R. W. Johnson, J. T. Tanskanen, N. Liu, M.‐G. Kim, C. Pang, C. Ahn, S. F. Bent, Z. Bao, Nat. Commun. 2014, 5, 1.Z. Q. Li, C. J. Lu, Z. P. Xia, Y. Zhou, Z. Luo, Carbon 2007, 45, 1686.A. S. Dobrota, S. Gutić, A. Kalijadis, M. Baljozović, S. v Mentus, N. v Skorodumova, I. A. Pašti, RSC Adv. 2016, 6, 57910.D. J. Lucas, L. A. Curtiss, J. A. Pople, J. Chem. Phys. 1993, 99, 6697.Y. Mulyana, M. Uenuma, Y. Ishikawa, Y. Uraoka, J. Phys. Chem. C 2014, 118, 27372.T. Sudare, A. Zenzai, S. Tamura, M. Kiyama, F. Hayashi, K. Teshima, CrystEngComm 2019, 21, 7211.A. Eckmann, A. Felten, A. Mishchenko, L. Britnell, R. Krupke, K. S. Novoselov, C. Casiraghi, Nano Lett 2012, 12, 3925.C. Casiraghi, Phys. Status Solidi RRL 2009, 3, 175.A. Sadezky, H. Muckenhuber, H. Grothe, R. Niessner, U. Pöschl, Carbon 2005, 43, 1731.A. Gao, C. J. Lee, F. Bijkerk, J. Appl. Phys. 2014, 116, 054312.A. Pirkle, J. Chan, A. Venugopal, D. Hinojos, C. W. Magnuson, S. McDonnell, L. Colombo, E. M. Vogel, R. S. Ruoff, R. M. Wallace, Appl. Phys. Lett. 2011, 99, 122108.R. Beams, L. G. Cançado, L. Novotny, J Phys Condens Matter 2015, 27, 083002.

Journal

Advanced Electronic MaterialsWiley

Published: Jul 1, 2023

Keywords: dielectric films; graphene field effect transistors; surface functionalization; UV‐assisted atomic layer deposition

There are no references for this article.