Nanoscale View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
COMMUNICATION
Cite this: Nanoscale, 2014, 6, 12315 Received 20th June 2014, Accepted 7th August 2014 DOI: 10.1039/c4nr03448a www.rsc.org/nanoscale
View Journal | View Issue
Stable charge storing in two-dimensional MoS2 nanoflake floating gates for multilevel organic flash memory† Minji Kang,a Yeong-A. Kim,a Jin-Mun Yun,b Dongyoon Khim,a Jihong Kim,a Yong-Young Noh,*c Kang-Jun Baeg*d and Dong-Yu Kim*a
In this study, we investigated chemically exfoliated two-dimensional (2-D) nanoflakes of molybdenum disulfide (MoS2) as chargestoring elements for use in organic multilevel memory devices (of the printed/flexible non-volatile type) based on organic fieldeffect transistors (OFETs) containing poly(3-hexylthiophene) (P3HT). The metallic MoS2 nanoflakes were exfoliated in 2-methoxyethanol by the lithium intercalation method and were deposited as nano-floating gates between polystyrene and poly(methyl methacrylate), used as bilayered gate dielectrics, by a simple spincoating and low temperature (102 times, and most importantly, quasi-permanent charge-storing characteristics, i.e., a very long retention time (longer than the technological requirement of commercial memory devices (>10 years)). In addition, we successfully developed multilevel memory cells (2 bits per cell) by controlling the gate bias magnitude.
a Heeger Center for Advanced Materials, Gwangju Institute of Science and Technology (GIST), 261 Cheomdan-gwagiro, Buk-gu, Gwangju 500-712, Republic of Korea. E-mail: [email protected] b Radiation Research Division for Industry and Environment, Korea Atomic Energy Research Institute (KAERI), Jeollabuk-do 580-185, Republic of Korea c Department of Energy and Materials Engineering, Dongguk University, 26 Pil-dong, 3 ga, Jung-gu, Seoul 100-715, Republic of Korea. E-mail: [email protected] d Nano Carbon Materials Research Group, Korea Electrotechnology Research Institute (KERI), 12 Bulmosan-ro 10beon-gil, Seongsan-gu, Changwon, Gyeongsangnamdo 642-120, Republic of Korea. E-mail: [email protected] † Electronic supplementary information (ESI) available: The memory characteristics with thickness of MoS2 nanoflakes as nano-floating-gate. See DOI: 10.1039/c4nr03448a
This journal is © The Royal Society of Chemistry 2014
Introduction As was very recently highlighted in Consumer Electronics Show (CES) 2014, future electronics are expected to be advanced enough to enable the fabrication of flexible, foldable, or even stretchable and wearable smart devices. For realizing such devices, most electrical or electronic components, such as transistors and circuitry,1–4 supercapacitors,5 nanogenerators,6 and displays,7,8 are being extensively developed with free form factors. Electrically reprogrammable/erasable non-volatile memory is one of the fundamental parts required to meet the rapidly increasing demands for high-density data storage devices. As suitable candidates for such memory devices, organic nano-floating gate memories (NFGMs) have received significant attention because of their potential for being used as flexible or stretchable charge storage media via cost-effective graphic art printing processes.9–11 In principle, organic NFGMs have the same framework as organic fieldeffect transistors (OFETs), except that in the former, nano sized floating gates (NFGs) are inserted as the charge storage medium in the middle of the gate dielectric layer.9,12,13 For programming and erasing data files, charge carriers are respectively trapped and detrapped in the embedded NFGs by applying an external gate electric field for modulating channel conductance and controlling the threshold voltage (VTh) of OFETs. In comparison with the conventional thin-film floating-gate devices, NFGMs have the advantages of stable charge retention and high data storage density achieved either by controlling the density of conductive nano sites embedded in the dielectric layer or by employing the multibit storage concept through the lateral and vertical stacking of NFGs without any increase in the leakage current.14,15 To date, various materials, including metal nanoparticles,16,17 small molecular organic semiconductors,18 graphene,19,20 and graphene oxide and its reduced form,21 have been used as NFGs in multilayered or composite dielectric films. These previous reports revealed that a proper selection of charge-trapping materials and optimization of their nanostructure (size and distribution) are equally important to
Nanoscale, 2014, 6, 12315–12323 | 12315
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Communication
achieve excellent and reliable non-volatile memory characteristics.18,22,23 Although organic NFGMs with metallic nanoparticles, as commonly used in NFGs, have been successfully fabricated by simple thermal evaporation23 or via precursor routes in nanostructured block copolymers,14,17 several limitations, such as metal nanoparticle deposition in a high vacuum chamber or a long curing process at high temperature, need to be overcome for realizing the practical applications of such NFGMs in printed/flexible electronic devices by costeffective methods. In addition, metal penetration into an organic tunneling layer with a smooth surface during the deposition of metallic nanoparticles makes it difficult to protect an organic semiconductor layer and reduce the operation voltage of the memory device due to the electrical short problem. Moreover, it becomes difficult to realize the vertical stacks of the charge storage layer for enabling high-density and multibit memory cells. Alternatively, two-dimensional (2-D) nanomaterials, such as graphene and reduced graphene oxide, have been considered to be promising candidates for substituting metal nanoparticles or clusters as charge-trapping materials in memory devices19,24–26 since Kim et al. reported organic NFGMs with graphene oxide nanosheets as chargetrapping layers.21 In particular, these 2-D materials used as NFGs in printed and flexible memory devices have the following advantages: (i) high mechanical strength under severe strain conditions, (ii) high optical transparency, and (iii) excellent electrical properties. Another advantage is that after being subjected to liquid exfoliation methods, these materials can be deposited by a simple solution process. Transition metal dichalcogenides (TMDs), such as molybdenum disulfide (MoS2), are emerging 2-D materials because of their unique electrical properties.27 MoS2 can crystallize in a graphite-like layered structure that shows strong anisotropy in chemical, mechanical, thermal, and optical properties.28–30 Further, its band-gap can be tuned from ∼1.29 eV (indirect, bulk MoS2) to 1.9 eV (direct, monolayer MoS2), depending on the number of MoS2 layers. Thus, MoS2 has various electronic and optoelectronic applications, including lithium ion batteries,31 gas sensors,32 photodetectors,33 and interlayer in organic photovoltaics.34,35 Monolayer or few-layered MoS2 exhibits semiconducting properties at room temperature: a band-gap of ∼1.9 eV and high charge carrier mobilities of 200–500 cm2 (V s)−1 along the basal plane.36,37 Choi et al. realized non-volatile memory devices by utilizing mechanically exfoliated MoS2 and graphene as the channel and floating-gate layer, respectively, with an atomically thin hexagonal boron nitride (h-BN) tunneling barrier.25 Notably, semiconducting 2-D TMDs are broadly utilized in the abovementioned applications but are mostly used as an active layer in transistors. Their use as NFG materials in organic flash memory has not been reported so far. In this study, we fabricated highly reliable multilevel organic NFGMs with chemically exfoliated MoS2 nanoflakes constituting a charge-trapping layer. Top-gate/bottom-contact (TG/BC) OFETs were fabricated with regio-regular poly(3-hexylthiophene) (P3HT) and polystyrene (PS)/poly(methyl metha-
12316 | Nanoscale, 2014, 6, 12315–12323
Nanoscale
crylate) (PMMA) as active and bilayered gate dielectric layers, respectively. Further, chemically exfoliated MoS2 nanoflakes in 2-methoxyethanol were inserted at the PS/PMMA interface by a low-temperature solution process. The optimized MoS2-based organic NFGMs showed reversible counter-clockwise hysteresis in transfer curves with an appropriately large memory window of >23 V, programming–reading–erasing–reading cycling endurance of >102 times, and quasi-permanent charge-storing capability, i.e., a very long retention time of >10 years. Moreover, P3HT NFGMs showed reproducible multilevel (2 bits per cell) memory characteristics because of their relatively effective and stable charge trapping behavior in discrete MoS2 NFGs.
Experimental section Chemical exfoliation of MoS2 MoS2 nanoflakes were obtained by the lithium intercalation method carried out in accordance with a procedure reported earlier.34 A small amount (0.3 g) of natural MoS2 crystals obtained from Sigma-Aldrich was dispersed in 3 ml of a 2 M butyl-lithium solution in cyclohexane (Sigma-Aldrich) for 48 h in a flask filled with nitrogen gas. LixMoS2 was retrieved by filtration and washed with hexane (60 ml) to remove excess lithium and organic residues.38 Exfoliation was achieved by ultrasonicating the obtained LixMoS2 slurry in DI-water for 1 h. The extracted mixture was centrifuged and redispersed in 2-methoxyethanol.
Fabrication of organic field-effect transistor (OFET) memory devices A Corning XG glass substrate was used to fabricate the devices. A Ni adhesion layer (4 nm thick) and Au source/drain electrodes (15 nm thick) were patterned on the substrates by conventional photolithography and lift-off processes. The channel width/length (W/L) was 1.0 mm/10 μm. The semiconducting polymer P3HT (Rieke Metal), having a concentration of 10 mg ml−1 in chlorobenzene, was spin-coated onto the as-cleaned substrates containing the BC Au electrodes. Both the first dielectric layer, PS (concentration of 2.5 mg ml−1 in n-butyl acetate (nBA)), and the second dielectric layer, PMMA (concentration of 50 mg ml−1 in 2-ethoxyethanol (2E)), were sequentially spin-coated on the top of the semiconductor films. All the dielectric materials and solvents were purchased from Sigma-Aldrich and used as received. The solutions were filtered through a 0.45 μm polytetrafluoroethylene filter prior to use. The PS film was covered with a sufficient amount of the MoS2 solution, allowed to stand for 30 s, and spin-coated at 2000 rpm for 60 s. Thereafter, MoS2 nanoflakes were dried at 100 °C in ambient conditions. The semiconductor and dielectric films were thermally annealed at 150 °C and 80 °C for 30 min, respectively. The organic NFGM devices were completed after the vapor deposition of Al gate electrodes (∼45 nm thick) through a metal shadow mask.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Nanoscale
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Thin film and device characterization The current–voltage (I–V) characteristics of the OFETs and non-volatile memory devices were measured in a nitrogenfilled glove box using a Keithley 4200 semiconductor characterization system. The thicknesses of the thin films were measured using an XP-1 surface profiling system (Ambios Technology, Inc.). The work function measurement was performed using a Mcallister Kelvin Probe (KP6500). The height of MoS2 nanoflakes was measured using a Digital Instruments Multimode AFM. XPS(AXIS-NOVA) measurement was performed using monochromatized Al Kα under a pressure of 5 × 10−8 Torr. The Raman characterization of MoS2 nanoflakes and bulk flake was performed using a confocal Raman microscope equipped with an excited 632 nm laser (Nanofinder 30, Tokyo Instruments).
Results and discussion Fig. 1a shows the configuration of TG/BC organic NFGMs that consist of P3HT as an organic semiconductor, an aluminum control gate, gold source/drain electrodes, and MoS2 NFGs inserted between bilayered PS/PMMA tunneling and blocking dielectrics. In NFGMs, channel current can be reversibly changed by the trapping/detrapping of charge carriers in the floating gate by the application of an external electric field. The energy diagram of the memory cell in the field-effectinduced carrier transfer state (Fig. 1b) shows that the stored charges during the first operation effectively shield the gate field, thus inducing electrical-gate modulation for the next operation. To generate charge-trapping sites, we chemically deposited exfoliated MoS2 nanoflakes via spin-coating. Liquid exfoliation is one of the most effective methods for preparing
Communication
few-layered MoS2 nanoflakes on a mass production scale.38 The flake size and number of layers can be simply controlled by adjusting the lithiation time with proper stirring carried out using a magnetic stirrer. It should be noted here that the intercalation with alkali metals induces phase changes in some TMDs. In this study, MoS2 nanoflakes were prepared by the intercalation of lithium; therefore, they lost their semiconducting properties due to a transition to the metallic phase.39 The intercalation of Li+ ions provides charge transfer from Li to MoS2 that leads to the transformation of the original crystal phase from a trigonal prismatic (2H: two layers, hexagonal symmetry) phase to an octahedral (1T: one layer, tetragonal symmetry) coordination geometry.28,40 Therefore, Kelvin probe measurement revealed a metallic work function of 4.3 eV instead of a semiconducting band-gap.35 PS and PMMA have relatively larger band-gap (>6 eV) and low electron affinity as insulators, and thus, charge carriers can be stably trapped in MoS2 NFGs of the memory cell, as shown in Fig. 1b.11 Although the barrier height (∼2.8 eV) to be overcome for electron injection from the highest occupied molecular orbital (HOMO) of P3HT to MoS2 nanoflakes is very high, charge transfer could occur via the Fowler–Nordheim (F–N) tunneling by applying a high electric field across the PS tunneling layer.41 Clearly, the reduction in the thickness of the PS tunneling layer is important to decrease the operating voltage of organic NFGMs without any increase in the leakage current.25,42 Although metal nanoparticles deposited by thermal evaporation can achieve a relatively uniform NFG size and distribution under specific deposition conditions, the reduction in PS thickness is limited because of the easy penetration of metal nanoparticles in the polymer film.23 In contrast to thermally deposited metal nanoparticles, MoS2 nanoflakes are free from this problem because of the relatively
Fig. 1 (a) Schematic device configuration of an organic NFGM fabricated using MoS2 floating gate based on TG/BC OFET geometry. The right-hand image illustrates MoS2 nanoflakes between PS and PMMA bilayered dielectrics. (b) Energy band diagram shows the operating mechanism of electrical-gate modulation in the MoS2 NFGM device.
This journal is © The Royal Society of Chemistry 2014
Nanoscale, 2014, 6, 12315–12323 | 12317
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Communication
big lateral size (100–500 nm, vide infra), which is sufficient to stack them onto the tunneling dielectric layer without any intermixing and penetration into the under-layer. Therefore, the operation voltage can be effectively reduced by reducing the PS layer thickness (∼20 nm). The device characteristics of NFGMs strongly depend on the size and distribution of conductive nano-floating gate materials on the tunneling dielectric layer. A relatively uniform distribution of MoS2 nanoflakes on the PS layer can be obtained by a careful selection of the dispersion solvent, e.g., 2-methoxyethanol (2ME). Li-intercalated MoS2 nanoflakes are well-dispersed in 2ME (see Fig. 2a); it provides well-distributed MoS2 nanoflake morphology by spin-coating without the dissolution of the under-laid PS layer because 2ME shows good surface coverage on PS during coating while being perfectly orthogonal to PS. Fig. 2b and 2c show an optical micrograph and atomic force microscopy (AFM) image of spin-coated MoS2 nanoflakes on a PS film. The average thickness of isolated MoS2 flakes and the estimated number of layers, as calculated from the height-mode AFM image, are represented in Fig. 2d and 2e, respectively. Notably, we calculated the number of MoS2 layers divided by Li-intercalated MoS2 monolayer thickness in the range of 1.1–1.3 nm, which is similar to MoS2 monolayer,35,39 due to thickness inhomogeneity in samples and the stacking disorder. The
Nanoscale
average height of deposited MoS2 nanoflakes was ∼3.8 nm, indicating that the majority of nanoflakes consisted of 2 to 8 stacked sheets (Fig. 2e). Fig. 2f shows the lateral sizes of the nanoflakes; it can be seen that MoS2 nanoflakes with different sizes are randomly distributed from tens of nanometers ( ±50 V, as shown in Fig. 4d. The memory window was remarkably increased beyond Vg = ±60 V, and reached up to a maximum value of more than 22.5 V at a programming/erasing voltage (VP/E) of ±80 V. Interestingly, when we applied different layer thicknesses of nanoflakes, the overall memory characteristics became very different, depending on the layer thickness of MoS2 NFGs. As shown in Fig. S1,† devices with thicker nanoflakes showed more enhanced memory behavior under the same VP/E conditions. This indicates that metallic nanoflakes contributed more to electron trapping than semiconducting nanoflakes. This is because monolayer MoS2 was mostly composed of semiconducting 2H-MoS2 with a bigger band-gap (∼1.9 eV),
Nanoscale, 2014, 6, 12315–12323 | 12319
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Communication
Nanoscale
Fig. 4 Schematic electrical characteristics of P3HT-based OFETs and NFGMs. (a) Transfer (Id vs. Vg) and (b) output (Id vs. Vd) curves of P3HT OFETs fabricated using bilayered dielectrics, PS and PMMA. (c) Memory characteristics of organic NFGMs with MoS2 floating gate in PS/PMMA dielectric at different Vg sweep values in the ranges of ±30 and ±80 V. (d) Memory window as a function of program/erase voltage (VP/E) from ±30 to ±80 V, where the maximum window is ∼22.5 V at VP/E = ±80 V.
while thicker flakes had a 1T–2H hybrid phase with a larger metallic 1T phase (Fig. 3a and b). Another scenario is related to the energy levels of trap states in thicker and thinner MoS2 flakes. Bulk and thicker MoS2 have deeper work function and/ or minimum conduction band compared to those of monolayer and few-layered MoS2 nanosheets. Therefore, MoS2 NFGs with multilayers could store transferred electrons more stably than MoS2 monolayer; thus, controlling the layer thickness of MoS2 nanoflakes is similar to tuning the work function of reduced graphene oxide as floating gates in organic transistor memory devices.52 In this study, we obtained the best memory characteristics using flakes with 2 to 8 stacked sheets. On the basis of the abovementioned device characteristics, it can be concluded that MoS2 nanoflakes show stable and electrical-bias-controllable charge storage behavior with an appropriately large memory window. To further improve the beneficial characteristics of MoS2 NFGs used in our memory device, we investigated their multilevel non-volatile memory
12320 | Nanoscale, 2014, 6, 12315–12323
behavior (Fig. 5a–d). For the memory cycling endurance test, consecutively four different Id levels were measured by the application of programming voltage steps of Vg = −80 (fully erased state), +30, +50, and +80 V (fully programmed state), as shown in Fig. 5b. ON current states were controlled by the applied Vg and cycling rate (>100 ms Vg pulse for high charge density) repeatedly. The NFGM devices were sustained during the reprogrammable cycles (write–read–erase–read) more than 102 times, as shown in Fig. 5a. However, programming speed of this device is very slow to be commercially usable. The programming time can be improved by reducing tunneling dielectric layer thickness and lowering charge injection barrier, while decreasing leakage current by increasing charge blocking dielectric layer thickness with a larger energy barrier.13 In addition, the programming speed could also be improved with the increasing the charge carrier mobility of organic semiconductor and the thickness and/or dielectric constant of the insulating layers.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Nanoscale
Communication
Fig. 5 Multilevel operation in P3HT NFGM device using MoS2 nanoflakes floating gate. (a) Programming–reading–erasing cycling endurance and (b) corresponding applied bias condition. (c) Retention characteristics of different current levels (measured Id at Vd = −5 V, Vg = 0 V) after programming/ erasing at Vg = −80, 30, 50, and 80 V. Data retention of currents in ON and OFF states was measured over a time interval of 60 s at Vd = −5 V and Vg = 0 V. (d) Retention time was estimated by the extrapolation of each ON- and OFF-state plot.
Retention characteristics were also independently measured by four different ON-state current values (see Fig. 5c). The multilevel NFGM device showed very stable four-states retention characteristics under a constant reading bias condition at Vd = −5 V and Vg = 0 V (see Fig. 5d). Importantly, for our multibit organic NFGM device, we obtained highly stable charge storage characteristics in MoS2 nanoflakes with a quasi-permanent retention time of more than 10 years, which is a technical requirement for commercialized non-volatile memory. There are viable demands for low cost and high-density data storage in current electronic devices, including emerging wearable smart devices with free shapes; 2-D nanomaterials are expected to facilitate meeting these demands when used as a charge storage medium not only as an active channel.
nanoflakes as a nano-floating gate. The main advantages of using chemically exfoliated MoS2 nanoflakes instead of commonly used metal nanoclusters as charge storage media are low-cost solution processability at a relatively low temperature and reduction in operation voltage by decreasing the tunneling layer thickness without obvious charge loss. Our organic NFGMs showed reliable non-volatile memory characteristics through improved charge confinement properties and trap density of MoS2 nanoflakes. In addition, a cycle endurance of more than 102 times and retention times longer than 10 years were achieved. We believe that organic NFGMs with charge trapping in MoS2 nanoflakes are promising and reliable candidates for the low-cost solution processing of flexible multilevel memory devices.
Conclusions
Acknowledgements
In conclusion, we have fabricated highly stable multilevel (2 bits per cell) non-volatile organic NFGMs by utilizing MoS2
This research was financially supported by the National Research Foundation of Korea (NRF) grant funded by the
This journal is © The Royal Society of Chemistry 2014
Nanoscale, 2014, 6, 12315–12323 | 12321
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Communication
Korea government(MSIP) (no. 2013-059210), a grant (no. 20080062606, CELA-NCRC), the Primary Research Program (14-12N0101-14) of Korea Electrotechnology Research Institute, the Center for Advanced Soft-Electronics funded by the Ministry of Science, ICT and Future Planning as Global Frontier Project (2013M3A6A5073183), and the Dongguk University Research Fund of 2014. We thank the Korea Basic Science Institute (KBSI) for AFM, XPS, and Raman measurements.
References 1 J. A. Rogers, T. Someya and Y. Huang, Science, 2010, 327, 1603–1607. 2 S. H. Chae, W. J. Yu, J. J. Bae, D. L. Duong, D. Perello, H. Y. Jeong, Q. H. Ta, T. H. Ly, Q. A. Vu, M. Yun, X. Duan and Y. H. Lee, Nat. Mater., 2013, 12, 403–409. 3 Y. Hu, C. Warwick, A. Sou, L. Jiang and H. Sirringhaus, J. Mater. Chem. C, 2014, 2, 1260–1263. 4 S. G. Bucella, G. Nava, K. C. Vishunubhatla and M. Caironi, Org. Electron., 2013, 14, 2249–2256. 5 Q. Wu, Y. Xu, Z. Yao, A. Liu and G. Shi, ACS Nano, 2010, 4, 1963–1970. 6 S. Lee, S.-H. Bae, L. Lin, Y. Yang, C. Park, S.-W. Kim, S. N. Cha, H. Kim, Y. J. Park and Z. L. Wang, Adv. Funct. Mater., 2013, 23, 2445–2449. 7 T. Sekitani, H. Nakajima, H. Maeda, T. Fukushima, T. Aida, K. Hata and T. Someya, Nat. Mater., 2009, 8, 494–499. 8 S. Steudel, K. Myny, S. Schols, P. Vicca, S. Smout, A. Tripathi, B. van der Putten, J.-L. van der Steen, M. van Neer, F. Schütze, O. R. Hild, E. van Veenendaal, P. van Lieshout, M. van Mil, J. Genoe, G. Gelinck and P. Heremans, Org. Electron., 2012, 13, 1729–1735. 9 T. Sekitani, T. Yokota, U. Zschieschang, H. Klauk, S. Bauer, K. Takeuchi, M. Takamiya, T. Sakurai and T. Someya, Science, 2009, 326, 1516–1519. 10 S.-T. Han, Y. Zhou and V. A. L. Roy, Adv. Mater., 2013, 25, 5425–5449. 11 K.-J. Baeg, D. Khim, J. Kim, B.-D. Yang, M. Kang, S.-W. Jung, I.-K. You, D.-Y. Kim and Y.-Y. Noh, Adv. Funct. Mater., 2012, 22, 2915–2926. 12 W. D. Brown and J. E. Brewer, Nonvolatile Semiconductor Memory Technology: a Comprehensive Guide to Understanding and to using NVSM Devices, IEEE Press, New York, 1998. 13 K.-J. Baeg, Y.-Y. Noh, H. Sirringhaus and D. Y. Kim, Adv. Funct. Mater., 2010, 20, 224–230. 14 J. S. Lee, Y. M. Kim, J. H. Kwon, J. S. Sim, H. Shin, B. H. Sohn and Q. Jia, Adv. Mater., 2011, 23, 2064–2068. 15 T. Z. Lu, M. Alexe, R. Scholz, V. Talelaev and M. Zacharias, Appl. Phys. Lett., 2005, 87, 202110. 16 S. Lee, J. Lee, H. Lee, Y. J. Yuk, M. Kim, H. Moon, J. Seo, Y. Park, J. Y. Park, S. H. Ko and S. Yoo, Org. Electron., 2013, 14, 3260–3266. 17 Q. Wei, Y. Lin, E. R. Anderson, A. L. Briseno, S. P. Gido and J. J. Watkins, ACS Nano, 2012, 6, 1188–1194.
12322 | Nanoscale, 2014, 6, 12315–12323
Nanoscale
18 S. Paydavosi, H. Abdu, G. J. Supran and V. Bulovic, IEEE Trans. Nanotechnol., 2011, 10, 594–599. 19 X. Wang, W. Xie and J.-B. Xu, Adv. Mater., 2014, 26, 5496– 5503. 20 S. Bertolazzi, D. Krasnozhon and A. Kis, ACS Nano, 2013, 7, 3246–3252. 21 T.-W. Kim, Y. Gao, O. Acton, H.-L. Yip, H. Ma, H. Chen and A. K.-Y. Jen, Appl. Phys. Lett., 2010, 97, 023310. 22 T.-C. Chang, F.-Y. Jian, S.-C. Chen and Y.-T. Tsai, Mater. Today, 2011, 14, 608–615. 23 M. Kang, K. J. Baeg, D. Khim, Y. Y. Noh and D. Y. Kim, Adv. Funct. Mater., 2013, 23, 3503–3512. 24 R. Yang, C. Zhu, J. Meng, Z. Huo, M. Cheng, D. Liu, W. Yang, D. Shi, M. Liu and G. Zhang, Sci. Rep., 2013, 3, 2126. 25 M. S. Choi, G. H. Lee, Y. J. Yu, D. Y. Lee, S. H. Lee, P. Kim, J. Hone and W. J. Yoo, Nat. Commun., 2013, 4, 1624. 26 Y. Ji, M. Choe, B. Cho, S. Song, J. Yoon, H. C. Ko and T. Lee, Nanotechnology, 2012, 23, 105202. 27 B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti and A. Kis, Nat. Nanotechnol., 2011, 6, 147–150. 28 Q. Wang, K. Kalantar-Zadeh, A. Kis, J. Coleman and M. Strano, Nat. Nanotechnol., 2012, 7, 699–712. 29 S. Bertolazzi, J. Brivio and A. Kis, ACS Nano, 2011, 5, 9703– 9709. 30 V. Stengl and J. Henych, Nanoscale, 2013, 5, 3387– 3394. 31 K. Chang and W. Chen, ACS Nano, 2011, 5, 4720–4728. 32 H. Li, Z. Yin, Q. He, H. Li, X. Huang, G. Lu, D. Fam, A. Tok, Q. Zhang and H. Zhang, Small, 2012, 8, 63–67. 33 O. Lopez-Sanchez, D. Lembke, M. Kayci, A. Radenovic and A. Kis, Nat. Nanotechnol., 2013, 8, 497–501. 34 G. Xing, C. Wei, L. Hai, W. Zhongwei, Z. Zhiyuan, L. ShuitTong, Z. Hua and S. Baoquan, Adv. Energy Mater., 2013, 3, 1262–1268. 35 J.-M. Yun, Y.-J. Noh, J.-S. Yeo, Y.-J. Go, S.-I. Na, H.-G. Jeong, J. Kim, S. Lee, S.-S. Kim, H. Y. Koo, T.-W. Kim and D.-Y. Kim, J. Mater. Chem. C, 2013, 1, 3777–3783. 36 H. Wang, L. Yu, Y.-H. Lee, Y. Shi, A. Hsu, M. L. Chin, L.-J. Li, M. Dubey, J. Kong and T. Palacios, Nano Lett., 2012, 12, 4674–4680. 37 H. Liu and P. D. Ye, IEEE Electron Device Lett., 2012, 33, 546–548. 38 M. Chhowalla, H. Shin, G. Eda, L.-J. Li, K. Loh and H. Zhang, Nat. Chem., 2013, 5, 263–275. 39 G. Eda, H. Yamaguchi, D. Voiry, T. Fujita, M. Chen and M. Chhowalla, Nano Lett., 2011, 11, 5111–5116. 40 Y. Wang, J. Ou, S. Balendhran, A. Chrimes, M. Mortazavi, D. Yao, M. Field, K. Latham, V. Bansal, J. Friend, S. Zhuiykov, N. Medhekar, M. Strano and K. KalantarZadeh, ACS Nano, 2013, 7, 10083–10093. 41 Z. Liu, C. Lee, V. Narayanan, G. Pei and E. C. Kan, IEEE Trans. Electron Devices, 2002, 49, 1606–1613. 42 G.-H. Lee, Y.-J. Yu, C. Lee, C. Dean, K. L. Shepard, P. Kim and J. Hone, Appl. Phys. Lett., 2011, 99, 243114.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Nanoscale
43 C. A. Papageorgopoulos and W. Jaegermann, Surf. Sci., 1995, 338, 83–93. 44 T. Hu, R. Li and J. Dong, J. Chem. Phys., 2013, 139, 174702. 45 H. Li, Q. Zhang, C. C. R. Yap, B. K. Tay, T. H. T. Edwin, A. Olivier and D. Baillargeat, Adv. Funct. Mater., 2012, 22, 1385–1390. 46 G. Eda, T. Fujita, H. Yamaguchi, D. Voiry, M. Chen and M. Chhowalla, ACS Nano, 2012, 6, 7311–7317. 47 S. Balendhran, J. Ou, M. Bhaskaran, S. Sriram, S. Ippolito, Z. Vasic, E. Kats, S. Bhargava, S. Zhuiykov and K. KalantarZadeh, Nanoscale, 2012, 4, 461–466.
This journal is © The Royal Society of Chemistry 2014
Communication
48 G. L. Frey, R. Tenne, M. J. Matthews, M. S. Dresselhaus and G. Dresselhaus, Phys. Rev. B: Condens. Matter, 1999, 60, 2883–2892. 49 S. Sandoval, D. Yang, R. Frindt and J. Irwin, Phys. Rev. B: Condens. Matter, 1991, 44, 3955–3962. 50 A. Facchetti, M. H. Yoon and T. J. Marks, Adv. Mater., 2005, 17, 1705–1725. 51 W. L. Leong, P. S. Lee, A. Lohani, Y. M. Lam, T. Chen, S. Zhang, A. Dodabalapur and S. G. Mhaisalkar, Adv. Mater., 2008, 20, 2325–2331. 52 S.-T. Han, Y. Zhou, Q. D. Yang, L. Zhou, L.-B. Huang, Y. Yan, C.-S. Lee and V. A. L. Roy, ACS Nano, 2014, 8, 1923–1931.
Nanoscale, 2014, 6, 12315–12323 | 12323
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
COMMUNICATION
Cite this: Nanoscale, 2014, 6, 12315 Received 20th June 2014, Accepted 7th August 2014 DOI: 10.1039/c4nr03448a www.rsc.org/nanoscale
View Journal | View Issue
Stable charge storing in two-dimensional MoS2 nanoflake floating gates for multilevel organic flash memory† Minji Kang,a Yeong-A. Kim,a Jin-Mun Yun,b Dongyoon Khim,a Jihong Kim,a Yong-Young Noh,*c Kang-Jun Baeg*d and Dong-Yu Kim*a
In this study, we investigated chemically exfoliated two-dimensional (2-D) nanoflakes of molybdenum disulfide (MoS2) as chargestoring elements for use in organic multilevel memory devices (of the printed/flexible non-volatile type) based on organic fieldeffect transistors (OFETs) containing poly(3-hexylthiophene) (P3HT). The metallic MoS2 nanoflakes were exfoliated in 2-methoxyethanol by the lithium intercalation method and were deposited as nano-floating gates between polystyrene and poly(methyl methacrylate), used as bilayered gate dielectrics, by a simple spincoating and low temperature (102 times, and most importantly, quasi-permanent charge-storing characteristics, i.e., a very long retention time (longer than the technological requirement of commercial memory devices (>10 years)). In addition, we successfully developed multilevel memory cells (2 bits per cell) by controlling the gate bias magnitude.
a Heeger Center for Advanced Materials, Gwangju Institute of Science and Technology (GIST), 261 Cheomdan-gwagiro, Buk-gu, Gwangju 500-712, Republic of Korea. E-mail: [email protected] b Radiation Research Division for Industry and Environment, Korea Atomic Energy Research Institute (KAERI), Jeollabuk-do 580-185, Republic of Korea c Department of Energy and Materials Engineering, Dongguk University, 26 Pil-dong, 3 ga, Jung-gu, Seoul 100-715, Republic of Korea. E-mail: [email protected] d Nano Carbon Materials Research Group, Korea Electrotechnology Research Institute (KERI), 12 Bulmosan-ro 10beon-gil, Seongsan-gu, Changwon, Gyeongsangnamdo 642-120, Republic of Korea. E-mail: [email protected] † Electronic supplementary information (ESI) available: The memory characteristics with thickness of MoS2 nanoflakes as nano-floating-gate. See DOI: 10.1039/c4nr03448a
This journal is © The Royal Society of Chemistry 2014
Introduction As was very recently highlighted in Consumer Electronics Show (CES) 2014, future electronics are expected to be advanced enough to enable the fabrication of flexible, foldable, or even stretchable and wearable smart devices. For realizing such devices, most electrical or electronic components, such as transistors and circuitry,1–4 supercapacitors,5 nanogenerators,6 and displays,7,8 are being extensively developed with free form factors. Electrically reprogrammable/erasable non-volatile memory is one of the fundamental parts required to meet the rapidly increasing demands for high-density data storage devices. As suitable candidates for such memory devices, organic nano-floating gate memories (NFGMs) have received significant attention because of their potential for being used as flexible or stretchable charge storage media via cost-effective graphic art printing processes.9–11 In principle, organic NFGMs have the same framework as organic fieldeffect transistors (OFETs), except that in the former, nano sized floating gates (NFGs) are inserted as the charge storage medium in the middle of the gate dielectric layer.9,12,13 For programming and erasing data files, charge carriers are respectively trapped and detrapped in the embedded NFGs by applying an external gate electric field for modulating channel conductance and controlling the threshold voltage (VTh) of OFETs. In comparison with the conventional thin-film floating-gate devices, NFGMs have the advantages of stable charge retention and high data storage density achieved either by controlling the density of conductive nano sites embedded in the dielectric layer or by employing the multibit storage concept through the lateral and vertical stacking of NFGs without any increase in the leakage current.14,15 To date, various materials, including metal nanoparticles,16,17 small molecular organic semiconductors,18 graphene,19,20 and graphene oxide and its reduced form,21 have been used as NFGs in multilayered or composite dielectric films. These previous reports revealed that a proper selection of charge-trapping materials and optimization of their nanostructure (size and distribution) are equally important to
Nanoscale, 2014, 6, 12315–12323 | 12315
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Communication
achieve excellent and reliable non-volatile memory characteristics.18,22,23 Although organic NFGMs with metallic nanoparticles, as commonly used in NFGs, have been successfully fabricated by simple thermal evaporation23 or via precursor routes in nanostructured block copolymers,14,17 several limitations, such as metal nanoparticle deposition in a high vacuum chamber or a long curing process at high temperature, need to be overcome for realizing the practical applications of such NFGMs in printed/flexible electronic devices by costeffective methods. In addition, metal penetration into an organic tunneling layer with a smooth surface during the deposition of metallic nanoparticles makes it difficult to protect an organic semiconductor layer and reduce the operation voltage of the memory device due to the electrical short problem. Moreover, it becomes difficult to realize the vertical stacks of the charge storage layer for enabling high-density and multibit memory cells. Alternatively, two-dimensional (2-D) nanomaterials, such as graphene and reduced graphene oxide, have been considered to be promising candidates for substituting metal nanoparticles or clusters as charge-trapping materials in memory devices19,24–26 since Kim et al. reported organic NFGMs with graphene oxide nanosheets as chargetrapping layers.21 In particular, these 2-D materials used as NFGs in printed and flexible memory devices have the following advantages: (i) high mechanical strength under severe strain conditions, (ii) high optical transparency, and (iii) excellent electrical properties. Another advantage is that after being subjected to liquid exfoliation methods, these materials can be deposited by a simple solution process. Transition metal dichalcogenides (TMDs), such as molybdenum disulfide (MoS2), are emerging 2-D materials because of their unique electrical properties.27 MoS2 can crystallize in a graphite-like layered structure that shows strong anisotropy in chemical, mechanical, thermal, and optical properties.28–30 Further, its band-gap can be tuned from ∼1.29 eV (indirect, bulk MoS2) to 1.9 eV (direct, monolayer MoS2), depending on the number of MoS2 layers. Thus, MoS2 has various electronic and optoelectronic applications, including lithium ion batteries,31 gas sensors,32 photodetectors,33 and interlayer in organic photovoltaics.34,35 Monolayer or few-layered MoS2 exhibits semiconducting properties at room temperature: a band-gap of ∼1.9 eV and high charge carrier mobilities of 200–500 cm2 (V s)−1 along the basal plane.36,37 Choi et al. realized non-volatile memory devices by utilizing mechanically exfoliated MoS2 and graphene as the channel and floating-gate layer, respectively, with an atomically thin hexagonal boron nitride (h-BN) tunneling barrier.25 Notably, semiconducting 2-D TMDs are broadly utilized in the abovementioned applications but are mostly used as an active layer in transistors. Their use as NFG materials in organic flash memory has not been reported so far. In this study, we fabricated highly reliable multilevel organic NFGMs with chemically exfoliated MoS2 nanoflakes constituting a charge-trapping layer. Top-gate/bottom-contact (TG/BC) OFETs were fabricated with regio-regular poly(3-hexylthiophene) (P3HT) and polystyrene (PS)/poly(methyl metha-
12316 | Nanoscale, 2014, 6, 12315–12323
Nanoscale
crylate) (PMMA) as active and bilayered gate dielectric layers, respectively. Further, chemically exfoliated MoS2 nanoflakes in 2-methoxyethanol were inserted at the PS/PMMA interface by a low-temperature solution process. The optimized MoS2-based organic NFGMs showed reversible counter-clockwise hysteresis in transfer curves with an appropriately large memory window of >23 V, programming–reading–erasing–reading cycling endurance of >102 times, and quasi-permanent charge-storing capability, i.e., a very long retention time of >10 years. Moreover, P3HT NFGMs showed reproducible multilevel (2 bits per cell) memory characteristics because of their relatively effective and stable charge trapping behavior in discrete MoS2 NFGs.
Experimental section Chemical exfoliation of MoS2 MoS2 nanoflakes were obtained by the lithium intercalation method carried out in accordance with a procedure reported earlier.34 A small amount (0.3 g) of natural MoS2 crystals obtained from Sigma-Aldrich was dispersed in 3 ml of a 2 M butyl-lithium solution in cyclohexane (Sigma-Aldrich) for 48 h in a flask filled with nitrogen gas. LixMoS2 was retrieved by filtration and washed with hexane (60 ml) to remove excess lithium and organic residues.38 Exfoliation was achieved by ultrasonicating the obtained LixMoS2 slurry in DI-water for 1 h. The extracted mixture was centrifuged and redispersed in 2-methoxyethanol.
Fabrication of organic field-effect transistor (OFET) memory devices A Corning XG glass substrate was used to fabricate the devices. A Ni adhesion layer (4 nm thick) and Au source/drain electrodes (15 nm thick) were patterned on the substrates by conventional photolithography and lift-off processes. The channel width/length (W/L) was 1.0 mm/10 μm. The semiconducting polymer P3HT (Rieke Metal), having a concentration of 10 mg ml−1 in chlorobenzene, was spin-coated onto the as-cleaned substrates containing the BC Au electrodes. Both the first dielectric layer, PS (concentration of 2.5 mg ml−1 in n-butyl acetate (nBA)), and the second dielectric layer, PMMA (concentration of 50 mg ml−1 in 2-ethoxyethanol (2E)), were sequentially spin-coated on the top of the semiconductor films. All the dielectric materials and solvents were purchased from Sigma-Aldrich and used as received. The solutions were filtered through a 0.45 μm polytetrafluoroethylene filter prior to use. The PS film was covered with a sufficient amount of the MoS2 solution, allowed to stand for 30 s, and spin-coated at 2000 rpm for 60 s. Thereafter, MoS2 nanoflakes were dried at 100 °C in ambient conditions. The semiconductor and dielectric films were thermally annealed at 150 °C and 80 °C for 30 min, respectively. The organic NFGM devices were completed after the vapor deposition of Al gate electrodes (∼45 nm thick) through a metal shadow mask.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Nanoscale
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Thin film and device characterization The current–voltage (I–V) characteristics of the OFETs and non-volatile memory devices were measured in a nitrogenfilled glove box using a Keithley 4200 semiconductor characterization system. The thicknesses of the thin films were measured using an XP-1 surface profiling system (Ambios Technology, Inc.). The work function measurement was performed using a Mcallister Kelvin Probe (KP6500). The height of MoS2 nanoflakes was measured using a Digital Instruments Multimode AFM. XPS(AXIS-NOVA) measurement was performed using monochromatized Al Kα under a pressure of 5 × 10−8 Torr. The Raman characterization of MoS2 nanoflakes and bulk flake was performed using a confocal Raman microscope equipped with an excited 632 nm laser (Nanofinder 30, Tokyo Instruments).
Results and discussion Fig. 1a shows the configuration of TG/BC organic NFGMs that consist of P3HT as an organic semiconductor, an aluminum control gate, gold source/drain electrodes, and MoS2 NFGs inserted between bilayered PS/PMMA tunneling and blocking dielectrics. In NFGMs, channel current can be reversibly changed by the trapping/detrapping of charge carriers in the floating gate by the application of an external electric field. The energy diagram of the memory cell in the field-effectinduced carrier transfer state (Fig. 1b) shows that the stored charges during the first operation effectively shield the gate field, thus inducing electrical-gate modulation for the next operation. To generate charge-trapping sites, we chemically deposited exfoliated MoS2 nanoflakes via spin-coating. Liquid exfoliation is one of the most effective methods for preparing
Communication
few-layered MoS2 nanoflakes on a mass production scale.38 The flake size and number of layers can be simply controlled by adjusting the lithiation time with proper stirring carried out using a magnetic stirrer. It should be noted here that the intercalation with alkali metals induces phase changes in some TMDs. In this study, MoS2 nanoflakes were prepared by the intercalation of lithium; therefore, they lost their semiconducting properties due to a transition to the metallic phase.39 The intercalation of Li+ ions provides charge transfer from Li to MoS2 that leads to the transformation of the original crystal phase from a trigonal prismatic (2H: two layers, hexagonal symmetry) phase to an octahedral (1T: one layer, tetragonal symmetry) coordination geometry.28,40 Therefore, Kelvin probe measurement revealed a metallic work function of 4.3 eV instead of a semiconducting band-gap.35 PS and PMMA have relatively larger band-gap (>6 eV) and low electron affinity as insulators, and thus, charge carriers can be stably trapped in MoS2 NFGs of the memory cell, as shown in Fig. 1b.11 Although the barrier height (∼2.8 eV) to be overcome for electron injection from the highest occupied molecular orbital (HOMO) of P3HT to MoS2 nanoflakes is very high, charge transfer could occur via the Fowler–Nordheim (F–N) tunneling by applying a high electric field across the PS tunneling layer.41 Clearly, the reduction in the thickness of the PS tunneling layer is important to decrease the operating voltage of organic NFGMs without any increase in the leakage current.25,42 Although metal nanoparticles deposited by thermal evaporation can achieve a relatively uniform NFG size and distribution under specific deposition conditions, the reduction in PS thickness is limited because of the easy penetration of metal nanoparticles in the polymer film.23 In contrast to thermally deposited metal nanoparticles, MoS2 nanoflakes are free from this problem because of the relatively
Fig. 1 (a) Schematic device configuration of an organic NFGM fabricated using MoS2 floating gate based on TG/BC OFET geometry. The right-hand image illustrates MoS2 nanoflakes between PS and PMMA bilayered dielectrics. (b) Energy band diagram shows the operating mechanism of electrical-gate modulation in the MoS2 NFGM device.
This journal is © The Royal Society of Chemistry 2014
Nanoscale, 2014, 6, 12315–12323 | 12317
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Communication
big lateral size (100–500 nm, vide infra), which is sufficient to stack them onto the tunneling dielectric layer without any intermixing and penetration into the under-layer. Therefore, the operation voltage can be effectively reduced by reducing the PS layer thickness (∼20 nm). The device characteristics of NFGMs strongly depend on the size and distribution of conductive nano-floating gate materials on the tunneling dielectric layer. A relatively uniform distribution of MoS2 nanoflakes on the PS layer can be obtained by a careful selection of the dispersion solvent, e.g., 2-methoxyethanol (2ME). Li-intercalated MoS2 nanoflakes are well-dispersed in 2ME (see Fig. 2a); it provides well-distributed MoS2 nanoflake morphology by spin-coating without the dissolution of the under-laid PS layer because 2ME shows good surface coverage on PS during coating while being perfectly orthogonal to PS. Fig. 2b and 2c show an optical micrograph and atomic force microscopy (AFM) image of spin-coated MoS2 nanoflakes on a PS film. The average thickness of isolated MoS2 flakes and the estimated number of layers, as calculated from the height-mode AFM image, are represented in Fig. 2d and 2e, respectively. Notably, we calculated the number of MoS2 layers divided by Li-intercalated MoS2 monolayer thickness in the range of 1.1–1.3 nm, which is similar to MoS2 monolayer,35,39 due to thickness inhomogeneity in samples and the stacking disorder. The
Nanoscale
average height of deposited MoS2 nanoflakes was ∼3.8 nm, indicating that the majority of nanoflakes consisted of 2 to 8 stacked sheets (Fig. 2e). Fig. 2f shows the lateral sizes of the nanoflakes; it can be seen that MoS2 nanoflakes with different sizes are randomly distributed from tens of nanometers ( ±50 V, as shown in Fig. 4d. The memory window was remarkably increased beyond Vg = ±60 V, and reached up to a maximum value of more than 22.5 V at a programming/erasing voltage (VP/E) of ±80 V. Interestingly, when we applied different layer thicknesses of nanoflakes, the overall memory characteristics became very different, depending on the layer thickness of MoS2 NFGs. As shown in Fig. S1,† devices with thicker nanoflakes showed more enhanced memory behavior under the same VP/E conditions. This indicates that metallic nanoflakes contributed more to electron trapping than semiconducting nanoflakes. This is because monolayer MoS2 was mostly composed of semiconducting 2H-MoS2 with a bigger band-gap (∼1.9 eV),
Nanoscale, 2014, 6, 12315–12323 | 12319
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Communication
Nanoscale
Fig. 4 Schematic electrical characteristics of P3HT-based OFETs and NFGMs. (a) Transfer (Id vs. Vg) and (b) output (Id vs. Vd) curves of P3HT OFETs fabricated using bilayered dielectrics, PS and PMMA. (c) Memory characteristics of organic NFGMs with MoS2 floating gate in PS/PMMA dielectric at different Vg sweep values in the ranges of ±30 and ±80 V. (d) Memory window as a function of program/erase voltage (VP/E) from ±30 to ±80 V, where the maximum window is ∼22.5 V at VP/E = ±80 V.
while thicker flakes had a 1T–2H hybrid phase with a larger metallic 1T phase (Fig. 3a and b). Another scenario is related to the energy levels of trap states in thicker and thinner MoS2 flakes. Bulk and thicker MoS2 have deeper work function and/ or minimum conduction band compared to those of monolayer and few-layered MoS2 nanosheets. Therefore, MoS2 NFGs with multilayers could store transferred electrons more stably than MoS2 monolayer; thus, controlling the layer thickness of MoS2 nanoflakes is similar to tuning the work function of reduced graphene oxide as floating gates in organic transistor memory devices.52 In this study, we obtained the best memory characteristics using flakes with 2 to 8 stacked sheets. On the basis of the abovementioned device characteristics, it can be concluded that MoS2 nanoflakes show stable and electrical-bias-controllable charge storage behavior with an appropriately large memory window. To further improve the beneficial characteristics of MoS2 NFGs used in our memory device, we investigated their multilevel non-volatile memory
12320 | Nanoscale, 2014, 6, 12315–12323
behavior (Fig. 5a–d). For the memory cycling endurance test, consecutively four different Id levels were measured by the application of programming voltage steps of Vg = −80 (fully erased state), +30, +50, and +80 V (fully programmed state), as shown in Fig. 5b. ON current states were controlled by the applied Vg and cycling rate (>100 ms Vg pulse for high charge density) repeatedly. The NFGM devices were sustained during the reprogrammable cycles (write–read–erase–read) more than 102 times, as shown in Fig. 5a. However, programming speed of this device is very slow to be commercially usable. The programming time can be improved by reducing tunneling dielectric layer thickness and lowering charge injection barrier, while decreasing leakage current by increasing charge blocking dielectric layer thickness with a larger energy barrier.13 In addition, the programming speed could also be improved with the increasing the charge carrier mobility of organic semiconductor and the thickness and/or dielectric constant of the insulating layers.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Nanoscale
Communication
Fig. 5 Multilevel operation in P3HT NFGM device using MoS2 nanoflakes floating gate. (a) Programming–reading–erasing cycling endurance and (b) corresponding applied bias condition. (c) Retention characteristics of different current levels (measured Id at Vd = −5 V, Vg = 0 V) after programming/ erasing at Vg = −80, 30, 50, and 80 V. Data retention of currents in ON and OFF states was measured over a time interval of 60 s at Vd = −5 V and Vg = 0 V. (d) Retention time was estimated by the extrapolation of each ON- and OFF-state plot.
Retention characteristics were also independently measured by four different ON-state current values (see Fig. 5c). The multilevel NFGM device showed very stable four-states retention characteristics under a constant reading bias condition at Vd = −5 V and Vg = 0 V (see Fig. 5d). Importantly, for our multibit organic NFGM device, we obtained highly stable charge storage characteristics in MoS2 nanoflakes with a quasi-permanent retention time of more than 10 years, which is a technical requirement for commercialized non-volatile memory. There are viable demands for low cost and high-density data storage in current electronic devices, including emerging wearable smart devices with free shapes; 2-D nanomaterials are expected to facilitate meeting these demands when used as a charge storage medium not only as an active channel.
nanoflakes as a nano-floating gate. The main advantages of using chemically exfoliated MoS2 nanoflakes instead of commonly used metal nanoclusters as charge storage media are low-cost solution processability at a relatively low temperature and reduction in operation voltage by decreasing the tunneling layer thickness without obvious charge loss. Our organic NFGMs showed reliable non-volatile memory characteristics through improved charge confinement properties and trap density of MoS2 nanoflakes. In addition, a cycle endurance of more than 102 times and retention times longer than 10 years were achieved. We believe that organic NFGMs with charge trapping in MoS2 nanoflakes are promising and reliable candidates for the low-cost solution processing of flexible multilevel memory devices.
Conclusions
Acknowledgements
In conclusion, we have fabricated highly stable multilevel (2 bits per cell) non-volatile organic NFGMs by utilizing MoS2
This research was financially supported by the National Research Foundation of Korea (NRF) grant funded by the
This journal is © The Royal Society of Chemistry 2014
Nanoscale, 2014, 6, 12315–12323 | 12321
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Communication
Korea government(MSIP) (no. 2013-059210), a grant (no. 20080062606, CELA-NCRC), the Primary Research Program (14-12N0101-14) of Korea Electrotechnology Research Institute, the Center for Advanced Soft-Electronics funded by the Ministry of Science, ICT and Future Planning as Global Frontier Project (2013M3A6A5073183), and the Dongguk University Research Fund of 2014. We thank the Korea Basic Science Institute (KBSI) for AFM, XPS, and Raman measurements.
References 1 J. A. Rogers, T. Someya and Y. Huang, Science, 2010, 327, 1603–1607. 2 S. H. Chae, W. J. Yu, J. J. Bae, D. L. Duong, D. Perello, H. Y. Jeong, Q. H. Ta, T. H. Ly, Q. A. Vu, M. Yun, X. Duan and Y. H. Lee, Nat. Mater., 2013, 12, 403–409. 3 Y. Hu, C. Warwick, A. Sou, L. Jiang and H. Sirringhaus, J. Mater. Chem. C, 2014, 2, 1260–1263. 4 S. G. Bucella, G. Nava, K. C. Vishunubhatla and M. Caironi, Org. Electron., 2013, 14, 2249–2256. 5 Q. Wu, Y. Xu, Z. Yao, A. Liu and G. Shi, ACS Nano, 2010, 4, 1963–1970. 6 S. Lee, S.-H. Bae, L. Lin, Y. Yang, C. Park, S.-W. Kim, S. N. Cha, H. Kim, Y. J. Park and Z. L. Wang, Adv. Funct. Mater., 2013, 23, 2445–2449. 7 T. Sekitani, H. Nakajima, H. Maeda, T. Fukushima, T. Aida, K. Hata and T. Someya, Nat. Mater., 2009, 8, 494–499. 8 S. Steudel, K. Myny, S. Schols, P. Vicca, S. Smout, A. Tripathi, B. van der Putten, J.-L. van der Steen, M. van Neer, F. Schütze, O. R. Hild, E. van Veenendaal, P. van Lieshout, M. van Mil, J. Genoe, G. Gelinck and P. Heremans, Org. Electron., 2012, 13, 1729–1735. 9 T. Sekitani, T. Yokota, U. Zschieschang, H. Klauk, S. Bauer, K. Takeuchi, M. Takamiya, T. Sakurai and T. Someya, Science, 2009, 326, 1516–1519. 10 S.-T. Han, Y. Zhou and V. A. L. Roy, Adv. Mater., 2013, 25, 5425–5449. 11 K.-J. Baeg, D. Khim, J. Kim, B.-D. Yang, M. Kang, S.-W. Jung, I.-K. You, D.-Y. Kim and Y.-Y. Noh, Adv. Funct. Mater., 2012, 22, 2915–2926. 12 W. D. Brown and J. E. Brewer, Nonvolatile Semiconductor Memory Technology: a Comprehensive Guide to Understanding and to using NVSM Devices, IEEE Press, New York, 1998. 13 K.-J. Baeg, Y.-Y. Noh, H. Sirringhaus and D. Y. Kim, Adv. Funct. Mater., 2010, 20, 224–230. 14 J. S. Lee, Y. M. Kim, J. H. Kwon, J. S. Sim, H. Shin, B. H. Sohn and Q. Jia, Adv. Mater., 2011, 23, 2064–2068. 15 T. Z. Lu, M. Alexe, R. Scholz, V. Talelaev and M. Zacharias, Appl. Phys. Lett., 2005, 87, 202110. 16 S. Lee, J. Lee, H. Lee, Y. J. Yuk, M. Kim, H. Moon, J. Seo, Y. Park, J. Y. Park, S. H. Ko and S. Yoo, Org. Electron., 2013, 14, 3260–3266. 17 Q. Wei, Y. Lin, E. R. Anderson, A. L. Briseno, S. P. Gido and J. J. Watkins, ACS Nano, 2012, 6, 1188–1194.
12322 | Nanoscale, 2014, 6, 12315–12323
Nanoscale
18 S. Paydavosi, H. Abdu, G. J. Supran and V. Bulovic, IEEE Trans. Nanotechnol., 2011, 10, 594–599. 19 X. Wang, W. Xie and J.-B. Xu, Adv. Mater., 2014, 26, 5496– 5503. 20 S. Bertolazzi, D. Krasnozhon and A. Kis, ACS Nano, 2013, 7, 3246–3252. 21 T.-W. Kim, Y. Gao, O. Acton, H.-L. Yip, H. Ma, H. Chen and A. K.-Y. Jen, Appl. Phys. Lett., 2010, 97, 023310. 22 T.-C. Chang, F.-Y. Jian, S.-C. Chen and Y.-T. Tsai, Mater. Today, 2011, 14, 608–615. 23 M. Kang, K. J. Baeg, D. Khim, Y. Y. Noh and D. Y. Kim, Adv. Funct. Mater., 2013, 23, 3503–3512. 24 R. Yang, C. Zhu, J. Meng, Z. Huo, M. Cheng, D. Liu, W. Yang, D. Shi, M. Liu and G. Zhang, Sci. Rep., 2013, 3, 2126. 25 M. S. Choi, G. H. Lee, Y. J. Yu, D. Y. Lee, S. H. Lee, P. Kim, J. Hone and W. J. Yoo, Nat. Commun., 2013, 4, 1624. 26 Y. Ji, M. Choe, B. Cho, S. Song, J. Yoon, H. C. Ko and T. Lee, Nanotechnology, 2012, 23, 105202. 27 B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti and A. Kis, Nat. Nanotechnol., 2011, 6, 147–150. 28 Q. Wang, K. Kalantar-Zadeh, A. Kis, J. Coleman and M. Strano, Nat. Nanotechnol., 2012, 7, 699–712. 29 S. Bertolazzi, J. Brivio and A. Kis, ACS Nano, 2011, 5, 9703– 9709. 30 V. Stengl and J. Henych, Nanoscale, 2013, 5, 3387– 3394. 31 K. Chang and W. Chen, ACS Nano, 2011, 5, 4720–4728. 32 H. Li, Z. Yin, Q. He, H. Li, X. Huang, G. Lu, D. Fam, A. Tok, Q. Zhang and H. Zhang, Small, 2012, 8, 63–67. 33 O. Lopez-Sanchez, D. Lembke, M. Kayci, A. Radenovic and A. Kis, Nat. Nanotechnol., 2013, 8, 497–501. 34 G. Xing, C. Wei, L. Hai, W. Zhongwei, Z. Zhiyuan, L. ShuitTong, Z. Hua and S. Baoquan, Adv. Energy Mater., 2013, 3, 1262–1268. 35 J.-M. Yun, Y.-J. Noh, J.-S. Yeo, Y.-J. Go, S.-I. Na, H.-G. Jeong, J. Kim, S. Lee, S.-S. Kim, H. Y. Koo, T.-W. Kim and D.-Y. Kim, J. Mater. Chem. C, 2013, 1, 3777–3783. 36 H. Wang, L. Yu, Y.-H. Lee, Y. Shi, A. Hsu, M. L. Chin, L.-J. Li, M. Dubey, J. Kong and T. Palacios, Nano Lett., 2012, 12, 4674–4680. 37 H. Liu and P. D. Ye, IEEE Electron Device Lett., 2012, 33, 546–548. 38 M. Chhowalla, H. Shin, G. Eda, L.-J. Li, K. Loh and H. Zhang, Nat. Chem., 2013, 5, 263–275. 39 G. Eda, H. Yamaguchi, D. Voiry, T. Fujita, M. Chen and M. Chhowalla, Nano Lett., 2011, 11, 5111–5116. 40 Y. Wang, J. Ou, S. Balendhran, A. Chrimes, M. Mortazavi, D. Yao, M. Field, K. Latham, V. Bansal, J. Friend, S. Zhuiykov, N. Medhekar, M. Strano and K. KalantarZadeh, ACS Nano, 2013, 7, 10083–10093. 41 Z. Liu, C. Lee, V. Narayanan, G. Pei and E. C. Kan, IEEE Trans. Electron Devices, 2002, 49, 1606–1613. 42 G.-H. Lee, Y.-J. Yu, C. Lee, C. Dean, K. L. Shepard, P. Kim and J. Hone, Appl. Phys. Lett., 2011, 99, 243114.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 11 August 2014. Downloaded by University of Western Ontario on 27/10/2014 06:42:47.
Nanoscale
43 C. A. Papageorgopoulos and W. Jaegermann, Surf. Sci., 1995, 338, 83–93. 44 T. Hu, R. Li and J. Dong, J. Chem. Phys., 2013, 139, 174702. 45 H. Li, Q. Zhang, C. C. R. Yap, B. K. Tay, T. H. T. Edwin, A. Olivier and D. Baillargeat, Adv. Funct. Mater., 2012, 22, 1385–1390. 46 G. Eda, T. Fujita, H. Yamaguchi, D. Voiry, M. Chen and M. Chhowalla, ACS Nano, 2012, 6, 7311–7317. 47 S. Balendhran, J. Ou, M. Bhaskaran, S. Sriram, S. Ippolito, Z. Vasic, E. Kats, S. Bhargava, S. Zhuiykov and K. KalantarZadeh, Nanoscale, 2012, 4, 461–466.
This journal is © The Royal Society of Chemistry 2014
Communication
48 G. L. Frey, R. Tenne, M. J. Matthews, M. S. Dresselhaus and G. Dresselhaus, Phys. Rev. B: Condens. Matter, 1999, 60, 2883–2892. 49 S. Sandoval, D. Yang, R. Frindt and J. Irwin, Phys. Rev. B: Condens. Matter, 1991, 44, 3955–3962. 50 A. Facchetti, M. H. Yoon and T. J. Marks, Adv. Mater., 2005, 17, 1705–1725. 51 W. L. Leong, P. S. Lee, A. Lohani, Y. M. Lam, T. Chen, S. Zhang, A. Dodabalapur and S. G. Mhaisalkar, Adv. Mater., 2008, 20, 2325–2331. 52 S.-T. Han, Y. Zhou, Q. D. Yang, L. Zhou, L.-B. Huang, Y. Yan, C.-S. Lee and V. A. L. Roy, ACS Nano, 2014, 8, 1923–1931.
Nanoscale, 2014, 6, 12315–12323 | 12323
