For those of you with particularly deep insight into pdms, i was wondering if there's a treatment method on the pdms surface or on the surface of something like mylar (transparency film), that reduces the sticky factor of the pdms when attached to this mylar. I'd like the pdms and mylar surface to still be optically transparent in the visible after the particular evaporation or chemical treatment method.
i learned from a colleague that kapton can be chemically placed on the pdms but this compromises the optical property somewhat & the lifetime is poor (~a few weeks).
We have had several reports of smells in the lab this week during the evening hours. Please continue reporting smells you notice.
I would also ask you to send me a description of the smell, time, location of even minor smells. We are trying to see if there is some pattern that will help us find the cause.
One of the smell was reported near the LPCVD furnaces so we are going to clean/inspect more of the LPCVD exhaust/burnbox assembly. We clean the burnbox every week but tomorrow we will clean the manifold were the individual exhaust lines join and enter the burnbox. This extra cleaning was scheduled for the shutdown but the reported smell in our area has us looking at everything.
Again, please continue reporting anything you notice.
Date: Thursday, September 29, 2011 Time: 9:30 AM Location: Paul Allen Auditorium (Formerly CISX-101) Title: Novel Phototransistors for Optical Interconnect
Abstract: Interconnect is one of the major problems in high performance silicon chips in terms of latency, power and bandwidth. Optical interconnect has potential benefits to solve such problems. Because of very high carrier frequency, optical interconnect avoids the resistive loss physics of metal wires, thus high bandwidth is achievable. However, energy/bit is still high compared to conventional copper wire in short on-chip/off-chip interconnects. While lasers and optical modulators promise power requirement for on-chip optical interconnect lower than copper wire, receivers still consumes high power. To decrease receiver-end power consumption, receiver-less scheme had been proposed. To realize this, a device complimentary to photodetector is needed.
First part of this talk presents the first complimentary device to photodetector operates at 850nm laser with output current reduction with incident light. Experiment result successfully showed voltage shift of Id-Vg curve, which is signature of phototransistor operation, as big as 0.7V. It also presents the first wavelength tunable complimentary to photodetector operates at optical communication wavelength range. These phototransistors have potential to replace current optical receiver circuit and reduce power dissipation to meet power requirement.
Second part of this talk presents photodetector utilizes phototransistor concept. Responsivity higher than 700A/W is experimentally shown with 150nW of incident light in 850nm wavelength, which is higher than external quantum efficiency of 1000. In addition, small output capacitance of ~2fF/µm and high scalability are verified from simulation. With these characteristic, this phototransistor can replace conventional photodiode, and further decrease power dissipation of optical interconnect.
As part of my roles and responsibilities, I am planning to send out a bimonthly, or even monthly, note about what is happening at SNF. I will hopefully keep it brief and informal, but also informative.
Attached is the first version. Included are comments on: Safety Lab Members meeting Renovation and associated fab shutdown Dec 14 to Jan 31. New tools being installed New Org chart for SNF
Please feel free to come see me if you have questions or inputs.
Regards,
John B
_______________________ John Bumgarner, PhD Operations Director SNF
At this point there is still gas in the lines, but it will run out. Be sure to monitor your gas flows closely. A replacement cylinder has been ordered and should be here in a week.
The Berkeley MEMS etch tables, available on the SNF website if you dig deep enough, list 50C H2O2 and Transcene NiCr etch, as well as harsher recipes, for Ta. I think it should be easy to dry etch as well. What are you stopping on?
jim
From: Arunanshu Roy <amroy@stanford.edu> To: labmembers@snf.stanford.edu Sent: Monday, September 26, 2011 9:07 PM Subject: Ta wet etching
Hi,
I was wondering if anyone has experience with wet etching of thin Ta films. I only need to etch 5 nm and I don't mind a slow etch rate. Most references I found online recommend using a concentrated solution of HF and HNO3 but I would prefer something milder, especially since I am not looking for a high etch rate.
I was wondering if anyone has experience with wet etching of thin Ta films. I only need to etch 5 nm and I don't mind a slow etch rate. Most references I found online recommend using a concentrated solution of HF and HNO3 but I would prefer something milder, especially since I am not looking for a high etch rate.
Title: "Photoacoustic Imaging using Capacitive Micromachined Ultrasonic Transducers (CMUTs)" Date: Sept. 30, 2011 (Friday) Time: 2:15 PM (Refreshments served at 2:00 PM) Location: CISX Auditorium
Abstract: Photoacoustic imaging is a relatively new medical imaging modality with tremendous clinical potential. In this talk, I will explain the 'photoacoustic effect' – the conversion of short light pulses into ultrasonic waves – and how it allows us to obtain structural/functional information with high spatial resolution and sensitivity. I will present details of a mechanical scanning photoacoustic imaging instrument that I developed including integrating the optics, acoustic transducer, electronics and software. Using this instrument a number of unique photoacoustic molecular contrast agents have been designed and tested to visualize tumors. I will then introduce capacitive micromachined ultrasonic transducer (CMUT) arrays as a possible technology platform to perform 3-D volumetric photoacoustic imaging with high sensitivity. CMUTs offer a number of advantages over traditional piezoelectric ultrasound transducers including batch fabrication, wide bandwidth transducers, ability to fabricate 2-D arrays with arbitrary geometries and tight integration with electronics. I will show photoacoustic imaging results in tissue phantoms using CMUTs and preliminary results of real-time photoacoustic imaging. Finally, I will introduce a new microfabrication process to make CMUTs that offers better reliability against dielectric breakdown and increased design flexibility.
Just a few days before EE412 begins anew, with Profs. Olav Solgaard and Roger Howe at the helm. Class meets Tuesdays, starting September 27, at 4 pm in the Allen 101 conference room.
If you aren't familiar with this, EE412 is a team-project-based course to develop, characterize and document processes of value to the SNF community (and your own research!) You can select from available projects or work with staff member or other mentor to come up with their own. You get not only course credit, but lab time and supplies to support your projects -- and of course the satisfaction of contributing to the lab community knowledge base.
This Fall, we are pleased to offer new tools (the Fiji ALD, Intlvac PVD sputter, and Intlvac evaporator) in addition to the usual suspects (EV Spraycoater, STSetch2, etc.)
-- Mary X. Tang, Ph.D. Stanford Nanofabrication Facility Paul G. Allen Room 136, Mail Code 4070 Stanford, CA 94305 (650)723-9980 mtang@stanford.edu http://snf.stanford.edu
Stanford University PhD Oral Defense - Department of Electrical Engineering Date: September 26, 2011 (Monday) Time: 1:00 PM (Refreshments at 12:45 PM) Location: McCullough 335 Title: Molecular Junctions of Self-Assembled Monolayers with Universal Soft Contacts
Abstract:
As transistor scaling following Moore's Law faces mounting obstacles from device leakage and heat dissipation, numerous disruptive "post-silicon" technologies are being explored. Devices consisting of a single or few organic molecules have been proposed as extremely scalable, low-power logic and memory elements. The diversity and consistency of molecules that may be obtained through chemical synthesis has resulted in numerous exciting device proposals. However, in the many studies of transport through molecular systems that have been presented in the literature, the difficulty in establishing contacts to a desired molecular system has arisen as a complication. There is an emerging consensus that the ideal molecular device geometry should consist of a self-assembled monolayer (SAM) sandwiched between two electrodes to average out molecule-contact variations and ensure consistent behavior with scaling.
This talk will begin with a discussion of the general characteristics of a scalable molecular device, and then briefly review recent device designs demonstrated in the literature, with an emphasis on the limitations of the various contact materials and deposition methods. A novel, scalable, high-yield molecular junction incorporating an isolating dielectric layer and a soft polymer top contact will then be presented. Studies of transport through alkanethiol SAMs with a variety of terminations formed on gold substrates will be presented. Models explaining the molecular length-decay and unique temperature dependence behavior will be demonstrated.Studies of devices incorporating platinum and silver substrates will be presented to provide insight into the packing behavior of the SAM. The limitations of devices with soft contacts will be explored with the presentation of studies of conjugated molecular systems and molecules with redox-active organometallic terminal groups. The demonstration of scalable, high-yield molecular junctions is an important step in the development of computing devices beyond Moore's Law.
It is safe to go back in to the fab. The alarm was from the Ozone detector monitoring the exhuast on the Fiji ALD systems. Working with the fire department, using their portable detector we could not find a trace of ozone in the fab area. As a precaution, I have shut down the Fiji systems until we can take a look at the system and the detector on Monday. Very low concentrations of ozone is generated within the system and exhausted through the scrubber and should never reach the fab air.
Thanks to the lab member community which called the duty phone and allowed us to quickly.
On Monday morning (9/19) the double doors near the ALD and GaAs wet bench will be open in order to roll out the AG4108 and AG4100. In the near future you will see more tool movement as we prepare the fab locations for the new etch systems which will be arriving this year.
If you have sensitive samples, please plan your use of the ALD and GaAs wet bench accordingly.
Stanford University PhD Oral Defense - Department of Electrical Engineering
Speaker: Byoungil Lee
Date: September 16, 2011 (Friday) Time: 1:00 PM (Refreshments served at 12:45 PM) Location: Paul Allen Auditorium (CISX-AUD)
Advisor: H.-S Philip Wong
Title: Fabrication and Characterization of Nanoscale Resistance Change Memory Abstract:
As the current charge-based memory technologies such as DRAM and Flash Memory are facing their fundamental scaling limits, new types of memory technologies such as spin-torque-transfer RAM (STTRAM), phase-change memory (PCM), and metal-oxide resistance change memory (RRAM) have been actively explored. Among these candidates, RRAM devices utilize the resistance change property of the metal-oxide films, and have shown promising results such as fast switching speed, low programming current, and good CMOS compatibility. Despite the potential advantages, they suffer from the critical issues that must be addressed for the high-density memory application, which includes existence of forming-process, large variation in switching characteristics, and cell selection issue in the cross-point structure.
In this talk, I will first demonstrate the process and characterization methodology for RRAM devices to verify the characteristics and potential issues in the nanoscale regime down to 50X50 nm2 size. Unipolar switching memory cells using NiO thin films were fabricated by ebeam lithography process. The device characteristics and the scaling trends will be discussed.
Next, I will present two novel RRAM structures developed by the process and characterization methods established in the previous section. In the first structure, both top and bottom electrodes are processed prior to the metal-oxide deposition, and therefore, contamination-less programming region is achieved in the memory cell. Moreover, the novel structure features more confined conduction path compared to the conventional structure, which results in narrower low-resistance distribution. The process flow and characteristics of the nanoscale NiO memory cells using this structure will be presented.
In the second structure, an oxide-to-oxide interface is incorporated inside the memory cell in order to utilize the initial leakage and eliminate the high-voltage forming process. Due to this initial leakage, the as-fabricated AlOx memory cells exhibit low-resistance sates, while the conventional AlOx devices show high-resistance initial states.The new memory cell does not require the high-voltage forming process since the as-fabricated cells can be switched to high-resistance states with a regular RESET voltage.
Finally, application of the forming-less AlOx cells to the complementary resistive switch (CRS) scheme will be discussed using the measured cell characteristics. The CRS scheme was proposed in 2010 as a potential solution to the high-density cross-point memory arrays. The novel memory cells with forming-less property demonstrate promising results that enables the CRS operation using bipolar switching RRAM cells.
