Home
Search
Collections
Journals
About
Contact us
My IOPscience
Temperature dependence of resistive switching behaviors in resistive random access memory based on graphene oxide film
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2014 Nanotechnology 25 185202 (http://iopscience.iop.org/0957-4484/25/18/185202) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 129.8.242.67 This content was downloaded on 13/06/2014 at 08:47
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 25 (2014) 185202 (7pp)
doi:10.1088/0957-4484/25/18/185202
Temperature dependence of resistive switching behaviors in resistive random access memory based on graphene oxide film Mingdong Yi1, Yong Cao1, Haifeng Ling1, Zhuzhu Du1, Laiyuan Wang1, Tao Yang1, Quli Fan1, Linghai Xie1 and Wei Huang1,2 1
Center for Molecular Systems and Organic Devices (CMSOD), Key Laboratory for Organic Electronics & Information Displays (KLOEID) and Institute of Advanced Materials (IAM), Nanjing University of Posts &Telecommunications (NUPT), Nanjing 210 023, People’s Republic of China 2 Jiangsu-Singapore Joint Research Center for Organic/Bio Electronics & Information Displays, Institute of Advanced Materials (IAM), Nanjing University of Technology (NJUT), Nanjing 211 816, People’s Republic of China E-mail: [email protected] and [email protected] Received 1 November 2013, revised 4 March 2014 Accepted for publication 12 March 2014 Published 16 April 2014 Abstract
We reported resistive switching behaviors in the resistive random access memory (RRAM) devices based on the different annealing temperatures of graphene oxide (GO) film as active layers. It was found that the resistive switching characteristics of an indium tin oxide (ITO)/GO/ Ag structure have a strong dependence on the annealing temperature of GO film. When the annealing temperature of the GO film was 20 °C, the devices showed typical write-once-readmany-times (WORM) type memory behaviors, which have good memory performance with a higher ON/OFF current ratio (∼104), the higher the high resistance state (HRS)/low resistance state (LRS) ratio (∼105) and stable retention characteristics (>103 s) under lower programming voltage (−1 V and −0.5 V). With the increasing annealing temperature of GO film, the resistive switching behavior of RRAM devices gradually weakened and eventually disappeared. This phenomenon could be understood by the different energy level distributions of the charge traps in GO film, and the different charge injection ability from the Ag electrode to GO film, which is caused by the different annealing temperatures of the GO film. Keywords: memory, temperature dependence, graphene oxide, resistive switching (Some figures may appear in colour only in the online journal)
1. Introduction
materials has been observed by many research groups, and these two-terminal devices, consisting of carbon materials sandwiched between two electrodes, exhibited good nonvolatile memory properties which have higher ON/OFF current ratio, lower set/reset voltage and stable data retention [8–10]. As a typical carbon material, graphene oxide (GO) has particular advantages in the RRAM due to its unique
Resistive random access memory (RRAM) has attracted great attention for its potential applications in next generation nonvolatile memory owing to its simple structure, low operational voltage, fast switching speed and easy preparation [1–7]. Recently, the resistive switching effect of carbon 0957-4484/14/185202+07$33.00
1
© 2014 IOP Publishing Ltd Printed in the UK
M Yi et al
Nanotechnology 25 (2014) 185202
Figure 1. (a) Typical current-voltage (I-V) characteristics of the ITO/GO/Ag devices when the annealing temperature of the GO films was 20 °C. The inset shows the schematic illustration of the GO memory devices. (b). The retention characteristics of the GO memory devices at the programming voltage (−1 V and −0.5 V). (c). The retention characteristics of the resistance ratio of the high resistance state (HRS) and the low resistance state (LRS).
electrical properties, and the conductivity of GO film can achieve transition between the low conductivity state (OFF state) and the high conductivity state (ON state) under the electric field induction. To date, many reliable nonvolatile memory devices based on GO film were reported, and these devices showed a promising performance of flash and writeonce-read-many-times (WORM) type memory characteristics on rigid and flexible substrate [11–15]. The several operation mechanisms of GO memory devices were correspondingly proposed, including the formation and rupture of conducting filaments, the movement of oxygen ions and charge trapping and forming space-charge field [5, 16]. However, the accurate memory mechanism of GO memory devices has not been clarified yet, which restricts the development of GO memory devices. It is well known that the conductivity of GO film depends on the oxygen-induced defects and disorders, and these defects and disorders are also closely related to the annealing temperature of GO film [17–19]. Therefore, it is useful to fully understand the microscopic origin of resistive
switching of GO memory devices by studying the relationship between memory behaviors and the annealing temperature of GO film. In this paper, we reported the resistive switching effect of two-terminal indium tin oxide (ITO)/GO/Ag devices based on the different annealing temperatures of GO film as an active layer. The obvious temperature dependence of resistive switching of GO memory devices was observed when the annealing temperature of GO film ranged from 20 °C to 180 °C. Based on the results of electrical measurements and atomic force microscopy (AFM) analysis, the mechanism of the temperature dependence of resistive switching of our devices is discussed in detail.
2. Experimental details The two-terminal devices were fabricated as an ITO/GO/Ag structure, as shown in the inset of figure 1(a). After standard 2
M Yi et al
Nanotechnology 25 (2014) 185202
Figure 2. (a) The electrical transition process from the low conductivity state to the high conductivity state of the GO memory devices when the annealing temperature of the GO film ranged from 20 °C to 180 °C. (b). The influence of the annealing temperature of the GO films on S and the ON/OFF current ratio of the GO memory devices.
cleaning using acetone, ethanol, and deionized water, the ITO substrate was exposed to UV light for five minutes to enhance the work function and remove the residual organic molecules [20]. Then the GO aqueous solution was spin-coated on the ITO substrate, followed by annealing at a certain temperature (20 °C, 40 °C, 60 °C, 80 °C, 100 °C, 120 °C, 140 °C, 160 °C, 180 °C) for 30 mins. Finally, an Ag top electrode of about 100 nm thickness was thermally evaporated onto the GO film through a shadow mask to form the top electrode. To avoid damaging the GO film, the evaporated rate of Ag and the evaporated distance between the Ag source and the GO film was controlled to 1 Å s−1 and 40 cm, respectively. The I-V characteristics of the two-terminal ITO/GO/Ag devices were measured by an Agilent B1500A semiconductor parameter analyzer in the probe station in ambient condition. Figure 1(a) shows the typical current-voltage (I-V) characteristics of the ITO/GO/Ag devices when the annealing temperature of the GO films was 20 °C. The positive voltage bias was applied to the bottom electrode (ITO), the top electrode (Ag) was grounded. As the negative voltage increased (sweep 1), the current jumped abruptly from 10−7 A to 10−3 A at the voltage value of −0.4 V. The electrical state of the device switched from the low conductivity state (OFF state) to the high conductivity state (ON state), which is the ‘writing’ process in memory devices. After the electrical transition was achieved, the ON state was maintained steadily during the subsequent voltage sweep (sweep 2 and 3). Moreover, the electrical state could not recover to the OFF state, even after the power supply was removed (sweep 4). The I-V characteristics of the device show that the switch process between the OFF and ON state is irreversible and nonvolatile. Therefore, the device exhibited obvious WORM memory behavior. Figure 1(b) shows the retention characteristics of the GO memory devices. When the programming voltages were −1 V and −0.5 V, respectively, both the ON and OFF state can be retained for more than 103 s without slight fluctuations, indicating that the GO memory device is very stable. Meanwhile, the ON/OFF current ratio of the GO memory device is about 104, and the high ON/OFF current
ratio in the memory device can effectively avoid the misreading of the data. For RRAM, the resistance change characteristics of the GO memory devices are essential for resistive switching behavior. Figure 1(c) shows the retention characteristics of the resistance ratio of the high resistance state (HRS) and the low resistance state (LRS). The HRS/ LRS resistance ratio of the GO memory devices can be maintained 105 over 103 s, showing that the resistive switching behavior of the GO memory devices is very stable and distinguishable, which favors data storage. The lower programming voltage, the longer retention time, the higher ON/OFF current ratio, and the higher HRS/LRS resistance ratio of the ITO/GO/Ag device make it potentially suitable for next generation memory application.
3. Results and discussion Figure 2(a) shows the electrical transition process from the low conductivity state to the high conductivity state of the GO memory devices when the annealing temperature of the GO films ranged from 20 °C to 180 °C. As the annealing temperature of the GO films increased, the electrical transition process of the GO memory devices from the low conductivity state to the high conductivity state gradually weakened, and eventually became very flat at the higher annealing temperature (160 °C, 180 °C). It can also be seen that the range of the set voltages of GO memory devices under the different annealing temperatures of the GO films was very narrow, which distributed about −0.6 V ±0.2 V, showing that the GO memory devices have excellent programming properties. However, the set voltages of GO memory devices did not demonstrate the regular change with the annealing temperature of GO film. In order to quantitatively describe the above electrical transition process, we use the formula (S = log(I)/log (V)) to define the intensity of the electrical transition process. Figure 2(b) shows the influence of the annealing temperature of GO films on S and ON/OFF current ratio of the GO memory devices. The experiment data showed that both S and 3
M Yi et al
Nanotechnology 25 (2014) 185202
Figure 3. The injection and conduction models of the GO memory devices with the different annealing temperatures of the GO film: (a) ON state. (b) OFF state. (c). The dependence of the resistance of the ON state on the different annealing temperatures of the GO films.
reduced by an electrical field becomes less and less. As a result, the variation in the set voltage of GO memory devices under the different annealing temperature of GO films was mainly ascribed to the sp3-hybridized oxygenated carbon species, which can be reduced to sp2 -bond carbon atoms by the combined action of electrical field and thermal annealing. The resistive switching of GO memory devices showed the dependence on the annealing temperature of GO film. To further investigate the mechanism of the temperature dependence of resistive switching of the GO memory devices, we analyzed the I-V characteristics of the OFF and ON state of the devices when the annealing temperature of the GO film ranged from 20 °C to 180 °C, as shown in figure 3. Figure 3(a) shows the injection and conduction models for the OFF state of the GO memory devices with the different annealing temperatures of the GO film. According to the linear slope of the logarithmic plot of the I-V data, the conduction state of the electrical transition process of the GO memory devices when the annealing temperature of the GO films ranged from 20 °C to 140 °C, obeyed the space-chargelimited current (SCLC) model. This model is regarded as a trap-filling process (I ∼ Vm + 1, m > 2) [23], indicating that the
the ON/OFF current ratio decreased gradually with the increase of the annealing temperature of the GO film. Moreover, it was also found that there was positive correlation between the S and ON/OFF current ratio, indicating that more intense resistive switching during the electrical transition process may cause a higher ON/OFF current ratio. These results revealed that resistive switching behaviors of the GO memory devices gradually weakened, and eventually disappeared, with the increasing of the annealing temperature of the GO films. It is well known that GO consists of high conductivity sp2-bond carbon atoms and low conductivity sp3-hybridized oxygenated carbon species. The sp3-hybridized oxygenated carbon species can be reduced to sp2-bond carbon atoms by an external action (electrical field, thermal annealing and chemical reaction) which could enhance the conductivity of GO [21]. Therefore, the amount of sp2-bond carbon atoms transfered by sp3-hybridized oxygenated carbon species in GO film would continue to increase under the action of the higher annealing temperature, meaning that the ratio of sp2-bond carbon atoms to sp3-hybridized oxygenated carbon species increases [22]. In this case, the amount of sp3hybridized oxygenated carbon species which could be 4
M Yi et al
Nanotechnology 25 (2014) 185202
Figure 4. The atomic force microscopy (AFM) images of the three annealing temperatures of the GO film: (a) 20 °C. (b) 100 °C. (c) 180 °C.
The three-dimensional AFM images of the three annealing temperatures of the GO film: (d) 20 °C. (e) 100 °C. (f) 180 °C.
occupied charge traps in GO films are exponentially distributed in the deeper energy level of the GO band gap. It was also seen that the S1 (S1 = m + 1) value of the fitting line gradually reduced with the increasing of the annealing temperature of the GO films. When the annealing temperature of the GO films went up to 160 °C and 180 °C, the S1 value of the logarithmic plot of the I-V data reduced to 1.50 and 1.00. The I-V data of the electrical transition process of the GO memory devices can be fitted well by using trap-free SCLC (I ∼ V2) and Ohmic model (I ∼ V), respectively [24]. The different conductive states (trap-filling, trap-free SCLC, Ohmic model) of the electrical transition process of the GO memory devices showed that the distribution energy level position of the occupied charge traps became shallower in the GO band gap with the increasing of the annealing temperature of the GO films [24, 25]. After the electrical transition process, the conduction states at the lower annealing temperature of the GO film turned into the trap-free SCLC conduction, and the conduction states at the higher annealing temperature of the GO films (160 °C, 180 °C) still kept the former carrier transport model. For the ON state of the GO memory devices, the logarithmic plot of the I-V data is shown in figure 3(b). The obtained slopes at the different annealing temperature of GO films were nearly equal to 1, indicating that the carrier transport conduction belonged to the Ohmic model when the conductivity of the GO memory devices switched from the OFF to ON state. Figure 3(c) shows the dependence of the resistance of the ON state on the annealing temperature of the GO films. The resistance value of the ON state was found to
be inversely proportional to the annealing temperature of the GO films. The results of the I-V characteristics of the OFF and ON state showed that resistive switching of the GO memory devices is closely related to the existence of charge trap sites in GO films. In addition, when the positive voltage was applied to the Ag electrodes of the devices, it also exhibited the obvious WORM memory behaviors, as was the case with the injection and conduction models for the OFF and ON state, demonstrating that the resistive switching of GO memory devices is not affected by the direction of the carriers transport. It should be noted that the silver (Ag) atoms could be oxidized to Ag+ and further form a Ag conductive filament in the active layers, which cause the resistive switching behaviors of RRAM [26]. However, the formation and rupture of the Ag conductive filament usually were observed in the reversible nonvolatile memory devices [27]. The ITO/GO/Ag devices showed the typical irreversible nonvolatile memory (WORM) behaviors, thus the roles of the Ag conductive filament could be ruled out in our GO memory devices. These results further confirmed that the resistive switching of GO memory devices is attributed to charge trap sites in GO film [24]. Figure 4 shows the atomic force microscopy (AFM) images of the three annealing temperatures (20 °C, 100 °C, 180 °C) of the GO film on ITO substrate. When the annealing temperature of the GO film was 20 °C, the film had larger ruffles and a higher root-mean-square (RMS) of surface roughness (5.31 nm), as shown in figure 4(a). As the annealing temperature of the GO films rose to 100 °C, the 5
M Yi et al
Nanotechnology 25 (2014) 185202
intensity of the ruffles of the films became less obvious and the RMS of surface roughness of the film also decreased to 3.29 nm, as shown in figure 4(b). When the annealing temperature went up to 180 °C, the ruffles of the GO film almost disappeared, and the surface of the film became smoother, with an RMS roughness of about 1.05 nm, as shown in figure 4(c). So the AFM images of the GO film suggested that the intensity of the ruffles and surface roughness of the film were inversely related to the annealing temperature of the film. The above conclusion was further confirmed by the three-dimensional AFM images of the GO films, as shown in figure 4(d)–(f). For the GO memory devices, the surface morphology of the GO film that had more ruffles and larger surface roughness can favor broadening the effective contact area between the Ag electrode and the GO film. In this case, the larger interface dipole layer can be built into the interface of the Ag electrode, and the GO film then reduces the charge injection barrier and enhances the electron injection efficiency from the Ag electrodes to the GO film. Thus, when the annealing temperature of the GO films was lower, the injected electrons from the Ag electrodes obtained enough energy to overcome the energy barrier and were captured by the charge trap sites in the deeper energy level position of the GO film due to the existence of the larger interface dipole layer. Eventually, the degree of resistive switching of the GO memory devices became more intense under the same operational voltage, and the ON/OFF current ratio were also achieved. Based on the results of the electrical measurements and AFM analysis, it can be concluded that the different resistive switching behaviors of the GO memory devices with the different annealing temperatures of the GO film relate to the energy level distribution of the charge traps in the GO film, and the charge injection ability from the Ag electrode to the GO film. When the annealing temperature of the GO film was lower, the film had a higher sp3-hybridized oxygenated carbon species content. This makes the energy level distribution of charge traps deeper due to the increased amount of oxygenrelated functional groups (epoxide, hydroxyl, and carboxyl groups, etc) [25]. As a result, more injected electrons from the Ag electrode were captured under the same operational voltage, which enhanced the degree of resistive switching and ON/OFF current ratio of the GO memory devices. Meanwhile, the GO film at lower annealing temperatures had more ruffles and larger surface roughness, which can generate a larger interface dipole layer between the Ag electrode and the GO film. Thus, the electron injection ability from the Ag electrode to the GO film was enhanced under the effect of the interface dipole layer, meaning that the injected electrons could obtain enough energy to reach the deeper energy level of the GO film and were captured by charge traps located on the energy level position. Therefore, under the combined effect of the energy level distribution of the charge traps and the charge injection ability, the resistive switching effect of the GO memory device gradually weakened, and eventually disappeared, with the increasing of the annealing temperatures of the GO film.
4. Conclusions In summary, we have investigated the electrical characteristics of two-terminal ITO/GO/Ag devices based on the different annealing temperatures of the GO film as active layers. It was found that the annealing temperature of the GO film had a great effect on the resistive switching behaviors of the ITO/GO/Ag devices. When the annealing temperature of the GO film was lower, the typical WORM memory behaviors were observed in these devices. However, by increasing the annealing temperature of the GO film, the resistive switching behavior of the GO memory devices gradually weakened and eventually disappeared. According to the results of the fitted I-V curves of both the ON and OFF conductance states, and the AFM images at the different annealing temperatures of the GO film, the different resistive switching behaviors of the GO memory devices were attributed to the combined effect of the energy level distribution of the charge traps in the GO film, and the charge injection ability from the Ag electrode to the GO film, which was caused by the different annealing temperatures of the GO film.
Acknowledgments The project was supported by the National Basic Research Program of China (2012CB723402, 2014CB648300), National Science Fund for Excellent Young Scholars (21322402), National Natural Science Foundation of China (61204095, 61136003, 21144004), the Key Project of Chinese Ministry of Education, China (20113223120003), the Natural Science Foundation of Jiangsu Province, China (BK2012431), the Natural Science Foundation of the Education Committee of Jiangsu Province, China (11KJB510017), and the Scientific Research Staring Foundation of Nanjing University of Posts and Telecommunications, China (NY211022).
References [1] Liu G, Ling Q D, Kangand E T and Neoh K G 2007 J. Appl. Phys. 102 024502 [2] Xie L H, Ling Q D, Hou X Y and Huang W 2008 J. Am. Chem. Soc. 130 2120 [3] Jung G Y, Ganapathiappan S, Ohlberg D A A, Olynick D L, Chen Y, Tong W M and Williams R S 2004 Nano Lett. 4 1225 [4] Wu W, Jung G Y, Olynick D L, Straznicky J, Ohlberg D A A, Chen Y, Wang S Y, Liddle J A, Tong W M and Williams R S 2005 Appl. Phys. A 80 1173 [5] Green J E et al 2007 Nature 445 414 [6] Sekitani T, Yokota T, Zschieschang U, Klauk H, Bauer S, Takeuchi K, Takamiya M, Sakurai T and Someya T 2009 Science 326 1516 [7] Hong A J et al 2011 ACS Nano 5 7812 [8] Jeong H Y et al 2010 Nano Lett. 10 4381 [9] Kwon S, Seo H, Lee H, Jeonand K J and Park J Y 2012 Appl. Phys. Lett. 100 123101 6
M Yi et al
Nanotechnology 25 (2014) 185202
[19] Chen J R, Lin H T, Hwang G W, Chan Y J and Li P W 2009 Nanotechnology 20 255706 [20] Lo M F, Ng T W, Mo H W and Lee C S 2013 Adv. Funct. Mater. 23 1718 [21] Kuila T, Bose S, Mishra A K, Khanra P, Kim N H and Lee J H 2012 Prog. Mater. Sci. 57 1061 [22] Lim J, Choi K, Rani J R, Kim J S, Lee C, Kim J H and Jun S C 2013 J. Appl. Phys. 113 183502 [23] Liu J, Gu P Y, Zhou F, Xu Q F, Lu J M, Li H and Wang L H 2013 J. Mater. Chem. C 1 3947 [24] Ling Q D, Liaw D J, Zhu C X, Cha D S H, Kang E T and Neoh K G 2008 Prog. Polym. Sci. 33 917 [25] Shen Y et al 2013 Carbon 62 157 [26] Liao Z M, Hou C, Zhao Q, Wang D S, Li Y D and Yu D P 2009 Small 5 2377 [27] Liao Z M, Hou C, Zhang H Z, Wang D S and Yu D P 2010 Appl. Phys. Lett. 96 203109
[10] Ji Y S, Choe M, Cho B J, Song S, Yoon J, Ko H C and Lee T 2012 Nanotechnology 23 105202 [11] Yi M D et al 2011 J. Appl. Phys. 110 063709 [12] Wang L H, Yang W, Sun Q Q, Zhou P, Lu H L, Dingand S J and Zhang D W 2012 Appl. Phys. Lett. 100 063509 [13] Jin C H, Lee J H, Lee E, Hwang E and Lee H 2012 Chem. Commun. 48 4235 [14] Liu J Q, Lin Z Q, Liu T J, Yin Z Y, Zhou X Z, Chen S F, Xie L H, Freddy B, Zhang H and Huang W 2010 Small 14 1536 [15] Hong S K, Kim J E, Kim S O and Cho B J 2011 J. Appl. Phys. 110 044506 [16] Zhao F et al 2012 ACS Nano 4 3027 [17] Cristina G N, Meyer J C, Sundaram R S, Chuvilin A, Kurasch S, Burghard M, Kern K and Kaiser U 2010 Nano Lett. 10 1144 [18] Todd A D, Christopher W and Bielawski C W 2013 Polymer 6 021
7
Search
Collections
Journals
About
Contact us
My IOPscience
Temperature dependence of resistive switching behaviors in resistive random access memory based on graphene oxide film
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2014 Nanotechnology 25 185202 (http://iopscience.iop.org/0957-4484/25/18/185202) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 129.8.242.67 This content was downloaded on 13/06/2014 at 08:47
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 25 (2014) 185202 (7pp)
doi:10.1088/0957-4484/25/18/185202
Temperature dependence of resistive switching behaviors in resistive random access memory based on graphene oxide film Mingdong Yi1, Yong Cao1, Haifeng Ling1, Zhuzhu Du1, Laiyuan Wang1, Tao Yang1, Quli Fan1, Linghai Xie1 and Wei Huang1,2 1
Center for Molecular Systems and Organic Devices (CMSOD), Key Laboratory for Organic Electronics & Information Displays (KLOEID) and Institute of Advanced Materials (IAM), Nanjing University of Posts &Telecommunications (NUPT), Nanjing 210 023, People’s Republic of China 2 Jiangsu-Singapore Joint Research Center for Organic/Bio Electronics & Information Displays, Institute of Advanced Materials (IAM), Nanjing University of Technology (NJUT), Nanjing 211 816, People’s Republic of China E-mail: [email protected] and [email protected] Received 1 November 2013, revised 4 March 2014 Accepted for publication 12 March 2014 Published 16 April 2014 Abstract
We reported resistive switching behaviors in the resistive random access memory (RRAM) devices based on the different annealing temperatures of graphene oxide (GO) film as active layers. It was found that the resistive switching characteristics of an indium tin oxide (ITO)/GO/ Ag structure have a strong dependence on the annealing temperature of GO film. When the annealing temperature of the GO film was 20 °C, the devices showed typical write-once-readmany-times (WORM) type memory behaviors, which have good memory performance with a higher ON/OFF current ratio (∼104), the higher the high resistance state (HRS)/low resistance state (LRS) ratio (∼105) and stable retention characteristics (>103 s) under lower programming voltage (−1 V and −0.5 V). With the increasing annealing temperature of GO film, the resistive switching behavior of RRAM devices gradually weakened and eventually disappeared. This phenomenon could be understood by the different energy level distributions of the charge traps in GO film, and the different charge injection ability from the Ag electrode to GO film, which is caused by the different annealing temperatures of the GO film. Keywords: memory, temperature dependence, graphene oxide, resistive switching (Some figures may appear in colour only in the online journal)
1. Introduction
materials has been observed by many research groups, and these two-terminal devices, consisting of carbon materials sandwiched between two electrodes, exhibited good nonvolatile memory properties which have higher ON/OFF current ratio, lower set/reset voltage and stable data retention [8–10]. As a typical carbon material, graphene oxide (GO) has particular advantages in the RRAM due to its unique
Resistive random access memory (RRAM) has attracted great attention for its potential applications in next generation nonvolatile memory owing to its simple structure, low operational voltage, fast switching speed and easy preparation [1–7]. Recently, the resistive switching effect of carbon 0957-4484/14/185202+07$33.00
1
© 2014 IOP Publishing Ltd Printed in the UK
M Yi et al
Nanotechnology 25 (2014) 185202
Figure 1. (a) Typical current-voltage (I-V) characteristics of the ITO/GO/Ag devices when the annealing temperature of the GO films was 20 °C. The inset shows the schematic illustration of the GO memory devices. (b). The retention characteristics of the GO memory devices at the programming voltage (−1 V and −0.5 V). (c). The retention characteristics of the resistance ratio of the high resistance state (HRS) and the low resistance state (LRS).
electrical properties, and the conductivity of GO film can achieve transition between the low conductivity state (OFF state) and the high conductivity state (ON state) under the electric field induction. To date, many reliable nonvolatile memory devices based on GO film were reported, and these devices showed a promising performance of flash and writeonce-read-many-times (WORM) type memory characteristics on rigid and flexible substrate [11–15]. The several operation mechanisms of GO memory devices were correspondingly proposed, including the formation and rupture of conducting filaments, the movement of oxygen ions and charge trapping and forming space-charge field [5, 16]. However, the accurate memory mechanism of GO memory devices has not been clarified yet, which restricts the development of GO memory devices. It is well known that the conductivity of GO film depends on the oxygen-induced defects and disorders, and these defects and disorders are also closely related to the annealing temperature of GO film [17–19]. Therefore, it is useful to fully understand the microscopic origin of resistive
switching of GO memory devices by studying the relationship between memory behaviors and the annealing temperature of GO film. In this paper, we reported the resistive switching effect of two-terminal indium tin oxide (ITO)/GO/Ag devices based on the different annealing temperatures of GO film as an active layer. The obvious temperature dependence of resistive switching of GO memory devices was observed when the annealing temperature of GO film ranged from 20 °C to 180 °C. Based on the results of electrical measurements and atomic force microscopy (AFM) analysis, the mechanism of the temperature dependence of resistive switching of our devices is discussed in detail.
2. Experimental details The two-terminal devices were fabricated as an ITO/GO/Ag structure, as shown in the inset of figure 1(a). After standard 2
M Yi et al
Nanotechnology 25 (2014) 185202
Figure 2. (a) The electrical transition process from the low conductivity state to the high conductivity state of the GO memory devices when the annealing temperature of the GO film ranged from 20 °C to 180 °C. (b). The influence of the annealing temperature of the GO films on S and the ON/OFF current ratio of the GO memory devices.
cleaning using acetone, ethanol, and deionized water, the ITO substrate was exposed to UV light for five minutes to enhance the work function and remove the residual organic molecules [20]. Then the GO aqueous solution was spin-coated on the ITO substrate, followed by annealing at a certain temperature (20 °C, 40 °C, 60 °C, 80 °C, 100 °C, 120 °C, 140 °C, 160 °C, 180 °C) for 30 mins. Finally, an Ag top electrode of about 100 nm thickness was thermally evaporated onto the GO film through a shadow mask to form the top electrode. To avoid damaging the GO film, the evaporated rate of Ag and the evaporated distance between the Ag source and the GO film was controlled to 1 Å s−1 and 40 cm, respectively. The I-V characteristics of the two-terminal ITO/GO/Ag devices were measured by an Agilent B1500A semiconductor parameter analyzer in the probe station in ambient condition. Figure 1(a) shows the typical current-voltage (I-V) characteristics of the ITO/GO/Ag devices when the annealing temperature of the GO films was 20 °C. The positive voltage bias was applied to the bottom electrode (ITO), the top electrode (Ag) was grounded. As the negative voltage increased (sweep 1), the current jumped abruptly from 10−7 A to 10−3 A at the voltage value of −0.4 V. The electrical state of the device switched from the low conductivity state (OFF state) to the high conductivity state (ON state), which is the ‘writing’ process in memory devices. After the electrical transition was achieved, the ON state was maintained steadily during the subsequent voltage sweep (sweep 2 and 3). Moreover, the electrical state could not recover to the OFF state, even after the power supply was removed (sweep 4). The I-V characteristics of the device show that the switch process between the OFF and ON state is irreversible and nonvolatile. Therefore, the device exhibited obvious WORM memory behavior. Figure 1(b) shows the retention characteristics of the GO memory devices. When the programming voltages were −1 V and −0.5 V, respectively, both the ON and OFF state can be retained for more than 103 s without slight fluctuations, indicating that the GO memory device is very stable. Meanwhile, the ON/OFF current ratio of the GO memory device is about 104, and the high ON/OFF current
ratio in the memory device can effectively avoid the misreading of the data. For RRAM, the resistance change characteristics of the GO memory devices are essential for resistive switching behavior. Figure 1(c) shows the retention characteristics of the resistance ratio of the high resistance state (HRS) and the low resistance state (LRS). The HRS/ LRS resistance ratio of the GO memory devices can be maintained 105 over 103 s, showing that the resistive switching behavior of the GO memory devices is very stable and distinguishable, which favors data storage. The lower programming voltage, the longer retention time, the higher ON/OFF current ratio, and the higher HRS/LRS resistance ratio of the ITO/GO/Ag device make it potentially suitable for next generation memory application.
3. Results and discussion Figure 2(a) shows the electrical transition process from the low conductivity state to the high conductivity state of the GO memory devices when the annealing temperature of the GO films ranged from 20 °C to 180 °C. As the annealing temperature of the GO films increased, the electrical transition process of the GO memory devices from the low conductivity state to the high conductivity state gradually weakened, and eventually became very flat at the higher annealing temperature (160 °C, 180 °C). It can also be seen that the range of the set voltages of GO memory devices under the different annealing temperatures of the GO films was very narrow, which distributed about −0.6 V ±0.2 V, showing that the GO memory devices have excellent programming properties. However, the set voltages of GO memory devices did not demonstrate the regular change with the annealing temperature of GO film. In order to quantitatively describe the above electrical transition process, we use the formula (S = log(I)/log (V)) to define the intensity of the electrical transition process. Figure 2(b) shows the influence of the annealing temperature of GO films on S and ON/OFF current ratio of the GO memory devices. The experiment data showed that both S and 3
M Yi et al
Nanotechnology 25 (2014) 185202
Figure 3. The injection and conduction models of the GO memory devices with the different annealing temperatures of the GO film: (a) ON state. (b) OFF state. (c). The dependence of the resistance of the ON state on the different annealing temperatures of the GO films.
reduced by an electrical field becomes less and less. As a result, the variation in the set voltage of GO memory devices under the different annealing temperature of GO films was mainly ascribed to the sp3-hybridized oxygenated carbon species, which can be reduced to sp2 -bond carbon atoms by the combined action of electrical field and thermal annealing. The resistive switching of GO memory devices showed the dependence on the annealing temperature of GO film. To further investigate the mechanism of the temperature dependence of resistive switching of the GO memory devices, we analyzed the I-V characteristics of the OFF and ON state of the devices when the annealing temperature of the GO film ranged from 20 °C to 180 °C, as shown in figure 3. Figure 3(a) shows the injection and conduction models for the OFF state of the GO memory devices with the different annealing temperatures of the GO film. According to the linear slope of the logarithmic plot of the I-V data, the conduction state of the electrical transition process of the GO memory devices when the annealing temperature of the GO films ranged from 20 °C to 140 °C, obeyed the space-chargelimited current (SCLC) model. This model is regarded as a trap-filling process (I ∼ Vm + 1, m > 2) [23], indicating that the
the ON/OFF current ratio decreased gradually with the increase of the annealing temperature of the GO film. Moreover, it was also found that there was positive correlation between the S and ON/OFF current ratio, indicating that more intense resistive switching during the electrical transition process may cause a higher ON/OFF current ratio. These results revealed that resistive switching behaviors of the GO memory devices gradually weakened, and eventually disappeared, with the increasing of the annealing temperature of the GO films. It is well known that GO consists of high conductivity sp2-bond carbon atoms and low conductivity sp3-hybridized oxygenated carbon species. The sp3-hybridized oxygenated carbon species can be reduced to sp2-bond carbon atoms by an external action (electrical field, thermal annealing and chemical reaction) which could enhance the conductivity of GO [21]. Therefore, the amount of sp2-bond carbon atoms transfered by sp3-hybridized oxygenated carbon species in GO film would continue to increase under the action of the higher annealing temperature, meaning that the ratio of sp2-bond carbon atoms to sp3-hybridized oxygenated carbon species increases [22]. In this case, the amount of sp3hybridized oxygenated carbon species which could be 4
M Yi et al
Nanotechnology 25 (2014) 185202
Figure 4. The atomic force microscopy (AFM) images of the three annealing temperatures of the GO film: (a) 20 °C. (b) 100 °C. (c) 180 °C.
The three-dimensional AFM images of the three annealing temperatures of the GO film: (d) 20 °C. (e) 100 °C. (f) 180 °C.
occupied charge traps in GO films are exponentially distributed in the deeper energy level of the GO band gap. It was also seen that the S1 (S1 = m + 1) value of the fitting line gradually reduced with the increasing of the annealing temperature of the GO films. When the annealing temperature of the GO films went up to 160 °C and 180 °C, the S1 value of the logarithmic plot of the I-V data reduced to 1.50 and 1.00. The I-V data of the electrical transition process of the GO memory devices can be fitted well by using trap-free SCLC (I ∼ V2) and Ohmic model (I ∼ V), respectively [24]. The different conductive states (trap-filling, trap-free SCLC, Ohmic model) of the electrical transition process of the GO memory devices showed that the distribution energy level position of the occupied charge traps became shallower in the GO band gap with the increasing of the annealing temperature of the GO films [24, 25]. After the electrical transition process, the conduction states at the lower annealing temperature of the GO film turned into the trap-free SCLC conduction, and the conduction states at the higher annealing temperature of the GO films (160 °C, 180 °C) still kept the former carrier transport model. For the ON state of the GO memory devices, the logarithmic plot of the I-V data is shown in figure 3(b). The obtained slopes at the different annealing temperature of GO films were nearly equal to 1, indicating that the carrier transport conduction belonged to the Ohmic model when the conductivity of the GO memory devices switched from the OFF to ON state. Figure 3(c) shows the dependence of the resistance of the ON state on the annealing temperature of the GO films. The resistance value of the ON state was found to
be inversely proportional to the annealing temperature of the GO films. The results of the I-V characteristics of the OFF and ON state showed that resistive switching of the GO memory devices is closely related to the existence of charge trap sites in GO films. In addition, when the positive voltage was applied to the Ag electrodes of the devices, it also exhibited the obvious WORM memory behaviors, as was the case with the injection and conduction models for the OFF and ON state, demonstrating that the resistive switching of GO memory devices is not affected by the direction of the carriers transport. It should be noted that the silver (Ag) atoms could be oxidized to Ag+ and further form a Ag conductive filament in the active layers, which cause the resistive switching behaviors of RRAM [26]. However, the formation and rupture of the Ag conductive filament usually were observed in the reversible nonvolatile memory devices [27]. The ITO/GO/Ag devices showed the typical irreversible nonvolatile memory (WORM) behaviors, thus the roles of the Ag conductive filament could be ruled out in our GO memory devices. These results further confirmed that the resistive switching of GO memory devices is attributed to charge trap sites in GO film [24]. Figure 4 shows the atomic force microscopy (AFM) images of the three annealing temperatures (20 °C, 100 °C, 180 °C) of the GO film on ITO substrate. When the annealing temperature of the GO film was 20 °C, the film had larger ruffles and a higher root-mean-square (RMS) of surface roughness (5.31 nm), as shown in figure 4(a). As the annealing temperature of the GO films rose to 100 °C, the 5
M Yi et al
Nanotechnology 25 (2014) 185202
intensity of the ruffles of the films became less obvious and the RMS of surface roughness of the film also decreased to 3.29 nm, as shown in figure 4(b). When the annealing temperature went up to 180 °C, the ruffles of the GO film almost disappeared, and the surface of the film became smoother, with an RMS roughness of about 1.05 nm, as shown in figure 4(c). So the AFM images of the GO film suggested that the intensity of the ruffles and surface roughness of the film were inversely related to the annealing temperature of the film. The above conclusion was further confirmed by the three-dimensional AFM images of the GO films, as shown in figure 4(d)–(f). For the GO memory devices, the surface morphology of the GO film that had more ruffles and larger surface roughness can favor broadening the effective contact area between the Ag electrode and the GO film. In this case, the larger interface dipole layer can be built into the interface of the Ag electrode, and the GO film then reduces the charge injection barrier and enhances the electron injection efficiency from the Ag electrodes to the GO film. Thus, when the annealing temperature of the GO films was lower, the injected electrons from the Ag electrodes obtained enough energy to overcome the energy barrier and were captured by the charge trap sites in the deeper energy level position of the GO film due to the existence of the larger interface dipole layer. Eventually, the degree of resistive switching of the GO memory devices became more intense under the same operational voltage, and the ON/OFF current ratio were also achieved. Based on the results of the electrical measurements and AFM analysis, it can be concluded that the different resistive switching behaviors of the GO memory devices with the different annealing temperatures of the GO film relate to the energy level distribution of the charge traps in the GO film, and the charge injection ability from the Ag electrode to the GO film. When the annealing temperature of the GO film was lower, the film had a higher sp3-hybridized oxygenated carbon species content. This makes the energy level distribution of charge traps deeper due to the increased amount of oxygenrelated functional groups (epoxide, hydroxyl, and carboxyl groups, etc) [25]. As a result, more injected electrons from the Ag electrode were captured under the same operational voltage, which enhanced the degree of resistive switching and ON/OFF current ratio of the GO memory devices. Meanwhile, the GO film at lower annealing temperatures had more ruffles and larger surface roughness, which can generate a larger interface dipole layer between the Ag electrode and the GO film. Thus, the electron injection ability from the Ag electrode to the GO film was enhanced under the effect of the interface dipole layer, meaning that the injected electrons could obtain enough energy to reach the deeper energy level of the GO film and were captured by charge traps located on the energy level position. Therefore, under the combined effect of the energy level distribution of the charge traps and the charge injection ability, the resistive switching effect of the GO memory device gradually weakened, and eventually disappeared, with the increasing of the annealing temperatures of the GO film.
4. Conclusions In summary, we have investigated the electrical characteristics of two-terminal ITO/GO/Ag devices based on the different annealing temperatures of the GO film as active layers. It was found that the annealing temperature of the GO film had a great effect on the resistive switching behaviors of the ITO/GO/Ag devices. When the annealing temperature of the GO film was lower, the typical WORM memory behaviors were observed in these devices. However, by increasing the annealing temperature of the GO film, the resistive switching behavior of the GO memory devices gradually weakened and eventually disappeared. According to the results of the fitted I-V curves of both the ON and OFF conductance states, and the AFM images at the different annealing temperatures of the GO film, the different resistive switching behaviors of the GO memory devices were attributed to the combined effect of the energy level distribution of the charge traps in the GO film, and the charge injection ability from the Ag electrode to the GO film, which was caused by the different annealing temperatures of the GO film.
Acknowledgments The project was supported by the National Basic Research Program of China (2012CB723402, 2014CB648300), National Science Fund for Excellent Young Scholars (21322402), National Natural Science Foundation of China (61204095, 61136003, 21144004), the Key Project of Chinese Ministry of Education, China (20113223120003), the Natural Science Foundation of Jiangsu Province, China (BK2012431), the Natural Science Foundation of the Education Committee of Jiangsu Province, China (11KJB510017), and the Scientific Research Staring Foundation of Nanjing University of Posts and Telecommunications, China (NY211022).
References [1] Liu G, Ling Q D, Kangand E T and Neoh K G 2007 J. Appl. Phys. 102 024502 [2] Xie L H, Ling Q D, Hou X Y and Huang W 2008 J. Am. Chem. Soc. 130 2120 [3] Jung G Y, Ganapathiappan S, Ohlberg D A A, Olynick D L, Chen Y, Tong W M and Williams R S 2004 Nano Lett. 4 1225 [4] Wu W, Jung G Y, Olynick D L, Straznicky J, Ohlberg D A A, Chen Y, Wang S Y, Liddle J A, Tong W M and Williams R S 2005 Appl. Phys. A 80 1173 [5] Green J E et al 2007 Nature 445 414 [6] Sekitani T, Yokota T, Zschieschang U, Klauk H, Bauer S, Takeuchi K, Takamiya M, Sakurai T and Someya T 2009 Science 326 1516 [7] Hong A J et al 2011 ACS Nano 5 7812 [8] Jeong H Y et al 2010 Nano Lett. 10 4381 [9] Kwon S, Seo H, Lee H, Jeonand K J and Park J Y 2012 Appl. Phys. Lett. 100 123101 6
M Yi et al
Nanotechnology 25 (2014) 185202
[19] Chen J R, Lin H T, Hwang G W, Chan Y J and Li P W 2009 Nanotechnology 20 255706 [20] Lo M F, Ng T W, Mo H W and Lee C S 2013 Adv. Funct. Mater. 23 1718 [21] Kuila T, Bose S, Mishra A K, Khanra P, Kim N H and Lee J H 2012 Prog. Mater. Sci. 57 1061 [22] Lim J, Choi K, Rani J R, Kim J S, Lee C, Kim J H and Jun S C 2013 J. Appl. Phys. 113 183502 [23] Liu J, Gu P Y, Zhou F, Xu Q F, Lu J M, Li H and Wang L H 2013 J. Mater. Chem. C 1 3947 [24] Ling Q D, Liaw D J, Zhu C X, Cha D S H, Kang E T and Neoh K G 2008 Prog. Polym. Sci. 33 917 [25] Shen Y et al 2013 Carbon 62 157 [26] Liao Z M, Hou C, Zhao Q, Wang D S, Li Y D and Yu D P 2009 Small 5 2377 [27] Liao Z M, Hou C, Zhang H Z, Wang D S and Yu D P 2010 Appl. Phys. Lett. 96 203109
[10] Ji Y S, Choe M, Cho B J, Song S, Yoon J, Ko H C and Lee T 2012 Nanotechnology 23 105202 [11] Yi M D et al 2011 J. Appl. Phys. 110 063709 [12] Wang L H, Yang W, Sun Q Q, Zhou P, Lu H L, Dingand S J and Zhang D W 2012 Appl. Phys. Lett. 100 063509 [13] Jin C H, Lee J H, Lee E, Hwang E and Lee H 2012 Chem. Commun. 48 4235 [14] Liu J Q, Lin Z Q, Liu T J, Yin Z Y, Zhou X Z, Chen S F, Xie L H, Freddy B, Zhang H and Huang W 2010 Small 14 1536 [15] Hong S K, Kim J E, Kim S O and Cho B J 2011 J. Appl. Phys. 110 044506 [16] Zhao F et al 2012 ACS Nano 4 3027 [17] Cristina G N, Meyer J C, Sundaram R S, Chuvilin A, Kurasch S, Burghard M, Kern K and Kaiser U 2010 Nano Lett. 10 1144 [18] Todd A D, Christopher W and Bielawski C W 2013 Polymer 6 021
7
