Industry News: Superatoms, Chrome Use Decline, Double Patterning, Solar Cell Profiles
Staff -- Semiconductor International, 9/1/2008
'Superatoms' Open Window to Nanoparticle Energy
The principles behind the stability and electronic properties of miniscule nanoclusters of magnetic gold have been described by researchers from the Georgia Institute of Technology (Atlanta), Stanford University (Palo Alto, Calif.), the University of Jyväskylä (Finland) and Chalmers University of Technology (Göterborg, Sweden). Gold and sulfur atoms aggregate in specific numbers and extremely symmetrical geometries, and can mimic the chemistry of single atoms of a different element. Researchers discovered that these clusters are stable because they behave like "superatoms" and exhibit a "divide and protect" bonding structure.
According to Robert Whetten, a professor at Georgia Tech's School of Physics and School of Chemistry and Biochemistry, although gold nanoparticles are widely used in many fields, nobody fully understood their molecular structures and physiochemical properties. Researchers use gold nanoparticles because of their stability and distinct optical, electronic, electrochemical and biolabeling properties.
In 2007, Stanford researchers initially reported the first determination of a 102-atom gold cluster. Their X-ray structure study showed that pairs of organic sulfur (thiolate) groups extract gold atoms from the gold layer to form a linear thiolate-gold-thiolate bridge while weakly interacting with the underlying metal surface. This formed a "protective crust" around the nanoparticles, contradicting the belief that the sulfur atom merely sat atop the uppermost gold layer, bound to three adjacent metal atoms. This was the first time that atomic coordinates — the location of all of the atoms in one of these clusters — became available.
| Space-filling (a) and ball-and-stick (b) representations of a 102-atom gold nanoparticle. A 79-atom gold core surface with a 23-atom protective layer is shown in c and d. Close-ups of protective layer units (e) and a 79-atom core (f, g). (Source: Georgia Institute of Technology) |
The 102-atom gold cluster "superatom" has a core of 79 atoms arranged into a truncated decahedron. Around this core, 23 gold atoms joined the thiolates. Results confirmed the "divide and protect" structure, showing that gold atoms divided themselves into two groups — those making up the metal core and those that protected it.
In the cluster, each gold atom donated one valence electron. Forty-four of these electrons were immobilized in bonds between gold atoms and thiolates, leaving 58 to fill a shell around the superatom. Because the cluster does not have to add or shed electrons, its structure is stable. A major energy gap was also exhibited. "The calculated energy gap between the highest occupied molecular orbit and the lowest unoccupied molecular orbital states for the 102-atom compound was significant: 0.5 eV. Metals typically have a gap of zero," Whetten said, adding that this indicates an atypical electronic stability.
There had been speculation for years about the nature of the chemical bonding, structure and electronic principles holding such a structure together. Nobody fully understood how common sulfur-containing molecules bind to metallic gold; it was not known that such a large metallic cluster could attain molecular precision instead of just forming into an arbitrary aggregate. Amazingly, little was understood of the common and practical metallic chemistry of films like gold or silver.
The research opens a door into another startling possibility. "The 58 electrons in this complex behave as if orbital details didn't matter and as if they were part of a large superatom that follows its own electronic shell structure," Whetten said. "These electrons fall into a nice pattern and open a large gap between the highest occupied energy level and the lowest unoccupied level. Similar to what one finds in the Periodic Table if you go from neon with its 10 electrons to sodium with 1; there's greater electronic stability."
— Alexander E. Braun, Senior Editor
Chrome Going the Way of the Dodo Bird?
Chrome, which has served the photomask industry well for decades, may gradually be displaced by a better metal: molybdenum silicide (MoSi).
Franklin Kalk, CTO at Toppan Photomasks Inc. (Round Rock, Texas), said he knows of two semiconductor companies that plan to use MoSi for 32 nm logic production starting next year. At leading-edge design rules, MoSi "just works better than chrome, which is a pretty good reason to use it," Kalk said. The composition of the MoSi alloy produces a sharper sidewall with less line edge roughness (LER).
Although MoSi has been used in phase-shift masks (PSMs), binary masks, thus far, have stuck with chrome. However, the immersion lithography optics with numerical aperture (NA) >1 do not work as well with PSMs as with binary masks. More importantly, MoSi is easier to etch than chrome. For decades, wet etching was used with chrome, but as modern dry etchers came into use, the mask shops found chrome to be difficult to etch. "Moly-Si has a much stronger ionic component than chrome, so the etch step is less of a chemical process," Kalk said. "Overall, moly-Si is a much easier material to deal with, with better defect qualities than chrome and better sidewall angles."
The lower rate of defects means less time repairing masks, reducing the turnaround time needed to make a defect-free mask.
| For chrome absorbers, dry etch conditions have a significant effect on the edge slope and quality. (Source: Grenon Consulting) |
Brian J. Grenon, president of Grenon Consulting Inc. (Colchester, Vt.), said he has been advocating MoSi as a replacement for chrome absorbers for many years, and gave an entire seminar on the problems with chrome at Sematech in 2003. MoSi's attributes include better CD control, improved image placement, more accurate metrology, easier inspection and higher resolution. MoSi "is basically a purer material that is easier to process," Grenon said. "The reason is compositional, which has to do with the basic properties of the metal, allowing more robust process conditions. Dry etching is more precise because the film chemistry is less amorphous."
Advanced Micro Devices Inc. (AMD, Sunnyvale, Calif.) is likely to be among the first companies to use MoSi, Grenon said, noting that AMD and Toppan, along with Infineon Technologies, are members of the Advanced Mask Technology Center in Dresden, Germany.
Kalk said two companies will begin prototyping late this year and move to volume use of MoSi masks in 2009 for 32 nm logic production. The memory makers, which measure their design rules by the lithographic half-pitch, will consider MoSi masks for the 45 nm node.
Toppan, which is wholly owned by Toppan Printing Co. (Tokyo), developed its MoSi process with Shin-Etsu Chemical Co. (Tokyo). In June, Toppan Printing announced that it was ready to begin volume production of 32 nm masks.
Asked if chrome will eventually become obsolete, Kalk took pause and then said, "I think yes, it will be, frankly. There are a bunch of reasons why. Fundamentally, the material is being processed in ways it was never intended. Chrome is difficult to etch in a modern dry etch W. It is being used in ways where alternative materials can provide clear advantages, so it makes sense that we would replace it."
— David Lammers, News Editor
Double Patterning Battles Cost, Complexity
This year's Sokudo Lithography Breakfast Forum at SEMICON West focused on the challenges of double patterning. To be sure, double patterning is not without its challenges, but it nonetheless is positioned as the most promising technology for 32 nm patterning, and likely 22 nm as well.
Senior AMD Fellow Harry Levinson set the stage for the remainder of the breakfast discussion from suppliers. As he put it, double patterning is indeed double the trouble, but the motivation to take that route is pretty simple. The industry has traditionally scaled to finer linewidths simply by shrinking the wavelength of the lithography source — from arc lamps to 365 nm to 248 nm to the current leading-edge 193 nm ArF source. Shrinking that wavelength further to 157 nm was not ultimately advantageous for the industry, and the big jump to the extreme ultraviolet (EUV) wavelength of 13.4 nm doesn't look to be happening anytime soon.
| According to Senior AMD Fellow Harry Levinson, although double patterning is indeed double the trouble, the motivation for using it is simple. |
So the industry generally agrees that double patterning is needed to bridge the gap until the next source wavelength change. "However, it's going to be very difficult," Levinson noted. Even with single patterning, parameters such as overlay, CD control and line edge roughness (LER) scale ~0.7× node to node.
Tighter overlay control is key for making double patterning viable, but another considerable concern is cost. Just how much overlay or cost issues there are depends, in part, on the type of double patterning scheme used. Several of the morning's presenters highlighted the main possibilities. Although there are variations, they are generally considered to be spacer, double patterning (litho-etch-litho-etch) and double exposure (litho-freeze-litho-etch).
Spacer double patterning — called various things, such as self-aligned double patterning by Applied Materials — does not have overlay concerns because it has only one critical exposure, explained Chris Ngai. However, as several speakers pointed out, it does have serious cost issues. If you're trying to print an SRAM cell, spacer is probably a good way to go, noted ASML's Bob Socha. But the cost-of-ownership for spacer technology is considerably higher than other double patterning techniques, he said, adding that not all designs can benefit from the spacer technology.
In fact, which double patterning scheme is chosen is a game of trade-offs — particularly between process complexity and materials complexity, Levinson noted. For example, the standard litho-etch-litho-etch scheme uses materials that are readily available today, but it takes a big hit in throughput and process complexity. Litho-freeze-litho-etch, on the other hand, sounds like a great simplifier, allowing the process to stay in the litho tool, but the idea is based on materials that are not yet available.
There seems to be a lot of hope placed on the pattern freezing technique, with several materials suppliers working on solutions. JSR Micro's Mark Slezak presented the latest on his company's resist freeze process, including the etch results done at IMEC. JSR Micro has successfully patterned 32 nm logic patterns, as well as contact hole patterns, with its resist freeze process.
Although the freeze process puts more pressure on the materials side, Slezak said, the fewer steps required in patterning can add up to significant cost savings. The idea of the technology is to image the first material, apply a freeze, and then the frozen first resist is put through the subsequent steps. The trick is coming up with a frozen resist that is resistant to subsequent steps.
— Aaron Hand, Executive Editor, Electronic Media
Scanning Probe Microscope Works Out Solar Cell Profiles
There is an ongoing effort by the photovoltaics (PV) industry to use polymers to create solar cells because this would not only enable processing to be done at room temperature and ambient conditions, but would also allow the application of processing techniques, such as various forms of printing, that are relatively inexpensive when compared with processing silicon. However, for the greatest efficiency possible, it is not enough to have the chemistry and chemical composition of these different materials well in hand, but to also know as much as possible about their nanostructures before they are all brought together in the finished cell.
Professor David Ginger and his students at the University of Washington's Department of Chemistry (Seattle) have been researching this problem, using scanning probe microscopy (SPM) to solve it. "The idea is that if we can measure the PV and some of its more fundamental electronic properties with this level of resolution, we might be able to find out what limits performance and how to improve it," Ginger said.
| Several scanning probe microscopy techniques are useful to study nanostructured organic solar cells. Illustrated (center) are atomic force microscopy’s (AFM) two primary modes. In non-contact mode (A), a vibrating cantilever measures electrostatic forces between the tip and surface. The photo-induced charging rate in a solar material is mapped using a pulsed light source (top left). In contact mode (B), a static cantilever measures tip sample currents to map photocurrent generation in solar materials (bottom right). Contact mode can be used to extract quantitative values of local charge carrier mobility through analysis of current voltage curves taken by the AFM tip. The figure at the bottom left shows numerical simulation results for current density (J) and potential (V) distributions generated in a solar material under a conductive AFM tip. (Source: University of Washington) |
For quite some time, Ginger and his group have used different kinds of (SPM) — from electrostatic force microscopy (EFM) to conductive atomic force microscopy (c-AFM) — to develop maps of the photocurrent distribution in these solar cells to be able to determine the size and kinds of domains that the photocurrent originates from. "That's really just step one — getting a more microscopic picture," Ginger said. "You don't just want to know where the photocurrent comes from — you also want to find out where these charges are being created, where they are being lost to recombination, and how charge transport varies region to region in these extremely heterogeneous thin films. We're trying to look at transport differences."
The hurdle is that there is a quantitative issue with how to interpret that data. Several researchers who have previously attempted it report that if SPM techniques are used to measure the charge carrier mobility, it is soon realized that the values obtained can be up to two orders of magnitude higher than they should have been. Ginger's group has been trying to solve that problem and determine whether it is possible to use the method to get quantitative, rather than just qualitative, measurements. "The fact that there's a two-orders-of-magnitude discrepancy between what the known mobility values in these materials are and what's coming out of the scanning probe measurements made us hesitate a bit before using this technique," Ginger said, "because even if it gives a relative value, you must wonder whether it's that accurate since it is so much larger than what you thought it would be."
The researchers hypothesized that the differences lay primarily in the experiment's electrode geometry. After doing the numerical work to check that possibility and experimenting to do comparisons with these numerical simulations, it was discovered that, indeed, most of the variation could be attributed to the very sharp needle plane geometry present in the conducting AFM experiment vs. planar geometry, where bulk photodiode measurements are performed. "When you compare these two, you realize that you must scale your experimental conductive AFM results based on the film thickness and the tip diameter," Ginger said. "When you take those two factors into account, you collapse what looks like a whole scatter plot of conductive AFM data for mobility values on different films taken with different tips onto a nice single line that agrees with a fundamental underlying film property."
This breakthrough enabled the group to reconcile why a certain tip gave a different measurement on a certain film, and that film gave a different value than another. Now it is possible to obtain reasonable carrier mobility, enabling the use of this technique to locally map quantitatively changes in the mobility from region to region in heterogeneous thin films. The next step is to use the technique and its results to characterize local variations, because it is now known that, quantitatively, it provides the right values.
By understanding these local variations in transport, it will be possible to correlate them with inter-device performance changes with processing and possibly come up with a more rational optimization of film processing, which is currently being done in a somewhat hit-or-miss fashion. "You try different solvents, different heating conditions, different anneals for different times and different temperatures, and you know that affects the texture and nanostructure morphology of the film," Ginger said. "However, you'd really want to correlate that with a change in a local hole mobility in the domain, for instance."
In the future, the researchers would like to couple these transport measurements with local measurements of carrier generation and carrier trapping and recombination. Once all of these are obtainable, it becomes possible to build up a truly microscopic picture of how these nanostructured organic solar cells work. This would transform what at present is an art into a technology. The overall goal is, of course, to actually develop new polymers that harvest a wider range of the solar spectrum and have greater stability; materials that have better energy level alignments so that not as much of the voltage obtained from the photons that are absorbed is not wasted. However, there still remain many outstanding fundamental research as well as technological — processing and characterization — challenges in the organic electronics and polymer PV field.
The UW researchers know what needs to be done. But as Ginger put it with a grin, "The doing is easier said than done."
— Alexander E. Braun, Senior Editor
