SI CHINA     SI JAPAN
Login  |  Register          Free Newsletter Subscription
Subscribe
Email
Print
Reprint
Learn RSS

TaOxNy: A Promising Alternative to Ta2O5

Youngho Lee, Munsu Park and Juho Song Jusung Engineering Co. Ltd., Kyunggi-Do, Korea -- Semiconductor International, 7/1/2001

  
 At a Glance

Although tantalum pentoxide (Ta2O5) is a promising material for the high dielectric constants that will be required in next-generation memories, there is room for improvement. We have developed a process for producing TaOxNy high-k dielectric materials that demonstrate characteristics that are superior to those of traditional Ta2O5.

Next-generation memory devices will require materials with higher dielectric constants. Tantalum pentoxide (Ta2O5) is one of the most promising materials because it has a high dielectric constant (~25), meets the requirements for conventional VLSI processes, and has comparatively low leakage current.1-3

However, compared with some types of ferroelectric dielectrics, such as (Ba, Sr)TiO3, the dielectric constant of Ta2O5 films is modest. If the dielectric constant of Ta2O5 films can be physically improved, their integration into next-generation ULSI devices may be more readily achievable.

It is known that a plasma anneal in a nitrogen ambient can significantly reduce leakage current and trap density in the film.4 In addition, it has been reported that the dielectric constant of Ta2O5 films deposited on Ru can be significantly improved by a rapid thermal nitridation.5 In this article, we describe how we deposited TaOxNy films using NH3 and Ta(OC2H5)5 source materials, analyzed the impact of nitrogen on the Ta2O5 film, and carefully compared the electrical characteristics of TaOxNy to traditional Ta2O5 films.

Experiment

Research was performed using Jusung's EUREKA2000 CVD tool, which is distinguishable from conventional single-wafer CVD reactors in that it utilizes a gas boundary layer. It uses an injector system to obtain good film uniformity without a showerhead. It has a warm wall heating method that uses two heat sources to maximize the heat capacity and thermal stability in the reactor. Therefore, the incoming gases are preheated to minimize heat loss at the wafer surface, providing optimal control of temperature uniformity.

The reactor can also perform plasma-enhanced CVD (PECVD) and plasma-enhanced dry cleaning. It is equipped with remote and afterglow plasma discharge technology to optimize the PECVD, in situ dry wafer cleaning, and in situ chamber cleaning processes.

1. RBS spectra of 30 nm-thick TaOxNy film deposited with 25 sccm of NH3.
The precursor for Ta2O5 deposition, Ta(OC2H5)5, is a liquid source. Thus, we had to convert liquid states of Ta(OC2H5)5 into gas states for the Ta2O5 deposition process, requiring a liquid delivery system (LDS). We used an LDS designed by Jusung that features an independent dry pump in to ensure the repeatability of liquid flow and wafer-to-wafer thickness uniformity. The LDS also contains ethanol and gas purge lines for cleaning.

TaOxNy-based capacitors were fabricated using phosphorus-doped polycrystalline silicon as a bottom electrode. As a first step, the wafers were cleaned using 100:1 DHF.

Next, 1.2 nm-thick nitride films were grown by NH3 plasma at 430°C to prevent oxidation from forming at the interface during the Ta2O5 deposition or during post-treatment. After nitridation, 8 nm-thick Ta2O5 films were deposited using Ta(OC2H5)5 and O2.

2. Capacitance (a) and leakage current (b) of Ta2O5 and TaOxNy films are compared relative to bias voltage.
Separately, for purposes of comparison, 8 nm-thick TaOxNy films were deposited using Ta(OC2H5)5 and NH3. For RBS (rutherford backscattering spectroscopy) measurements, 30 nm-thick Ta2O5, TaOxNy films were also deposited.

After deposition, wafers were annealed with an O2 plasma at 430°C and 700°C. In the case of the wafers annealed at 430°C, an additional annealing process and wet reoxidation were performed at 450°C for 10 min to improve film quality. This also reduced the leakage current.

Finally, Pt or TiN films were deposited by evaporation or by CVD to form the top electrode of the MOS capacitor.

Results and discussion

Figure 1 shows the atomic composition (measured by RBS) of the 30 nm-thick TaOxNy film deposited with 25 sccm of NH3. As expected, nitrogen incorporation was detected in the TaOxNy films. The Table shows that the composition of nitrogen increases, and the composition of oxygen relatively decreases in the TaOxNy films with an increasing NH3 flow rate. Therefore, we find a possibility that nitrogen substitutes for oxygen sites.

Table. Composition of As-Deposited TaOxNy Films
NH3(sccm)Ta (at%)O (at%)N (at%)O/Ta
030.469.62.29
1522.056.022.02.55
2522.052.026.02.36
3522.051.027.02.32

In addition, compared with the O/Ta ratio of as-deposited Ta2O5 and TaOxNy, the O/Ta ratio of TaOxNy is higher than that of Ta2O5. Now we can think of a model in which nitrogen plays the role of a catalyst, with oxygen being filled to the vacant sites of tantalum pentoxide films.

However, it has been reported that nitrogen in films disappears by outgassing, and then the TaOxNy phase changes to the Ta2O5 phase with less than 3% of nitrogen when TaOxNy films are annealed at high temperature.6 If the nitrogen makes a stable chemical bonding with tantalum, it is possible to make phases like Ta3N5, TaN or TaON. If the TaOxNy films change, even a little, into conductors such as Ta3N5 or TaN, the films will have a higher leakage current than Ta2O5. But, as shown in Figures 2b and 3, the leakage current of TaOxNy films is lower than that of Ta2O5. So we believe that TaOxNy films will have the structure of nitrogen-doped Ta2O5 involving nitrogen in interstices or voids.

3. Leakage current behavior of 8 nm-thick Ta2O5 and TaOxNy films annealed with O2 plasma at 700°C.
As shown in Figure 2a, after deposition, anneal with O2 plasma at 430°C, wet reoxidation at 450°C and then deposition of the Pt electrode, the capacitance of TaOxNy films is slightly higher than that of Ta2O5. As shown in Figure 2b, the leakage current of TaOxNy is lower than that of Ta2O5. Therefore, we believe that the better leakage current of TaOxNy film is due to the lower oxygen vacancy.

The Table shows that the O/Ta ratio of films increases from 2.29 to 2.36 when O2 gas changes to NH3 gas. That is to say, during the deposition of TaOxNy film, nitrogen in NH3 plays a catalytic role in deposition, then decreases the concentration of O vacancy in TaOxNy, which results in better leakage characteristics of TaOxNy compared with Ta2O5. Figure 3 shows the plot of leakage current of 8 nm-thick films of Ta2O5 , TaOxNy annealed with O2 plasma at 700°C. The leakage current of Ta2O5 film is 8 × 10-8 A/cm2 while that of TaOxNy is 7 × 10-9 A/cm2 at -1.0 V. When compared with the results shown in Figure 2, we know that the better leakage current of TaOxNy can be taken with the higher temperature, which is attributed to the increased oxidation of interface poly-Si during post-treatment.

Conclusion

We have investigated capacitors made with high-quality TaOxNy films, and compared them with ones made with traditional Ta2O5. The TaOxNy films deposited using Ta(OC2H5)5 and NH3 show a higher dielectric constant and lower leakage current than Ta2O5 films. We were able to obtain TaOxNy film with a current of <1 × 10-8 A/cm2 and cell capacitance of ~10.0 fF/cm2 at
-1.0 V with the optimized O2 plasma treatment. We believe this leads to a promising alternative to Ta2O5 films for capacitor dielectrics in ULSI semiconductor devices.

Youngho Lee is a senior engineer at Jusung Engineering, focusing on development of UHV-CVD and MOCVD systems. He has bachelor's and master's degrees in materials science from Seoul National University.
e-mail: yh_lee@jseng.com

Munsu Park is an assistant manager at Jusung Engineering, focusing on development of MOCVD systems. He has a bachelor's degree in electrical engineering from In-ha University.
e-mail: ms_park@jseng.com

Juho Song is the senior managing director of the R&D center at Jusung Engineering. He has a B.S. in electronic engineering from Kyungpook National University, and an M.S. and Ph.D. in electrical engineering, focusing on solid-state electronics, from Louisiana State University. He has worked in the semiconductor industry for 10 years and served as a faculty member at Changwon National University for two years.
e-mail: juho_song@jseng.com


REFERENCES
  1. H.K. Kang, K.H. Kim, Y.G. Shin, I.S. Park, Tech. Dig., Int. Electron Devices Meet. 94, p. 635.
  2. K.W. Kwon, I.S. Park, D.H. Han, E.S. Kim, Tech. Dig., Int. Electron Devices Meet. 94, p. 835.
  3. Y. Takaishi, M. Sakao, S. Kamiyama, H. Suzuki, Tech. Dig., Int. Electron Devices Meet. 94, p. 839.
  4. G.B. Alers, et al, Applied Physics Letters, Vol. 72, 1998, p. 1308.
  5. J. Lin, N. Massaki, A. Tsukune, M. Yamada, Applied Physics Letters, Vol. 74, 1999, p. 2370.
  6. H.S. Jung, K.J. Im, H.S. Hwang, D.Y. Yang, Applied Physics Letters, Vol. 76, 2000, p. 3630.

Email
Print
Reprint
Learn RSS

Talkback

We would love your feedback!

Post a comment

» VIEW ALL TALKBACK THREADS

Related Content

Related Content

 

By This Author

There are no other articles written by this author.

SPONSORED LINKS



 
Advertisement
SPONSORED LINKS

More Content

  • Blogs
  • Podcasts
  • Videos

Blogs

Podcasts

Videos

Advertisements





NEWSLETTERS
Plug in and get the latest SI news, trends and industry updates delivered free, directly to your inbox!

SI NewsBreak and Special Reports (Weekdays)
Wafer Processing Report (Monthly)
Lithography Report (Monthly)
Metrology Report (Monthly)
Clean Processing Report (Monthly)
Packaging Report (Twice Monthly)
©2008 Reed Business Information, a division of Reed Elsevier Inc. All rights reserved.
Use of this Web site is subject to its Terms of Use | Privacy Policy
Please visit these other Reed Business sites