Quick Online Roundup

Intel Speeds Up Silicon Photonics: Who Says Silicon Isn’t Good For Photonics?

Silicon is ubiquitous in the world of microelectronics, and yet, silicon photonics isn’t nearly as well-developed as other materials. There are a number of fundamental reasons for this, the main ones being the fact that silicon has an indirect bandgap, a relatively large bandgap, and is difficult to make electro-optic modulators out of because of its electric susceptibility.

It would be nice, though, to realize useful silicon-based photonic integrated circuits e.g. for optical-based means of communication and data transmission. This necessitates a post in itself, and has been duly added to the list of things that will be blogged about at some indefinite point in the future.

The point is, though, that a number of very smart people have been able to do interesting things with silicon in the context of photonics; for example, a lot of progress has been made recently in developing silicon lasers (see, for example, this or this). Well, here’s another breakthrough:

Researchers at Intel have announced the world’s fastest silicon modulator–an advance that could cut bandwidth costs and make computers run faster and cooler.

MIT’s Tech Review has an easy-to-read article on this; the paper itself is available here.

Scientists Turned Filmmakers

Interestingly (and perhaps, somewhat amusingly), the Pentagon is funding screenwriting classes for scientists-turned-budding-filmmakers, in an effort to enhance the US’s strength in science and technology. As many have previously pointed out, the US appears to be suffering from a dearth of young scientists and engineers - it turns out that rock stars and i-bankers have more sex appeal than scientists, mathematicians and engineers. (What?!) Taking its cues from (among other things) the popular TV-series CSI - which apparently helped boost careers in forensics - the Pentagon’s decided that “by writing and producing movies that have more scientific themes - and more authentic and appealing science protagonists… the US could encourage more young people to pursue careers in plasma physics, molecular biology, and other fields.“

“If I want to watch sports, I can turn on any one of four to 12 channels, 24 hours a day, seven days a week,” says Alvin Chin, a professor of pediatrics at the University of Pennsylvania in Philadelphia… “Imagine if science programming had that kind of presence.”

The article is here. 

This Week’s Science Roundup

This week, there have been some interesting papers dealing with new magnetic materials; using thin-films in new and interesting ways (such as in transistor memory devices and gate dielectrics in carbon nanotube transistors); nanoscale photonics using nanowires and nanotubes; exploring the possibility of creating quantum dots in graphene using electrostatic potential barriers; using scanning tunneling microscopy to look at the Kondo effect in molecules and carrier dynamics in p-n junctions while they’re being operated; figuring out what part of the brain is responsible for our wandering minds; and two interesting applications of quantum mechanics in biology - theoretically considering phonon-assisted tunneling of electrons in elucidating how we smell, and using computational quantum mechanical calculations to study protein splicing. Whew.

Nanoscale/Condensed Matter-Related:
- Hybrid metal-organic materials that are magnetic at room temperature
- Thin-film ferromagnetic devices whose magnetization is modulated via an applied electric field
- A new organic (pentacene) thin-film field-effect transistor (FET) as a possible non-volatile memory device
- Using self-assembled monolayers (SAMs) as the gate dielectric in carbon nanotube FETs
- Nanoscale photonics: nanowire LEDs and nanotube coaxial cables
- Creating quantum dots electrostatically in graphene
- Manipulating the Kondo effect in molecular systems using STM
- Using STM to study carrier dynamics in a p-n junction

Bio-related:
- Why your brain wanders when you’re bored
- Could Humans Recognize Odor by Phonon Assisted Tunneling?
- Studying Protein Cleavage Using Quantum Mechanical Calculations

Neural Networks II

Last week’s subject was McCulloch-Pitts neurons and perceptrons, particularly the Perceptron Convergence Theorem and linearly separable problems. This week builds on that, and this post’s subject is Pigeons, the Credit Assignment Problem, and Backpropagation.

Schön: Where Is He Now?

Nature catches up with some past fraud investigations, and sees where they are today. Among them: Jan Hendrik Schön who, among other things, is well-known for fabricating data that would otherwise have revolutionized the field of molecular electronics, eventually withdrawing 21 high-profile papers from PRL, Nature, and Science.

Schön’s whereabouts are now unknown, and he has declined to comment on his research while investigations are continuing (the University of Konstanz in Germany is still assessing the work he did to earn his PhD there in the mid-1990s). But shock waves from the case have hit physicists, mainly in the shape of new research guidelines. Some of Schön’s co-authors were criticized for not spotting their colleague’s misconduct, prompting the American Physical Society to issue new rules stating that co-authors should be accountable for important data in papers they sign off.

Doug Natelson has a post from a while back in which he points out that with Alcatel taking over Lucent, the results of the Schön investigation (basically detailing everything fraudulent Schön had done) are apparently nowhere to be found on the new Bell Labs website - and so, he’s hosting the documents on his own website at Rice. Interestingly, in the comments, Peter Armitage reveals that he heard recently that “Schön had applied for a research job at some lab in Sweden. His CV was complete sans ONLY the papers that had been specifically flagged by the Beasley report”.

Nature Education

The nice people who bring us Nature (and its many progeny) have embarked on a new venture “to develop innovative educational resources and tools for science students and their professors”: Nature Education. As their recent press release states:

Nature Education will take a non-traditional approach to the rapidly-evolving college education market, focusing primarily on creating leading edge, digitally-based, learning solutions in biology, chemistry and physics…

Instructors and students are thirsty for learning environments that move beyond traditional textbooks and even course management systems to provide a highly interactive and personalized experience that simultaneously builds understanding, inspires career and research aspirations, and connects the student to a worldwide community of likeminded thinkers. With its excellent content, brand, global reach, and community of practicing scientists, NPG and Macmillan are superbly positioned to catalyze and capitalize on a radical shift in education.

So what exactly are these “digitally-based learning solutions”? The press release doesn’t say, and a few seconds of Googling don’t turn up too much, either. Is Nature Education a new journal? (Probably not, given the emphasis on breaking with tradition and being digital and all.) An online repository of open-access course materials, à la MIT’s OpenCourseWare? Something more?

I wonder where they’re going with this. It’s hard to say, but it looks like the Nature Publishing Group’s really focusing on being innovative, inventive, and breaking with tradition, in an interesting kind of way. Another example that springs to mind is last year’s peer-review trial (which although widely unpopular was an intriguing debate/experiment). It should be interesting for educators/anyone interested in science education to see what results from this.