Electronic Devices |


Advanced Lithography |
Nanofabrication |
Nanomechanical Devices |
   Carbon Nanotube Electronics
nano-MOSFET
Quantum Computing
Droplet-on-Demand Technology
Large Area Nanowires & Nanodots
Nanomechanical Resonators

 

Electronic Devices

Carbon Nanotube Electronics | Integration of Carbon Nanotube Devices with Silicon Technology

It has been shown extensively that semiconducting carbon nanotubes (CNTs) can be made into transistors. However, it is unlikely that CNT devices will become, in the near future, the basis of the integrated circuit technology. The introduction of CNT devices into electronic products will probably happen through their integration with silicon-based integrated circuits. Such a hybrid technology benefits from the advantages of both materials, and could enable novel devices and circuit design.

This project involves catalyst-mediated chemical vapor deposition (CVD) growth of single-walled carbon nanotubes on a MOS integrated circuit fabricated using the UC Berkeley Microlabys baseline process. Integration with NMOS devices has been successfully demonstrated, and a similar work with PMOS devices is in progress. As part of this project, we aim to develop a facility to perform CVD growth of carbon nanotubes in the Microlab. (Yu-Chih Tseng and Ali Javey[1] -Professors Hongjie Dai[2] and Jeffrey Bokor)

[1] Department of Chemistry, Stanford University
[2] Department of Chemistry, Stanford University


SEM image of CNT integrated in silicon circuitry.

 

nano-MOSFET | Study of the Electrical Transport in Nano-Scale Schottky Contacts

One of the great concerns about nano-scale devices is how to make good (ohmic and low resistance) and robust contacts to them. Conventional low resistance ohmic contacts are made of metal silicides on top of heavily doped semiconductor regions. The contact resistance is inversely proportional to contact area. In nano-scale devices, the contact size is on the order of nano-meter or smaller. Contact resistance could be a serious limitation to performance. It is well known that electrical transport through a large size Schottky contact is significantly limited by a large barrier height (Thermionic emission). In most systems, the barrier height does not depend on metal work functions because the surface Fermi level is pinned around the charge neutrality level by surface dipoles, which are seen as a consequence of metal induced gap states near the interface. At nanometer scale, a low resistance Schottky contact can be obtained only if we have a low Schottky barrier or the barrier is thin enough to allow significant tunneling current. Recently, theoretical studies have suggested that the surface Fermi level may not be strongly pinned in nano-scale Schottky contacts. Thus, we may have the opportunity to control the Schottky barrier height by using metals with different work functions.

We are seeking theoretical and experimental understanding of the transport in nano-scale Schottky contacts. Currently, we are studying the contacts to carbon nanotubes. Achieving highly transparent metal-nanotube contacts is our final goal. (Shiying Xiong)

Quantum Computing | Spin-state Readout for Silicon QC

Spin states of bounded donor electrons at low temperatures can be utilized as qubits in various silicon-based quantum computation schemes. Two crucial aspects of such schemes are the controlled placement of donor atoms in silicon and reliable electron spin-state read-out.

Single ion implantation [1] or scanning tunneling microscopy (STM) based hydrogen lithography [2] can be used to achieve donor placement with atomic precision. The latter is based on the electron stimulated desorption (ESD) of a hydrogen-terminated silicon surface with a STM tip, and subsequent exposure to PH3 for phosphorous placement. However, the fabrication of electrical contacts for these nano-scale donor patterns is challenging, due to the limited field of view of the STM used for ESD. In this project, pre-implanted As contact arrays are designed and fabricated for the electrical characterization of these 1 and 2D donor patterns.

Spin-state read-outs for silicon-based quantum computers can potentially be achieved by using field-effect transistor (FET) architectures, with the donor atoms positioned in the channel of the FET. The donor electron spin states can be detected by exploiting the difference in singlet and triplet scattering cross sections of the spin-polarized source-injected channel electrons and that of the bounded donor electrons [3]. Accumulation mode field-effect transistors have been fabricated for this purpose, and characterization and measurements are currently in progress.

[1] T. Schenkel et al., J. Vac. Sci. Technol. B, Vol. 20, No. 6.
[2] T.-C. Shen et al., J. Vac. Sci. Technol. B, Vol. 22, No. 6.
[3] R. N. Ghosh and R. H. Silsbee, Physical Review B, Vol. 46, No. 19.
 
Advanced Lithography

Droplet-on-Demand Technology | Maskless Lithography with Inkjet Printing Method

Writing patterns directly on a substrate with liquid droplets generated by drop-on-demand (DOD) inkjet devices offers a low cost, additive, non-contact, low temperature and data driven process. The existing inkjet systems could only form liquid droplets with a volume larger than several picoliters, which limited their minimum printable size to tens of microns. In our research, we focus on building a monolithic inkjet printhead with an array of inkjet devices and seek to reduce the liquid droplet size to fulfill the requirement of micron scale maskless lithography.

In our previous work, we have fabricated thermal bubble inkjet printheads by wafer-wafer epoxy stamping. The process had poor yield because the epoxy tends to clog the small inkjet channels during re-flowing. We then modified the process with Ni eutectic bonding where nickel silicide is formed at the wafer-wafer interface. The new process is much more reliable, and water droplets varying from 15 to 5 micron meter in diameter have been generated from our printhead.

The wafer bonding process, even though successful, has difficulty in further reducing the droplet size due to the large nozzle membrane and chamber dimensions. In the past year, we have also developed another printhead fabrication process based on a single silicon wafer. In this process, inkjet chambers and refilling channels are formed on top of the Pt heaters by a Ge sacrificial etching process. The walls and ceilings of inkjet chambers are made of thick LPCVD silicon dioxide, while the nozzles are patterned on thin PECVD silicon nitride membranes. The inkjet devices are connected to a fluid reservoir by deep etching through-wafer holes (Figure 1). Using this process, we have been able to generate water droplets as small as 3.5 micron meters (Figure 2), which is about a factor of 100 times smaller in volume compared with what the existing inkjet systems could print. The droplet generation process is also found to be stable, satellite-drop free and uniform in both droplet size and velocity.

An important issue in building up an inkjet printing maskless lithography system is to find suitable materials for pattern formation. Au nano-particles suspended in propylbenzene have been tested with our printhead. Stable, drop-on-demand operation with droplet size around 6 micron meters is verified. Using the material, Au dots with a diameter of about 10 micron meters have been formed on a silicon substrate. In the next step, we will install our printhead into a computer controlled printing system to print process patterns for organic transistor and other micro-scale device fabrication. (Yan Wang)


(a) Array of inkjet devices fabricated by the single wafer process; (b) Top view of a single inkjet device, the nozzle diameter is about 5 micron meters.


3.5 micron meter water droplet generated from test chip of the single wafer process.

 
 
Nanofabrication

Large Area Nanowires and Nanodots | Spacer and Nanoimprint Lithography

High throughput patterning of sub-100nm periodic features on surfaces has been of great interest for a number of scientific and engineering applications, such as sensors, soft X-ray optical device components, electronic circuit elements, or catalysts. In chemical and physical applications such as catalysis and sensors, low-cost periodic patterning is required rather than the arbitrary-shape, patterning capability of e-beam lithography. To satisfy this requirement and overcome the low throughput and high cost of electron beam lithography, a residual conformal film on the side wall of a photolithographically-defined pattern (a so-called ‘spacer’) has been used to generate nanoscopic line features, with a line width that is well controlled by the deposited film thickness [1][2]. If this spacer lithography is used n times in succession, 2n lines can generated from a single lithographically defined line. With nanoimprint technology, Pt nanowires and Pt nanoparticles can be fabricated by the spacer lithogrpahy for catalysts and nanowires composed of other materials can be used for chemical and biosensors. However, these techniques are limited to closed loop shaped line patterns and are not able to produce discrete dot patterns since spacers are formed along the side walls of the original features. In this work, we develop an advanced method for the conversion of spacer nanowires into dots by Mold-To-Mold Cross Imprint (MTMCI) based on the spacer and nanoimprint lithogrphy. First, 15nm wide silicon nanowire molds with 250nm pitch were fabricated by deep UV lithography and spacer lithography. Then the conversion of the wire pattern into a dot pattern by redefining an nanowire imprint mold with another nanowire imprint mold with perpendicular arrangement. This silicon nanodot array is to be used as an imprint mold for patterning well ordered nanoscopic metal islands with uniform size distribution. These, in turn, are useful for catalyst research or surface enhanced Raman spectroscopy. (Sunghoon Kwon and Xiaoming Yan[4])

[1] Choi, Y.K.;King, T.J.;Hu, C. IEEE Electron. Device Lett.,2002,46,1595
[2] Choi, Y.K.;Zhu, J., Grunes, J.;Bokor, J.;Somorjay, G.A.; J. Phys. Chem. B 2003, 107, 3340
[3] Kwon, S; Yan, Xiaoming; Contreras, A; Somorjai, G.A., Liddle, J.A., and Bokor, J, NNT04, 2004
[4] Lawrence Berkeley National Laboratory


(L) Schematic of Mold-to-Mold Cross Imprint (MTMCI) process for converting nanowires into nanodots. (R) 15nm wide, 150nm high silicon nanowires with 250nm pitch by DUV based spacer lithography.


(L) SEM of 30nm silicon nanodots converted from silicon spacer nanowire. (C) This large area silicon nanopillar is used as imprint mold. (R) Imprinted PMMA nanohole pattern using the mold from MTMCI.
 
Nanomechanical Devices

Nanomechanical Resonators |

Development of Scanning Force Microscopy Methods for the Chacterization of Nanomechanical Resonators

The development of new nanomechanical resonators for RF applications requires the availability of high resolution nanometrology tools for the characterization of the electromechanical properties of prototype devices. In this project we take advantage of the high resolution and versatility of atomic force microscopy (AFM) further improved with the combination with a lock-in detection method to develop such a nanometrology tool. This technique allows the determination of the resonance frequencies, quality factors and acoustic mode shapes of the resonators with sub-nanometer scale resolution. (Alvaro San Paulo and Xuchun Liu)

Scanning Microscopy Probe for Nanomechanical Resonators

This project is one part of a large project named Integrated Microwatt Transceivers at BSAC. The purpose of our project is to characterize the nanomechanical resonators by interferometer or atomic force microscope (AFM) combined with optical actuation. So far, an interferometry-based system was built to characterize the MEMS resonators, and paticularly the Agilent's FBAR. The system can detect the RF movement in the sub-nanometer range. With this system, the power-dependent response and frequency-dependent response of FBAR were measured and the results are consistent with the measurement made by AFM. In the next phase, a stage will be included into our system to detect the mode shape of FBAR and other MEMS resonstors. (Xuchun Liu, Cornell University)