Synthetic Metals 183 (2013) 24–28
Contents lists available at ScienceDirect
Synthetic Metals journal homepage: www.elsevier.com/locate/synmet
Nano-floating gate organic memory devices utilizing Ag–Cu nanoparticles embedded in PVA-PAA-glycerol polymer A.I. Ayesh a,∗ , S. Qadri b , V.J. Baboo a , M.Y. Haik c , Y. Haik d,e a
Department of Physics, United Arab Emirates University, Al Ain, United Arab Emirates Department of Chemistry, University of North Carolina at Greensboro, Greensboro, NC, USA Department of Engineering Mechanics, Virginia Tech, Blacksburg, VA, USA d Center for Research Excellence in Nanobiosciences, University of North Carolina at Greensboro, Greensboro, NC, USA e Department of Mechanical Engineering, United Arab Emirates University, Al Ain, United Arab Emirates b c
a r t i c l e
i n f o
Article history: Received 15 May 2013 Received in revised form 13 August 2013 Accepted 18 September 2013 Keywords: Organic memory devices Ag–Cu nanoparticles Nano-floating gate memory
a b s t r a c t We report on the fabrication of nonvolatile organic memory devices that utilize silver–copper (Ag–Cu) nanoparticles as charge storage elements. Herein, Ag–Cu nanoparticles of an average size of 12 nm were impeded between thin layers of polymers: poly(methyl methacrylate) (PMMA), and poly-vinylalcohol/poly acrylamide co-acrylic acid with glycerol ionic liquid (PVA-PAA-glycerol). PMMA acts as a dielectric layer, while our newly developed PVA-PAA-glycerol polymer acts as a semiconducting layer. The conductivity of PVA-PAA-glycerol could be controlled conductivity by adjusting the percentage of glycerol. Aluminum films of desired thickness were produced by thermal evaporation and used as electrical electrodes. Capacitance–voltage (C(V)) measurements of the fabricated devices revealed hysteresis with a 10 V window. This is an indication of charge storage within Ag–Cu nanoparticles. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Highly efficient organic memory devices using inorganic nanoparticles are currently the focus of active research because of their applications in many fields [1]. The advantages of organic memory devices over inorganic device which are driving them to nano-electronics technology are the easiness in fabrication, low cost, and their flexibility [2]. Many researchers reported on the fabrication of hybrid organic memory device that utilize silicon substrates and nanoparticles as charge storage elements [3]. When a voltage is applied across the electrodes of those devices, charge trap occurs as the electrons are likely to flow from the electrode toward the organic film [4]. However, memory devices based on fully organic materials that include inorganic nanoparticles are much attractive toward enabling applications such as transparent electronic devices which are power saving, size compactable and easily portable [3]. Silver (Ag) and copper (Cu) nanoparticles are well known for their low cost and the antifouling properties which work effectively for reducing the growth of various microorganisms [5]. Previous investigations [6] revealed high potential of using Ag–Cu nanoparticles for organic devices with enhanced
∗ Corresponding author at: Department of Physics, United Arab Emirates University, P.O. Box 15551, Al Ain, United Arab Emirates. Tel.: +971 3 7136315. E-mail address:
[email protected] (A.I. Ayesh). 0379-6779/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.synthmet.2013.09.018
functionality. Thus, memory devices with those nanoparticles could be suitable for biocompatible implementable devices. In our previous work we demonstrated the fabrication of organic polymers that are based on poly-vinyl-alcohol (PVA) and poly-vinyl-alcohol/poly acrylamide co-acrylic acid (PAA) with engineered conductivity [7,8]. The control of the conductivity was achieved by doping the organic polymers with ionic liquids. In addition, we utilized poly(methyl methacrylate) (PMMA) as organic dielectric in memory device fabrication [9] because of its high resistivity and stability [3,4,10,11]. While the memory behavior of either Ag or Cu nanoparticles was investigated previously [12,13] the memory behavior of the composite Ag–Cu nanoparticles was not investigated previously, to the best of our knowledge. Therefore, we report in this work on memory storage behavior of Ag–Cu nanoparticles impeded between layers of PMMA insulator and our newly developed PVA-PAA-glycerol organic polymer with controlled conductivity. 2. Experimental 2.1. Preparation of organic polymer PVA and PAA with average molecular weights of 61,000 and 5,000,000 g/mol, respectively, were used in this work and purchased from Sigma–Aldrich. Polymer blends were produced by stirring 5 g of each polymer separately in 100 ml of deionized water. Each mixture was stirred at 90 ◦ C using a Stuart-CB16Z magnetic
A.I. Ayesh et al. / Synthetic Metals 183 (2013) 24–28
25
stirrer until it turned to a clear homogeneous jelly solution free of air bubbles. The solution was then left to cool down. 10 ml of each PVA and PAA solutions were mixed together with 0.2 g (1%) of glycerol:choline-chloride (1:2). This mixture was subjected to shaking in a Benchmark-Incu shaker at 35 ◦ C and stirring at 200 rpm for 4 h. Membranes of PVA-PAA and PVA-PAA-glycerol were obtained using the solution casting technique [7,8]. Herein, 20 ml of the mixture was casted onto a PTFE plate to form a thin film and left inside an oven at 60 ◦ C for 12 h. Membranes were tested for their dryness, free standing, and uniformity of thickness. PMMA solution was prepared by diluting 8 mg of PMMA powder (molecular weight 97,000 g/mol) in 1 ml of chloroform and left overnight under stirring to ensure a perfect dissolution.
2.2. Preparation of Ag–Cu nanoparticle Ag–Cu nanoparticles were prepared by mixing 61.6 mmol of silver nitrate and 26.4 mmol of copper acetate in a 50 ml conical flask that contains 10 ml of double distilled deionized water and 2 mol of glycine. The solid phase reaction mixture was heated at 170 ◦ C for 20 min. The reduction of Ag–Cu was monitored by change in color from black to brown. The precipitate was dried on a filter paper. The dried mixture pellet was re-suspended in 10 ml of distilled water and sonicated for 2 h using micro-tip probe of 450S-Branson Sonicator. To remove the large aggregates, the solution was centrifuged at 4000 rpm for 20 min. The free glycine was removed by dialyzing the solution in water for 24 h, using a Slide-A-Lyzer Dialysis Cassettes with MWCO 2 K (Thermo Scientific, Inc.), with replacing the water every 6 h. Solution was collected from dialysis cassette and stored in 10 ml screw cap glass vial.
2.3. Device fabrication Fabrication of a memory device involves the following steps. Glass substrates of dimensions 2 cm × 2 cm were cleaned by the standard cleaning method inside an ultrasonic cleaner using acetone, ethanol and deionized water (one after the other for repeated times), and then dried with nitrogen gas. The bottom electrode of the device was created by depositing aluminum (Al) film with thickness of 100 nm on the glass substrates using thermal evaporation. PMMA solution was spin coated on each substrate (on the Al electrode) initially at 500 rpm for 10 s, then at 5000 rpm for 50 s. Next, the sample is annealed at 120 ◦ C for 20 min. Ag–Cu nanoparticles were spread over the PMMA layer by spin coating for 90 s at a speed of 5000 rpm, and dried on a hotplate at 50 ◦ C. Another PMMA layer was spin coated over the nanoparticle layer and annealed. The polymer blend of PVA-PAA-glycerol was spin coated at 5000 rpm for 2 min and dried at 80 ◦ C for 5 min. A shadow mask consisting of an array of 2 mm diameter circular holes was used to fabricate an array of Al top electrodes by thermal evaporation with a thickness of 100 nm.
2.4. Tests and measurements Charge storage in nanoparticles was investigated using capacitance–voltage (C(V)) measurements using a computer controlled Keithley 590 CV Analyzer. All C(V) measurements were performed at 298 K. Electrical conductivity of the polymer membranes was measured using a Solartron SI1260 Impedance Gain Phase Analyzer within a frequency range of 1 Hz to 1 MHz as a function of temperature between 298 and 398 K inside a test chamber. Ag–Cu nanoparticle size was analyzed using a Malvern Instruments Zeta analyzer. The Ag and Cu contents of the nanoparticles were investigated using energy-dispersive X-ray spectroscopy (EDX).
Fig. 1. The temperature dependence of the natural logarithm of the resistivity for PVA-PAA and PVA-PAA-(1%) glycerol polymer membrane films.
Scanning electron microscope (SEM) was used to investigate the surface morphology of the devices. 3. Results and discussion Impedance measurements of pure and glycerol doped PVA-PAA membranes (without nanoparticles) were used to calculate the resistivity (), as discussed in Ref. [8]. Fig. 1 shows the temperature dependence of the natural logarithm of the resistivity. A linear fit of the results enables the calculation of the activation energy using: = 0 exp(Ea /kb T), where 0 is the pre-exponential factor, T is the absolute temperature, Ea is the activation energy, and kb is the Boltzmann constant. Herein, the activation energy represents the potential the holes need for hopping. The figure depicts that doping PVA-PAA membranes with glycerol changes the resistivity significantly. Furthermore, the slope of the curve is lower for the glycerol doped membranes indicating the decrease of the activation energy as a result of glycerol doping (Ea = 0.84 and 0.36 eV for pure and glycerol doped PVA-PAA membranes, respectively). For nanoparticle size analysis, 1 l of Ag–Cu nanoparticles was mixed with 3 l of deionized water which acted as the dispersant medium. The nanoparticle size was determined using the Zeta analyzer and illustrated in Fig. 2(a). The figure reveals an average particle size of 12 nm. To determine the elemental contents within Ag–Cu nanoparticles, a drop of the solution containing Ag–Cu nanoparticles was spread over carbon tape and left for 4 h at 60 ◦ C to dry. EDX measurement (shown in Fig. 2(b)) depicts that the average elemental percentages are 26.47% of Ag and 73.53% of Cu. Therefore, Ag24.5 Cu73.5 will be used in the text below to describe the correct composition of the nanoparticles. Three types of devices were fabricated and presented in this work. Device #1 represents the reference device and it exhibits the structure of glass\Al\PMMA\PVA-PAA-glycerol\Al. Devices #2 and #3 have the structures of glass\Al\PMMA\Ag24.5 Cu73.5 \PVAPAA-glycerol\Al and glass\Al\PMMA\Ag24.5 Cu73.5 \PMMA\PVAPAA-glycerol\Al, respectively. Fig. 3 shows schematic illustrations of the reference and memory devices embedded with Ag24.5 Cu73.5 nanoparticles. SEM images were produced for the Ag24.5 Cu73.5 nanoparticles to confirm their distribution over the PMMA layer. A sample SEM image is shown in Fig. 2(c). The image shows qualitatively that the nanoparticle coverage on the PMMA layer is
26
A.I. Ayesh et al. / Synthetic Metals 183 (2013) 24–28
Fig. 2. (a) Size distribution by volume analysis from the nano Zeta analyzer. (b) EDX spectrum of the Ag–Cu nanoparticles. (c) SEM image of Ag24.5 Cu73.5 nanoparticles over the PMMA layer. (d) Thickness of the spin coated PVA-PAA-glycerol measured by profile measurements using a nano-indenter.
approximately 50%. The thickness of the spin coated PVA-PAAglycerol layer was estimated by a profile test using a nano-indenter (Micro Materials Ltd., Wrexham, UK) to 16.5 m as shown in Fig. 2(d). The PMMA layer thickness was found to be ∼85 nm. C(V) measurements were performed for all fabricated devices with dual voltage sweep in the range of −20 to 20 V at a scan rate of 1 V/s and a frequency of 100 kHz inside a glovebox. The C(V) measurements produced for all devices shows the conventional accumulation (the device is charged), depletion, and inversion (the device is discharged) characteristics that are similar to a typical metal-insulator-semiconductor (MIS) devices based on a p-type semiconductors.
Fig. 4 shows the C(V) double sweep measurement of devices #1. For the reference device, no hysteresis appears on the double voltage sweep which is indicative of the absence of charge trapping in the bulk dielectrics and at the surfaces of the PVA-PAA-glycerol and PMMA layers. A flat band voltage (VFB ) of ∼−11 V is noted and an accumulation capacitance of 4.5 pF. All C(V) measurements were stable and reproducible as long as the device is inside the glovebox. The maximum accumulation capacitance can be used to calculate the dielectric constant of the PVA-PAA-glycerol and PMMA stack (εp ). Using Cmax ∼ = Aεp /dp [9], we obtained εp = 2.69, where A = 3.14 mm2 , and dp is the thickness of PVA-PAA-glycerol and PMMA stack ∼16.6 m.
Fig. 3. Schematic illustrations of the three types of devices tested. Device #1 is the reference device with glass\Al\PMMA\PVA-PAA-glycerol\Al structure. Devices #2 has the structure of glass\Al\PMMA\Ag24.5 Cu73.5 \PVA-PAA-glycerol\Al. Device #3 has the structure glass\Al\PMMA\Ag24.5 Cu73.5 \PMMA\PVA-PAA-glycerol\Al.
A.I. Ayesh et al. / Synthetic Metals 183 (2013) 24–28
27
Fig. 4. The C(V) curves of device #1 (reference) and device #2 with Ag24.5 Cu73.5 nanoparticles.
Fig. 5. The C(V) curves of device #1 (reference) and device #3 embedded with Ag24.5 Cu73.5 nanoparticles.
The doping concentration of PVA-PAA-glycerol film (Na ) can be calculated using [14]:
charge storing is offered by the conducting polymer membrane PVA-PAA-glycerol [26,27]. Nevertheless, structure #2 is an unfavorable structure for memory devices as the stored charges are likely to leak out through PVA-PAA-glycerol layer to the metal electrode. Fig. 5 shows the C(V) curves of device #3 compared with the reference device. The C(V) curve of device #3 depicts a clockwise hysteresis with a ∼10 V window and less noise compared with device #2. Ag24.5 Cu73.5 nanoparticles are charged through the PMMA insulating layer in the accumulation region when negative gate voltage is applied, and discharged to the electrode at the inversion region. The voltage shift of the center of the hysteresis toward the negative voltage (∼−1 V) is an indication of the presence of positive charges within the insulating layer where more negative voltage has to be applied to the top metal electrode to achieve the flat band conditions in the PVA-PAA-glycerol layer. The C(V) curve points to the fact that it needs a voltage less than −6 V to charge the nanoparticles (write) and about +3 V to discharge them (erase) with wide reading voltage range between −6 V and +3 V. The clockwise nature of the hysteresis for devices #3 is normally related to ion drift or polarization of the insulator indicating that the device exhibit characteristics similar to the MIS devices based on p-type semiconductors [3] (PVA-PAA-glycerol layer). Charge storage in the nanoparticle layer is responsible for the hysteresis effect because of the lack of any hysteresis for the reference device. We therefore suggest that, in accumulation, electrons are injected from the bottom electrode to the Ag24.5 Cu73.5 nanoparticles through the PMMA layer, which subsequently become negatively charged. The opposite effect occurs in inversion and electrons are extracted from the nanoparticles to the bottom electrode. On the other hand, when a positive voltage is applied to the bottom electrode, electrons are injected from the inversion layer on the PVA-PAA-glycerol layer into the Ag24.5 Cu73.5 nanoparticles. Charges stored in the Ag24.5 Cu73.5 nanoparticles (Q) can be estimated using Q = CFB VFB [9], where VFB is the flat band voltage shift (∼10 V). The calculation produced an approximate value of the charge density of 6.12 × 1011 cm−2 .
CFB = A
εIns
dIns + (εIns /εs )
(kb Tεs /e2 Na )
(1)
where CFB is the flat band capacitance, εs is the dielectric constant of PVA-PAA-glycerol film ∼2.69, εIns is the dielectric constant of the PMMA ∼ 4.88 [9], T is the temperature ∼300 K, e is the electron charge, and dIns is the thickness of PMMA layer. The calculations result a doping concentration of 1.4 × 1010 cm−3 . In addition, the maximum depletion width (Wmax ) can be calculated using the capacitance corresponding to the full depletion (Cmin ) by [15]: Cmin = A
εIns dIns + (εIns Wmax /εs )
(2)
Eq. (2) produces a maximum depletion length of 37 m which is almost double of the thickness for PVA-PAA-glycerol film. The overestimation of maximum depletion length could be assigned to the inhomogeneous mixing of glycerol within the film. The appearance of accumulation at the negative voltage of the scan is an indication of the presence of incorporated positive charges within the PVA-PAA-glycerol conducting polymer. When the applied gate voltage (Ag24.5 Cu73.5 floating gates) is further reduced lower than VFB , electrons are injected from the gate through the PMMA thin film into the floating gate resulting in charging the Ag24.5 Cu73.5 nanoparticles. Negative charges on the gate attract the holes in the PVA-PAA-glycerol leading to accumulation of holes in the polymer interface which is typical characteristic of a p-type material [3,16–18]. In the reverse sweep, the positive gate voltage drains out the holes from the PVA-PAA-glycerol interface leading to depletion and then to inversion with further increase the gate voltage. Fig. 4 also depicts the C(V) measurements of device #2, red curve. The curve shows a clockwise hysteresis pattern with wide window which is an indication of the charge storage within the nanoparticles. The flat band voltage is shifted to a negative value of −16 V with the addition of the Ag24.5 Cu73.5 nanoparticle. The direct contact between the charge storage elements and the PVA-PAAglycerol layer increases the amount of stored charges within the nanoparticles, thus, the window size. Herein, electric charges are injected to the nanoparticles from the metal electrodes through the PVA-PAA-glycerol film [19–22] during the voltage sweep levels by the induced electric field [23–25]. The wide hysteresis is associated with high charge transport mobility of carriers; and also is clearly evident that Ag24.5 Cu73.5 nanoparticles are responsible for the majority charge storage however minority contribution of
4. Conclusion In summary, we presented nano-floating gate organic memory devices based on Ag24.5 Cu73.5 nanoparticles and utilizing our newly developed poly-vinyl-alcohol/poly acrylamide co-acrylic acid/glycerol (PVA-PAA-glycerol) polymer with controlled electrical conductivity. Electrical conductivity investigation on the PVA-PAA-glycerol polymer membrane confirmed the suitability of using the membrane to replace the p-type semiconductor used
28
A.I. Ayesh et al. / Synthetic Metals 183 (2013) 24–28
previously in hybrid memory devices. Capacitance–voltage (C(V)) measurements depicts a hysteresis with ∼10 V window indicating the charge storage within the Ag24.5 Cu73.5 nanoparticles. The fabrication process was simple and cost effective which indicates that those devices have a potential to be used on larger scale. References [1] M. Alba-Martin, T. Firmager, J. Atherton, M.C. Rosamond, D. Ashall, A. Al Ghaferi, A. Ayesh, A.J. Gallant, M.F. Mabrook, M.C. Petty, D.A. Zeze, J. Phys. D: Appl. Phys. 45 (2012) 295401. [2] A. Sleiman, M.C. Rosamond, R.R. Nejm, A. Ayesh, A. Al Ghaferi, D.A. Zeze, M.F. Mabrook, J. Appl. Phys. 112 (2012) 024509. [3] M.F. Mabrook, C. Pearson, D. Kolb, D.A. Zeze, M.C. Petty, Org. Electron. 9 (2008) 816–820. [4] J.A. Avila-Nino, A.O. Sustaita, M. Reyes-Reyes, R. Lopez-Sandoval, J. Nanotechnol. 2011 (2011) 702464. [5] N. Cioffi, N. Ditaranto, L. Torsi, R.A. Picca, E. De Giglio, L. Sabbatini, L. Novello, G. Tantillo, T. Bleve-Zacheo, P.G. Zambonin, Anal. Bioanal. Chem. 382 (2005) 1912–1918. [6] S. Gamerith, A. Klug, H. Scheiber, U. Scherf, E. Moderegger, E.J.W. List, Adv. Funct. Mater. 17 (2007) 3111–3118. [7] M. Mohsin, A. Hossin, Y. Haik, J. Polym. Sci. 122 (5) (2011) 3102–3109. [8] A.I. Ayesh, M.A. Mohsin, M.Y. Haik, Y. Haik, Curr. Appl. Phys. 12 (2012) 1223–1228. [9] A. Sleiman, M.C. Rosamond, M. Alba Martin, A. Ayesh, A. Al Ghaferi, A.J. Gallant, M.F. Mabrook, D.A. Zeze, Appl. Phys. Lett. 100 (2012) 023302. [10] M.F. Mabrook, Y. Yun, C. Pearson, D.A. Zeze, M.C. Petty, Appl. Phys. Lett. 94 (2009) 173302.
[11] S.-J. Kim, Y.-S. Park, S.-H. Lyu, J.-S. Lee, Appl. Phys. Lett. 96 (2010) 033302. [12] J.H. Jung, J.H. Kim, T.W. Kim, M.S. Song, Y.H. Kim, S. Jin, Appl. Phys. Lett. 89 (2006) 122110. [13] S.M. Wang, C.W. Leung, P.K.L. Chan, Appl. Phys. Lett. 97 (2010) 023511. [14] J. Singh, Semiconductor Devices: Basic Principles, Wiley, New York, 2001. [15] Y. Yun, C. Pearson, M.C. Petty, J. Appl. Phys. 105 (2009) 034508. [16] L. Wei, X. Ling, Z. Wei-Ming, D. Hong-Lin, M. Zhong-Yuan, X. Jun, C. Kun-Ji, Chin. Phys. B 19 (2010) 047308. [17] R.K. Gupta, D.Y. Kusuma, P.S. Lee, M.P. Srinivasan, Mater. Lett. 68 (2012) 287–289. [18] J. Bird, Electrical and Electronic Principles and Technology, 3rd ed., Newnes Publication, Oxford, UK, 2007, ISBN-13: 978-0750685566. [19] A. Prakash, J. Ouyang, J.-L. Lin, Y. Yang, J. Appl. Phys. 100 (2006) 054309. [20] J. Ouyang, C.W. Chu, R.J.H. Tseng, A. Prakash, Y. Yang, Proc. IEEE 93 (2005) 1287–1296. [21] Y. Yang, L. Ma, J. Wu, MRS Bull. 29 (2004) 833–837. [22] S. Kolliopoulou, P. Dimitrakis, P. Normand, H.-L. Zhang, N. Cant, S.D. Evans, S. Paul, C. Pearson, A. Molloy, M.C. Petty, D. Tsoukalas, J. Appl. Phys. 94 (2003) 5234–5239. [23] H. Anderson, A. Rusu, A. Manuilskiy, S. Haller, S. Ayoz, H.E. Nilsson, Microelectron. J. 42 (2011) 21–27. [24] S. Paul, C. Pearson, A. Molloy, M.A. Cousins, M. Green, S. Kolliopoulou, P. Dimitrakis, P. Normand, D. Tsoukalas, M.C. Petty, Nano Lett. 3 (2003) 533–536. [25] H.J. Kim, S.M. Jung, Y.H. Kim, B.J. Kim, S. Ha, Y.S. Kim, T.S. Yoon, H.H. Lee, Thin Solid Films 519 (2011) 6140–6143. [26] D. Ick, D.H. Park, J.B. Kim, J.-W. Choi, T.W. Kim, B. Angadi, Y. Yi, W.K. Choi, J. Phys. Chem. C 115 (2011) 2341–2348. [27] Y.-H. Kim, M. Kim, S. Oh, H. Jung, Y. Kim, T.-S. Yoon, Y.-S. Kim, H.H. Lee, Appl. Phys. Lett. 100 (2012) 163301.
