The Effect of Al and Ni Top Electrodes in Resistive Switching Behaviors of Yb2O3-Based Memory Cells

(1)

P22 ECS Solid State Letters, 1 (2) P22-P25 (2012) 2162-8742/2012/1(2)/P22/4/$28.00©The Electrochemical Society

The Effect of Al and Ni Top Electrodes in Resistive Switching

Behaviors of Yb2O3-Based Memory Cells

Somnath Mondal,a_{Jim-Long Her,}b_{Fu-Hsiang Ko,}c_{and Tung-Ming Pan}a,z

a_{Department of Electronics Engineering, Chang Gung University, Taoyuan 333, Taiwan}

b_{Division of Natural Science, Center for General Education, Chang Gung University, Taoyuan 333, Taiwan} c_{Department of Materials Science and Engineering, Institute of Biological Science and Technology, National} Chiao-Tung University, Hsinchu 300, Taiwan

In this letter, the effect of Al and Ni top electrodes in resistive switching behavior of Al/Yb2O3/TaN and Ni/Yb2O3/TaN memory devices is proposed. The Al/Yb2O3/TaN memory device demonstrates no such switching performance as applying bias on both top and bottom electrodes, whereas the Ni/Yb2O3/TaN reveals the bipolar memory switching behavior with a high resistance ratio of 104_{for over 200 cycles of switching responses and good data retention with memory window of about 10}5_{at 85}◦_{C, as extrapolated} up to 10 years. The resistance switching dynamic is ascribed to the conductivity modulation by oxygen ions/vacancies controlled electrochemical reaction process in the Yb2O3switching layer.

Resistive random access memory (ReRAM) is an emerging stor-age device for next generation high potential nonvolatile memory (NVM) applications due to its scaling ability, low cost, and com-patibility of being integrated in back-end-of-line (BEOL) process in complementary metal-oxide-semiconductor (CMOS) fabrication.1–7

A simple ReRAM device structure usually consists of an oxide layer sandwiched between two metal electrodes. Application of low volt-age electrical pulse with certain level can change the oxide’s resis-tance significantly between two stable states, namely high resisresis-tance state (HRS) and low resistance state (LRS). Nonvolatile resistive switching behavior has been demonstrated for several binary ox-ides, including TiO2, Ta2O5, Al2O3, ZrO2, and Yb2O3.8–13Although

resistance-switching phenomena in several binary metal oxides have been known from decades, the details of the switching mechanisms and the nature of the different resistive states are still under debate. Different switching models such as domain and tunneling model,14

filament model,15,16_{and field induced crystallinity and charge trapping}

model17–19_{have been studied extensively. However, exact switching}

phenomena are not clearly understood yet. In this letter, we explored an improved resistive switching behavior of Ni/Yb2O3/TaN memory

cells in compare with Pt/Yb2O3/TiN memory cell.13 Additionally,

the influence of top electrode material (Ni and Al) on the electrical properties of Yb2O3-based ReRAM is discussed to realize for next

generation universal memory application.

It is found that the bipolar resistive switching behavior in Ni/Yb2O3/TaN ReRAM cell keeps stable resistance ratio of 104with

switching responses over 200 cycles. Good data retention for 105 _s

with sufficient memory window at 85◦C is also demonstrated. The re-sistance switching dynamic is attributed to the filamentary conduction of oxygen ions/vacancies by electrochemical reaction process in the Yb2O3switching layer.

Experimental

The TaN bottom electrode was deposited on SiO2/Si substrates

by sputter deposition technique. Then, a 30-nm Yb2O3thin film was

deposited on the TaN by radio frequency (rf) magnetron sputtering with an Ar/O2 atmosphere using a metallic Yb target. The

cham-ber working pressure was maintained at 10−2Torr, while the Ar/O2

flow was set at 3:1 proportion. A shadow mask with circular areas (100μm diameter) was used to define the top electrode. Finally, a 100-nm top electrode (Al or Ni) was deposited by means of ther-mal evaporation technique. The merits of Ni electrode are low cost, high work function and compatibility with existing CMOS process. X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS)

z_E-mail:_{[email protected]}

were employed to determine the structure and the chemical bond-ing state of the Yb2O3 thin films, respectively. The electrical

mea-surements were performed in dark with Agilent E5260A high speed semiconductor parameter analyzer. The bottom electrode TaN was grounded and the signals were applied to the top electrode in all the measurements.

Results and Discussion

Fig.1ashows the XRD micrograph of as-deposited Yb2O3 thin

film grown on TaN bottom electrode. Presence of no Yb2O3 peak

in the data revealed an amorphous phase of the film, whereas the previous result showed some nanocrystal in the oxide film.13 _This

may be due to different deposition conditions in sputtering system. Inset of Fig.1ashows cross sectional TEM image of the Yb2O3thin

film, corroborates well with the XRD data. The core level O 1s spectra with their appropriate peak curve-fitting lines for the Yb2O3film are

shown in Fig.1b. The O 1s spectrum at the surface of Yb2O3thin film

consists of a low peak binding energy at 529.4 eV for Yb2O3lattice

oxygen ions, and a high peak binding energy at 531.9 eV indicates a large amount of non-lattice oxygen ions usually attributed to OH− groups. Rare-earth oxide is known to actively react with OH due to its hydration characteristics.20

Fig.2ashows the current-voltage (I-V) characteristics of the fab-ricated Al/Yb2O3/TaN ReRAM cell before and after electro-forming

process. The memory device using Al top electrode exhibited no switching characteristics after electroforming process. This can be at-tributed to the formation of stable Al-O layers at the Al and Yb2O3

interface due to the lower workfunction and higher oxygen affinity of Al electrode.21_{By applying voltage stress on either polarity at the}

Al electrode, an irreversible switching occurs which is commonly depicted as hard breakdown of dielectric. Furthermore, to look at the electrode thickness dependence resistive switching behavior, the Al/Yb2O3/TaN memory device with different Al thicknesses was

in-vestigated. All devices with an Al top electrode show no resistive switching characteristics in this study. In contrast, the ReRAM cell with Ni top electrode exhibit a reproducible bipolar resistive switch-ing behavior by dc double sweep for more than 200 sweepswitch-ing cycles, as depicted in Fig.2b. An electrical forming process is required to initiate the switching process by dc negative voltage sweeping with a compliance current of 0.1 mA. After the forming process, the device switches to LRS. By sweeping the voltage to opposite polarity above reset value (Vreset) of about 2 V, a sudden decrease in conduction

current is observed. The resistance of the memory cell switches from LRS to HRS. Contrariwise, the cell turns back to LRS while applying a negative bias over set voltage (Vset) about−2.5 V. A current

com-pliance of 1 mA was assigned to prevent the permanent breakdown of the memory devices during switching process. We consider that, the

) unless CC License in place (see abstract).

ecsdl.org/site/terms_use

address. Redistribution subject to ECS terms of use (see

140.113.38.11

(2)

ECS Solid State Letters, 1 (2) P22-P25 (2012) P23 537 534 531 528 525 (b) In ten sity (a.u.)

Binding energy (eV)

Data Fitted Deconvulated O1s Yb₂O₃ O0+ 20 30 40 50 60 70 (a) Intensity (a.u.) 2θ (degree) Yb₂O₃ TaN (101) θθθθθθθθθθθ

Figure 1. (a) XRD micrograph of Yb2O3 thin film de-posited on TaN electrode. The inset shows the cross sec-tional TEM image of the Yb2O3film. (b) the corelevel O1s spectrum on the surface of Yb2O3thin film.

high work-function metal electrode with low oxygen affinity as anode and high oxygen gettering electrode as cathode facilitate to produce stable resistive switching characteristics in Ni/Yb2O3/TaN memory

device.13,22

To clarify the origin of a bipolar resistive switching behavior, the current conduction mechanism of HRS and LRS was investigated. The I-V curves were plotted on the logarithmic scale, as shown in Fig.3a. In the HRS, the I-V characteristics of a Ni/Yb2O3/TaN device show

lin-ear behavior up to approximately 0.1 V, suggesting that the thermionic emission limited conduction (TELC) is the dominant mechanism at the lower regime, where carrier injection from the electrode plays a minor role. Beyond that, the increase in the slope is observed, as determined by the fit I∞ Vn_(n_{= 2 and 3), which can be understood as a}

tran-sition of conduction mechanism from Ohmic TELC to space charge limited current (SCLC). In the LRS, the slope is nearly constant as 1, which represents that the Ohmic conduction is dominant in all bias range during LRS. This may be due to the filamentary conduction of the accumulated oxygen vacancies into the oxide.23_Fig._3b_presents

the resistances of the LRS and the HRS as a function of temperature. In HRS, the resistance value decreases as the temperature increase from 300 K to about 400 K and this can be associated with a semi-conducting behavior of the oxide film while, the resistance at LRS increases as the temperature rises, which indicates weak metallic like behavior. As a consequence, the resistance ratio decreases about one order over a temperature range of 300 - 400 K. This can be attributed to the built-up of oxygen vacancy related traps inside the Yb2O3

film.24

Based on the above observation, we proposed the switching mecha-nism in the Ni/Yb2O3/TaN device as the migration of charge particles,

most possibly ions. It has been reported that oxygen vacancies in some metal oxides give rise to delocalized conduction electrons rather than localized in the band-gap region25,26 _{and, hence, it is assumed that}

the oxygen vacancies in Yb2O3 are ionized and act as mobile space

charges. This is confirmed by the presence of large amount of nonlat-tice oxygen ions in the XPS spectrum in Fig.1b. It is also believed that the low work function metals have more stable oxide and act as a sink for oxygen creating more oxygen vacancies in the oxide.27_Therefore,

it can be speculated that the physisorption or chemisorption of oxygen ions into the TaN electrode leads to the formation of oxygen vacancies

or metallic defects in the Yb2O3layer. The electrochemical reaction

with the Kr¨oger-Vink notation is given by

O×_O(Me)+ V_O••(O_×)↔ V_O••(Me)+ O×_O(O_×) (1) where Me and Oxin the parentheses represent the metal electrode and

oxide, respectively, and oxygen ions are assumed to have a charge of −2. During the forming process at sufficient large electric field, the nonlattice oxygen ions in the Yb2O3film drift toward TaN electrode.

The accumulated large number of oxygen vacancies in the oxide/TaN interface modulates the conductivity of the oxide, resulting in the de-vice switched from HRS to LRS. Stressing back by opposite polarity of bias, oxygen ions releasing from the TaN electrode will neutralize oxygen vacancies; hence the bipolar behavior is prompted upon the annihilation of oxygen vacancies.12,13_Fig.₄_{shows the driving}

mech-anism of the resistive memory switching due to the physisorption or chemisorption of oxygen ions near the TaN electrode interface at dif-ferent bias polarities on the device. Conversely, for positive forming, the electrochemical reaction of Eq.(1)occurs near the top electrode. At the Ni/Yb2O3interface, the electrode is oxidized by oxygen ions

and no switching is observed likely in the Al/Yb2O3/TaN ReRAM

cell.

Fig. 5shows the dependence of the resistance on the repetitive switching cycles of a Ni/Yb2O3/TaN ReRAM cell at both room

temperature and 85◦C. The device resistance values were read out at 0.2 V, while set and reset were performed at negative and positive voltage sweeping, respectively. It can be seen from the experimental data, the LRS was dispersed in a narrow range around 100, while the HRS value was somewhat fluctuant in the range around 10 M. The resistance ratio of HRS to LRS was in the range of 4-5 orders of magnitude within the 200 cycles. To compare with other work,12,13

the Ni/Yb2O3/TaN ReRAM cell exhibits a higher memory window

for operation. This may be attributed to different conditions of the ox-ide film in the ReRAM device. Moreover, the uniformity of memory parameters such as resistance values is also a very important param-eter for practical application. In the inset of Fig.5, the statistics of two resistance states at room temperature are depicted. A clear and non overlap window of resistances between HRS and LRS ensured

-15 -10 -5 0 5 10 10-12 10-10 10-8 10-6 10-4 10-2 100 Al/Yb 2O3/TaN (a) Fresh Forming After forming Cur re n t (A) Bias (V) -10 -8 -6 -4 -2 0 2 4 10-12 10-10 10-8 10-6 10-4 10-2 Curr ent (A) Ni/Yb 2O3/TaN (b) Bias (V) Forming 1st 100th 200th

Figure 2. (a) I-V characteristics of Al/Yb2O3/TaN ReRAM devices before and after electroforming pro-cess. (b) Typical resistive switching characteristics of Ni/Yb2O3/TaN ReRAM devices cell after standard elec-troforming process for 1st, 10th, and 200th sweeping cycles.

140.113.38.11

(3)

P24 ECS Solid State Letters, 1 (2) P22-P25 (2012) 0.01 0.1 1 10-11 10-9 10-7 10-5 10-3 10-1 HRS Current (A) Bias (V) 1.03 3.02 2.01 LRS (a) 1.01 300 320 340 360 101 103 105 107 109 LRS Resistance ( Ω ) Temperature (K) HRS (b)

Figure 3. (a) Double logarithmic plots of |I|-|V| curves

of Ni/Yb2O3/TaN ReRAM device. (b) Temperature de-pendent resistance values of the memory cell at HRS and LRS.

the Yb2O3-based device can be well operated in the application of

ReRAM.

To confirm the potentiality in memory application, the data reten-tion properties of the fabricated Ni/Yb2O3/TaN ReRAM cell in HRS

and LRS were also measured at room temperature (RT) and 85◦C, respectively. As shown in Fig.6, the time-dependent resistance evo-lution exhibits little change in the magnitude of resistance in HRS, while no significant change of the same in LRS after the duration for 4×104 _{s, and demonstrates extrapolated 10-year retention with}

nondestructive read-out under both temperatures. The superior data retention characteristics of the Ni/Yb2O3/TaN ReRAM cell reveal the

potential for nonvolatile memory applications.

Figure 4. Schematic of driving mechanism of resistive memory switching by

filamentary conduction due to physisorption/chemisorption of oxygen ions at TaN electrode.

0

50

100

150

200

10

1

10

3

10

5

10

7

10

9 10 1 k 1 0 0 k 1 0 M 0 3 6 9 1 2 1 5 1 8 C o unt s R (Ω)

Resistance (

Ω

)

No. of cycle

Solid synbol @ RT Open symbol @ 85 oC

Figure 5. Endurance characteristics of Ni/Yb2O3/TaN ReRAM device for DC sweeping cycles at room temperature and 85◦C. The inset shows the distribution of resistance values for 200 switching cycles.

10

2

10

4

10

6

10

8

10

1

10

3

10

5

10

7

10

9 HRS@ RT LRS @ RT HRS @ 85 oC LRS @ 85 oC

Resistance (

Ω

)

Time (s)

10-year life-time

Figure 6. Data retention performance of the Ni/Yb2O3/TaN ReRAM device in HRS and LRS under room temperature and 85◦C, respectively. The extrap-olation method is employed to give a long-term (10-year) prediction result.

140.113.38.11

(4)

ECS Solid State Letters, 1 (2) P22-P25 (2012) P25

Conclusions

In summary, the influence of Al and Ni electrodes in resistive switching characteristics of Al/Yb2O3/TaN and Ni/Yb2O3/TaN

ory devices was investigated for nonvolatile random access mem-ory applications. The switching behavior of Yb2O3-based memory

cell is related with the properties of the top electrode material. The Ni/Yb2O3/TaN ReRAM device exhibited good endurance of over

200 dc switching cycles and nondestructive readout for 10-year data retention performance maintaining the memory window of∼105 _at

85◦C. The resistive switching in Ni/Yb2O3/TaN ReRAM cell could

be regarded as the formation/annihilation of conducting filaments in Yb2O3layer by electrochemical reaction process. The Ni/Yb2O3/TaN

memory cell is very promising candidate for future nonvolatile mem-ory applications.

Acknowledgment

This project was supported by the National Science Council (NSC) of Republic of China under contract no. NSC-97-2221-E-182-050-MY3.

References

1. B. J. Choi, D. S. Jeong, S. K. Kim, C. Rohde, S. Choi, J. H. Oh, H. J. Kim, C. S. Hwang, K. Szot, R. Waser, B. Reichenberg, and S. Tiedke,J. Appl. Phys.,

98, 033715 (2005).

2. H. Akinaga and H Shima,Proc. IEEE, 98, 2237 (2010).

3. K. Szot, W. Speier, R. Carius, U. Zastrow, and W. Beyer,Phys. Rev. Lett., 88, 75508 (2002).

4. S. Tsui, A. Baikalov, J. Cmaidalka, Y. Y. Sun, Y. Q. Wang, Y. Y. Xue, C. W. Chu, L. Chen, and A. J. Jacobson,Appl. Phys. Lett., 85, 317 (2004).

5. S. Seo, M. J. Lee, D. H. Seo, E. J. Jeoung, D.-S. Suh, Y. S. Joung, I. K. Yoo, I. R. Hwang, S. H. Kim, I. S. Byun, J.-S. Kim, J. S. Choi, and B. H. Park,Appl. Phys. Lett., 85, 5655 (2004).

6. S. Zhang, S. Long, W. Guan, Q. Liu, Q. Wang, and M. Liu,J. Phys. D: Appl. Phys.,

42, 055112 (2009).

7. A. Beck, J. G. Bednorz, Ch. Gerber, C. Rossel, and D. Widmer,Appl. Phys. Lett., 77, 139 (2000).

8. K. M. Kim, B. J. Choi, Y. C. Shin, S. Choi, and C. S. Huang,Appl. Phys. Lett., 91, 012907 (2007).

9. K. L. Chopra,J. Appl. Phys., 36, 184 (1965). 10. T. W. Hickmott,J. Appl. Phys., 33, 2669 (1962).

11. Q. Liu, W. Guan, S. Long, R. Jia, and M. Liu,Appl. Phys. Lett., 92, 012117 (2008). 12. H.-C. Tseng, T.-C. Chang, J.-J. Huang, P.-C. Yang, Y.-T. Chen, F.-Y. Jian, S. M. Sze,

and M.-J. Tsai,Appl. Phys. Lett., 99, 132104 (2011).

13. H.-C. Tseng, T.-C. Chang, J.-J. Huang, Y.-T. Chen, P.-C. Yang, H.-C. Huang, D.-S. Gan, N.-J. Ho, S. M. Sze, and M.-J. Tsai,Thin Solid Films, 520, 1656 (2011). 14. M. J. Rozenberg, I. H. Inoue, and M. J. S´anchez,Phys. Rev. Lett., 92, 178302 (2004). 15. J. F. Gibbons and W. E. Beadle,Solid-State Electron., 7, 785 (1964).

16. K. Tsunoda, Y. Fukuzumi, J. R. Jameson, Z. Wang, P. B. Griffin, and Y. Nishi,Appl. Phys. Lett., 90, 113501 (2007).

17. N. A. Tulina, S. A. Zver’kov, Y. M. Mukovskii, and D. A. Shulyatev,Europhys. Lett.,

56, 836 (2001).

18. J. G. Simmons and R. R. Verderber,Proc. R. Soc. London, Ser. A, 301, 77 (1967). 19. S. Q. Liu, N. J. Wu, and A. Ignatiev,Appl. Phys. Lett., 76, 2749 (2000).

20. D. Eom, S. Y. No, H. Park, C. S. Hwang, and H. J. Kim,Electrochem. Solid-State Lett., 10, G93 (2007).

21. S. Jou, B. R. Hwang, and C. J. Li, Proceedings of the World Congress on engineering 2011, Vol II, WCE 2011, July 6–8, 2011, London, U.K.

22. P. Zhou, M. Yin, H. J. Wan, H. B. Lu, T. A. Tang, and Y. Y. Lin,Appl. Phys. Lett.,

94, 053510 (2009).

23. C. Chen, Y. C. Yang, F. Zeng, and F. Pan,Appl. Phys. Lett., 97, 083502 (2010). 24. Z. Fang, H. Y. Yu, W. J. Liu, K. L. Pey, X. Li, L. Wu, Z. R. Wang, P. G. Q. Lo, B. Gao,

and J. F. Kang, Proc. IEEE Int. Reliab. Phys. Symp., 964 (2010).

25. E. Cho, S. Han, H.-S. Ahn, K.-R. Lee, S. K. Kim, and C. S. Hwang,Phys. Rev. B, 73, 193202 (2006).

26. D. S. Jeong, H. Schroeder, and R. Waser,Phys. Rev. B, 79, 195317 (2009). 27. C. Vallee, P. Gonon, C. Jorel, and F. E. Kamel,Appl. Phys. Lett., 96, 233504 (2010).

140.113.38.11