Tech Talk

The potential of thin film bulk acoustic wave resonators (FBARs—no sniggering from the back row, please) to measure mass at the nanogram level has marred by one persistent obstacle.

An FBAR membrane is electronically driven to vibrate at a characteristic resonant frequency. When a particle—a protein molecule, perhaps—adsorbs to membrane, the resonant frequency drops; the frequency change is proportional to the glued-on mass. So, measure the frequency drop, measure the mass. And because the frequency shift for a given mass is proportional to the square of the initial resonant frequency, the higher the initial resonant frequency, the more sensitive the measurement and the smaller the mass you can expect to measure.

And now for the obstacle. The FBAR’s acoustic characteristics change significantly with temperature. The speed of sound changes, and so does the thickness of the membrane. And this temperature sensitivity grows along with its measurement sensitivity as the resonant frequency climbs.

Up until now, solutions to this gravimetric problem have required two separate measurements—either measuring the frequency change in two distinct environments (an isolated test environment and the field environment) or building a second temperature sensor into the device.

But a collaboration among researchers from four British Universities (Cambridge, Manchester, Sheffield, and Bolton) and from Korea’s Kyung Hee University—has yielded a new twist on an old solution to the problem of temperature effects on precise measurement.

Temperature changes have dogged metrologists at least since John Harrison labored through much of the 18^th Century to construct the first accurate clocks. Balance springs sag as temperatures rise. Iron pendulum rods stretch.  Time seems to slow down as the weather warms up.  The solution is to combine materials to counteract these effects. Harrison devised a “compensation curb” (similar to the bimetallic spring at the heart of many home thermostats) to “rewind” the balance spring as temperatures rose. He also invented the gridiron pendulum, in which brass or zinc compression beams lift the pendulum weight to counteract the iron rod’s expansion. (See Dava Sobel’s Longitude or this Royal Society inventory of Harrison’s innovations.)

Though others had suggested a Harrisonian two-material solution to FBARs’ problems, this is the first time it’s been made to work over a broad range of real-world temperatures. To do this, Cambridge’s Luis Garcia-Gacendo and the team built a two-material device that resonates in two characteristic modes—arranged so that one of the composite's resonant modes increases in frequency as temperature rises while the other decreases in frequency. By measuring the shifts in both modes simultaneously, scientists can calculate both the temperature at the device and the mass of adsorbed particles.

The prototype FBAR device “for parallel sensing of temperature and mass loading” consists of a 2-micrometer-thick piezoelectric film of crystalline zinc oxide sputtered onto a 2 μm layer of silicon dioxide, which sits atop a silicon wafer sandwiched between chromium-gold electrodes. The prototype measurement tool showed native resonant modes at 754 megahertz and 1.44 gigahertz. Both silicon dioxide and zinc oxide expand with temperature (though at different rates), so both layers get thicker as the temperature rises. This would normally mean that the resonant frequencies of both modes would fall with temperature. In this case, though, rising temperatures cause the acoustic wave velocity to increase in the silicon dioxide layer and decrease in the zinc oxide layer. So a one-Kelvin change in temperature causes a 79.5 parts per million frequency increase in the SiO₂ and a 7 parts per million frequency drop in ZnO.

Thus, measuring frequency changes in both modes simultaneously indicates both temperature and mass.

The team successfully measured loads of human fibrinogen and bovine serum albumin in concentrations of about 1 to 1000 micrograms per milliliter (The total mass measured is unstated; the calibration tests measured masses on the nanogram level). Note, though, that the protein solutions were deposited on the membrane and then dried. The device as built relies on thickness longitudinal mode data, and cannot be applied to direct sensing of masses in liquids. The authors note, however, that repositioning the electrodes could enable thickness shear mode sensing in a successor device, opening up, for example, opportunities for direct sensing in biological or biotechnology applications. 

Images: Top: L. Garcia-Gacendo, Cambridge University. Bottom: Public Domain.

Partially overshadowed by the dislocations of Hurricane Sandy was Oak Ridge National Laboratory’s unveiling of its Titan supercomputer, a 20-petaflop Cray XK7 that will crunch massive numbers to run simulations  in materials science, combustion, and, appropriately, climate change. (In the shadow of the storm, it’s interesting to note how many of Titan’s non-weather applications also have environmental implications.)

The system contains 18,688 nodes, each containing a 16-core AMD Opteron 6274 CPU and an NVIDIA Tesla K20 graphics processing unit. The design is 10 times as powerful as ORNL’s previous supercomputer, the Jaguar, but it fits into the same space and uses only a little more power.

“Combining GPUs and CPUS in a single system requires less power than CPUs alone, and is a responsible move towards lowering our carbon footprint,” said ORNL associate director Jeff Nichols in the debut announcement. Titan’s 299,008 CPUs will guide the complex simulations, while the even faster multi-core GPUs will handle the details.

It will take a while before Titan finishes acceptance testing. When it goes online, its biggest client will be the Department of Energy’s INCITE (Innovative and Novel Computational Impact on Theory and Experiment) program.

The biggest-iron to-do list includes:

• Calculating nanoscale magnetic properties and temperature sensitivities of steels, nickel-iron alloys, and advanced permanent magnets using Wang-Landau locally self-consistent multiple scattering (WL-LSMS) methods.
• Modeling combustion in the turbulent environment of an internal combustion engine—potentially important to improving engine designs that will both conserve fossil fuel resources and reduce greenhouse gas production.
• Modeling the behavior of neutrons in a nuclear power reactor—part of a study intended to help extend the working lives of aging reactors that still provide about 20% of America’s power. (ORNL says Titan will be able to simulate one fuel-rod service cycle in 13 hours, less than a quarter of the time Jaguar needed.)
• Simulating the long-term evolution of the world’s climate, helping to anticipate future air quality and the behavior of suspended particles. The simulation will reduce the world to an array of 14x14 km cells, “imagining” five years of real time per day of computing time. (Jaguar could simulate just three months in a day of calculation.)

Scientists have harvested energy from a guinea pig's inner ear and used it to power a small wireless transmitter. With further design work, researchers could harvest this biological battery to power implanted devices near the human ear, such as molecular sensors and drug delivery vehicles for hearing loss and other disorders, according to a study to be published today in Nature Biotechnology.

It has been known for decades that the inner ear contains this biological battery, but until now, no one has harvested it. The authors of the paper, led by Anantha Chandrakasan at Massachusetts Institute of Technology and Konstantina Stankovic at Massachusetts Eye and Ear Infirmary, succeeded without damaging the guinea pigs' hearing. 

The inner ear's biological battery is located in a spiral-shaped auditory region called the cochlea. The electric potential in this region arises from the electrical difference between two different chambers in the cochlea, which contain charged particles such as potassium and chloride ions. A nearby specialized structure known as the stria vascularis transports the ions through its unique arrangement of electrogenic ion pumps, generating an electrochemical potential known as the endocochlear potential. 

At 70-100 mV, the electrochemical potential of the inner ear is the highest in the mammalian body. But it's still a very small amount of energy, and only a fraction of it can be extracted without disrupting hearing. To address this challenge, the researchers chose to power a specially designed chip equipped with an ultralow-power radio transmitter. 

In the experiments, the researchers implanted electrodes in the cochlea of anesthetized guinea pigs. The electrodes were connected to the chip, which was located outside the animals' ears. (It is small enough to fit in a human ear.) The chip included power-conversion circuitry that gradually builds up charge in a capacitor. To kick-start the control circuit, the researchers applied a one-time burst of radio waves. The device wirelessly transmitted measurements of the endocochlear potential to an external receiver. About 1 nW of power was extracted for up to 5 hours—long enough to enable the 2.4 GHz radio to transmit measurements every 40-360 seconds.

Harvesting energy from the human ear to power small electronic devices could be a huge breakthrough for people grappling with hearing loss and other disorders. Implantable electronics usually require large energy reservoirs to operate reliably over long periods of time. But human anatomy limits the size of implantable batteries, and often requires surgical re-implantation or cumbersome external wireless power sources. Harvesting enough energy from the body's own energy sources is a way to extend implant life, and maybe even allow it to operate autonomously, the authors report.

Images: Patrick P. Mercier

A lung-on-a-chip looks nothing like a human lung: It's a clear, flexible piece of silicone rubber that's smaller than your thumb, with human lung cells growing inside the microscopic channels carved into it. But researchers have shown that this gizmo can not only mimic the essential functions of a healthy human lung, it can also be used to reproduce the conditions inside a diseased lung. This proof-of-concept research shows that organ-on-a-chip devices can aid medical research and drug development, and may reduce the need for animal testing in the future. 

The researchers hail from Harvard's Wyss Institute for Biologically Inspired Engineering, which is at the forefront of organ-on-a-chip research. We've covered prior triumphs from the Wyss researchers like their gut-on-a-chip, which mimicked human intestines and came complete with peristaltic motions, and their plans to link together ten different organ-chips to create a "human-on-a-chip." They describe their latest advance in the journal Science Translational Medicine. 

The lung-on-a-chip is fabricated using techniques learned from computer microchip manufacturing. Its channels have a porous matrix in the middle that host lung cells on one side, where air flows over them, and capillary cells on the other side, where a blood-like fluid flows over them. Vacuum pumps on both sides of the chip cause it to expand and contract, mimicking the way the human lung's air sacs expand and contract with every breath. 

In the latest research, the scientists reproduced the symptoms of pulmonary edema, a potentially deadly condition characterized by fluid and blood clots in the lungs. The cancer chemotherapy drug interleukin-2 (IL-2) is known to cause pulmonary edema in some patients, so the researchers introduced IL-2 into the lung-on-a-chip and watched to see what happened. Just as in a real lung, on the chip the drug caused fluid and proteins to cross over the matrix and leak into the air flow channel.

The researchers also tested a new class of drug that's being developed by GlaxoSmithKline to treat pulmonary edema symptoms. The drug was effective on the chip, and in a separate study the pharmaceutical scientists validated the results in animal experiments. These results suggest that organ-on-chip technology could soon reduce the need for animal testing, which is expensive, slow, and controversial. 

Donal Ingber, founding director of the Wyss Institute and a senior author of this study, spoke in a press release about the utility of this cutting-edge technology:

"In just a little more than two years, we've gone from unveiling the initial design of the lung-on-a-chip to demonstrating its potential to model a complex human disease, which we believe provides a glimpse of what drug discovery and development might look like in the future."

Images: Wyss Institute

President Barack Obama’s victory on Tuesday in the U.S. Presidential election was not only predictable, it was predicted. The guy who got it totally right this time around was Nate Silver of the New York Time’s FiveThirtyEight blog (538 is the total number of votes in the Electoral College).

The guy who got it 100 percent right in 2004, and missed the final count in 2008 by a single vote, got one state wrong this time around—Florida. Sam Wang, of the Princeton Election Consortium, who we profiled back in September, knew he was on shaky ground in that state, and that state alone. Here’s what he wrote Tuesday afternoon:

Florida is a hard case.  Several new polls came out this morning, making the median basically zero. As a tie-breaker I resorted to mean-based statistics. I will be unsurprised for it to go either way. Nate Silver and Drew Linzer went the other way. We are all tossing coins. I am prepared to lose the coin toss.

The final election count is looking to be Obama 332, Romney 206. Wang’s final prediction total was Obama 303, Romney 235. The difference is Florida’s 29 electoral votes (which is still undecided as of this writing).

David Rothschild, an economist now working at Microsoft Research, has been studying—and prognosticating—the election all year. He was on IEEE Spectrum’s weekly podcast, Techwise Conversations, twice: in March and in October. Both times, he predicted an Obama win. In fact, the March prediction followed a blog post back in February that called the election with the same near-perfect accuracy. He blogged today,

“Last February, the Signal predicted that President Barack Obama would win reelection with 303 electoral votes to his opponent's 235--a prediction we made before the Republican party had chosen the identity of that challenger. 

Needless to say, Rothschild’s predictions varied between February and November—indeed, at some points, Romney was ahead. I asked him today what, then, was the point of advanced prediction. After all, if he’s going to brag about being nearly perfectly accurate back in February, he has to acknowledge that he was wildly wrong for much of the summer. Here’s what he said.

Forecasts serve two purposes which I believe we delivered on this cycle. First, they provide efficiency in a multi-billion dollar industry. For forecasts to be useful, they need to early, accurate, and consistent. Second, forecasts provide insight in how and why things happen. Granular forecasts, like mine, will help answer questions about the value of debates, big TV buys, etc., I look forward to pouring over the data in the coming weeks and months and hopefully provide answers to some major political science questions regarding campaigns and elections.

That makes sense. It’s not that the summer forecasts were wrong—Romney really was ahead, and the odds are, if the election had been held then, Romney would be president and the forecasts would have been right. Something else Rothschild told me today also makes a lot of sense:

The election was another vindication of scientific and statistical forecasting versus punditry. Polling, prediction markets, and the statistical models that surrounded them were extremely accurate on the final outcome.

If you want to know what’s going to happen, increasingly, don’t turn to MSNBC or Fox News. Go to sites like Rothschild’s The Signal, the Princeton Election Consortium, and FiveThirtyEight. Some things are predictable—if you go to the people who rely on data and not their gut.

Editor's note: this post was updated to reflect the still undecided outcome of the Florida vote as of this writing.

Image: Princeton Election Consortium

Does your table-top laser plasma accelerator (LPA) just sit there, taking up kitchen-counter space next to the unused toaster oven (too big for toast, too small for “oven”)?

Perhaps not. But the utility of LPAs—compact particle accelerators called “table-top” because they are only a few meters long, rather than a few kilometers—has been limited because their output is hard to calibrate. LPAs produce short, intense pulses of synchronized high-energy electrons, or jolts of terahertz, x-ray, or gamma radiation. Their output is around one gigaelectron volt (GeV) or less—about one ten-thousandth of the 7 TeV generated by a Large Hadron Collider main beam crossing, but more than enough to illuminate research into photosynthesis, catalytic chemistry, structural molecular biology, low-temperature superconductors, spintronics, and graphene nanostructures, according to the Lawrence Berkeley National Laboratory (LBNL), which is studying the devices.

LBNL’s experimental LPA (left, with a simulation of its wakefield) attains these energies by blasting a hydrogen plasma with an intensely bright laser, a 24-femtosecond  pulse from an titanium aluminum-oxide laser delivering more than an exawatt—10¹⁸ watts—per square centimeter. The sudden energy boost drives the plasma’s electrons and nuclei apart, creating enormously steep potential gradients—on the order of hundreds of billions of volts per meter, some five orders of magnitude higher than the paltry few million volts per meter attained by conventional accelerators. (To quote the Genie in Aladdin, “Phenomenal cosmic powers. Itty-bitty living space.”)

Electrons then “surf” (LBNL’s term) down this potential cliff, accelerating to near-light-speed in an instant to follow the laser pulse through the plasma, generating a beam of light and electrons. The result is a concentrated clump—physicists refer to it as a “bunch”—of highly synchronized relativistic electrons predicted to be about 0.1 micrometer in radius. Let’s see: a blob roughly 10^-7 meters in radius moving about 300 million meters per second…. Ah, this little bundle of high-octane joy comes and goes in roughly one-third of a back-of-the-envelope femtosecond.

The bunch’s high energy, small dimensions, and short duration make it difficult to measure any of its characteristics directly without disrupting its structure. It is especially difficult to measure the beam’s emittance, its tendency to spread in phase space—the key indicator of how much energy it can deliver to how tightly defined an area. A low-emittance beam will be tightly focused and the particles will have nearly uniform momentum.

The traditional approach is measuring emittance has been the “pepper-pot”—a sieve of small holes that splits the bunch into beamlets whose impacts can be measured individually via a CCD camera.  Unfortunately, running the electron beam through the pepper-pot destroys the beam—and possibly the pepper pot apparatus itself.

Now, however, an LBNL research team has hit upon a method for measuring electron-beam emittance without disturbing the electrons themselves.

The key is the background betatron x-ray radiation—radiation that all electrons produce as they accelerate through a plasma. These energetic photons join the electrons and the stimulating laser light in the energy stream pouring out of the LPA. Assigned to evaluate this x-ray background, one LBNL team member, UCLA post-doc Guillaume Plateau, realized that it would be possible to separate these streams and calculate the beam’s emittance from the betatron x-ray pulse.

Plateau, his LBNL colleagues, and collaborators from Germany’s ExtreMe Matter Institute and Jena Helmholtz Institute worked out a method (see diagram). First they used an electron spectrometer to curve the electrons’ path and divert them out of the beam.  Next, layers of plastics and beryllium filtered out the laser light, and a two-foot-thick concrete wall filtered out residual radiation, so that the x-ray beam could enter a high-performance CCD camera without destroying it. The camera collected the incoming x-ray photons and recorded their energy and positions.

By fitting the data back to theoretical curves, the LBNL team could confirm the theoretical calculations: the LPA delivers electron bunches 0.1 μm in radius. Combining the x-ray data with electron spectra, they calculated that the electron beam’s emittance is also in accord with the predicted value.

Good, now that I can measure bright and tight my LPA beam will be, I can get it off the kitchen counter and use it to tease the cat. His name, of course, is Schrödinger.

Images: (Top): Photos Roy Kaltschmidt;  simulation Cameron Geddes; Lawrence Berkeley National Laboratory. (Bottom) Lawrence Berkeley National Laboratory

Hurricane Sandy’s power outages have certainly provided perspective on the progress of consumer electronics.  Though lithium-ion batteries are more capacious than ever, the gadgets they power are more voracious, too. It seems we’re hardly better off in a crisis.
 
In the ideal case, you’d be able to lobotomize your device to dumb or dumber: a plain cellphone with just enough on the ball to handle email. The standard in frugality is set by the humble pager, which needs just 90 and 70 mW to send and receive email, respectively.

Compare that to a smart phone, in which the ravenous display alone sucks around 400 mW. The non-display parts are none too frugal, either. In a 2010 analysis Aaron Carroll and Gernot Heiser of  the University of New South Wales, in Australia, found that those parts of a Samsung 2.5-G phone, the Openmoko Neo Freerunner, needed 610 mW to send an email message over the GPRS system—the telephonic one that you must resort to when you haven’t got WiFi. That figure drops to 302.2 mW when sending a text message.
 
How can phones and laptops be designed with emergency conservation in mind? If you’re into full-survival mode, you might want to prearrange for your phone to turn off its display and be dead to all but the most basic telephonic signals. You might send a single text message to a pre-arranged set of phone numbers saying:  “I am alive and can receive text messages, but I will turn on the display to read them only every few hours. If you absolutely, positively must reach me immediately, send the following text to my number, and it will sound an alarm.” 
 
This idea, refined considerably, is the gist of a 2011 proposal by Peter Cole, Suwannit Chareen and Hong Xie of the school of information technology at Murdoch University, in Perth Australia. (Hmm. Why are the Australians so prominent in disaster planning, seeing as they live on the most geologically stable part of the planet?)  To allow a phone to save power by going idle, thus deactivating circuits that handle signals from many different systems—4G, 3G, Wi-Fi, and so forth, the engineers suggest what they call a Wireless Interface Notification and Activation system, which would send emergency signals to phones that activated only the relevant wireless interface, which could then a message.
 
Of course, such a system would have the added advantage of making a charge last longer even when there’s no particular emergency. That means it could attract customers in times of plenty, while protecting them in times of want.
 
Meanwhile, those living in Sandy’s wake can do a little lobotomization by hand. First off, dim your display, the power hog par excellence. Next, turn off Wifi (probably useless anyway). Then revert to 3G- from your 4-G network, and so on.  Just strip away the smarts, much as Dave the astronaut did when he disabled Hal, the insane computer, in “2001: A Space Odyssey”—
 
“Dave, my mind is going,” pleaded Hal. “I can feel it. I can feel it. My mind is going. There is no question about it. I can feel it. I can feel it. I can feel it. I'm a... fraid.”
 

Back in August, a Wave Glider robot named Alex, from Liquid Robotics, headed out into the Caribbean on a mission to measure ocean temperatures and improve hurricane forecasting.

This week, Alex’s sibling robot, Mercury, battled directly through Hurricane Sandy 160 km due east of Toms River, NJ, and the now-decimated Jersey Shore. It met the storm at the point labeled 110 in the map below and traveled with the hurricane to the point labeled 100.

The wave-powered robot transmitted weather data in real time, recording a plunge in barometric pressure of over 54.3 millibars to a low of 946 millibars as Sandy approached the coast. (Typically, atmospheric pressure at sealevel is 1013 millibars). It clocked winds at up to 70 knots, or 130 km/hour.

Photo: top: a Wave Glider robot in the Pacific earlier this year. Below: Wave Glider Mercury's path during Hurricane Sandy. Credit: Liquid Robotics

Whether you vote Tuesday on a touch-screen voting machine or use a paper ballot, a host of computer systems are making it possible to collect and count your vote. These systems maintain registration databases, manage the information that goes on ballots and enable them to be printed, scan paper ballots, capture votes electronically, and collect and count scanned and electronic votes. And, for the most part, these different pieces of technology that together make up the U.S. voting process are made by a wide variety of vendors and handle data in diverse ways.

“In any one state, it could be a hodgepodge,” says John Wack, manager of the National Institute of Standards and Technology (NIST) common data format project and Vice Chair of IEEE Standards Project 1622 (more on that project in a moment).

“Because there is no common data format,” Wack says, “a state may have databases exporting in one format, being input by systems in another format, and exporting again in yet another format. A lot of proprietary software is being written in individual states to get these systems to talk to each other.”

IEEE Standards Project 1622 is working on electronic data interchange for voting systems. The plan is to create a common format, based on the Election Markup Language (EML) already recommended for use in Europe. This is a subset of the popular XML (eXtensible Markup Language) that specifies particular fields and data structures for use in voting.

The IEEE effort first started back in 2002, stalled, and then got going again in February of 2011. In January this year, the group published a standard for electronic distribution of blank ballots, and now is readying a draft of a standard for election results reporting.

“Election results reporting is very complicated,” Wack explained. “States look at how many ballots are cast, how many were overvotes [voting twice in a single contest], how many were undervotes [failing to vote on a specific contest], how many were cast absentee—they don’t just look at the winners. We are giving them a format that should ease life for them.”

Why is getting this standard adopted important? Wack says not having a standard format means it’s tough for states to switch voting systems vendors, and it’s tough for smaller companies to break into the market. The lack of a standard also slows the adaption of new innovations, such as online blank ballot distribution systems now being tested in several states, and use of iPads, being tested in Oregon.

“People developing these kinds of new technologies want a format for the ballot data; they’d rather not have to invent one,” Wack says.

While the standard is not complete, it will be introduced this year in one region of the country. “The D.C. Board of Elections was contacted by technologists at the Washington Post, who were looking for a way of simplifying the collection of election results from the various jurisdictions on which it reports,” said Paul Stenbjorn, the board's chief technology officer. Working with colleagues in Virginia, Maryland, and West Virginia, Stenbjorn settled on the same subset of the European EML being developed into the IEEE 1622 standard, and will use it to deliver election results in real time where possible on Tuesday evening.

Diagram: IEEE Project 1622 lays out the scope of its effort to create a CDF—Common Data Format—for election systems.

This morning, the U.S. Department of Energy’s Office of Electricity Delivery & Energy Reliability issued a “situation report” summarizing Hurricane Sandy-related outages, plant shutdowns, and the like. Here are a few of the mostly grim highlights:

·      As of 9:00 am this morning, more than 8 million customers in 17 states plus the District of Columbia were without power. New Jersey residents have been by far the hardest hit, with nearly 2.5 million customers—62 percent of the state—having no electricity.

·      Three of the region's nuclear units—PSEG Nuclear’s Salem Unit 1 in southern New Jersey, Entergy Nuclear’s Indian Point Unit 3 in New York, and Constellation’s Nine Mile Unit 1 near Oswego, N.Y.—were shut down.

·     Two other nuclear plants—Exelon’s Limerick Unit 1 near Philadelphia and Dominion Resources’ Millstone nuclear Unit 3 near New London, Conn.—had their power reduced. The cause in each case was different. For example, Salem I’s shutdown was triggered by four of the station’s six circulating water pumps, which rely on Delaware Bay/River water, being unavailable due to Sandy.

·      Two oil refineries lost power: the Phillips 66 Bayway refinery in Linden, N.J.—the second-largest refinery on the East Coast—and the Hess refinery in Port Reading, N.J. Both had been shut down ahead of the storm, and as of Tuesday remained offline and without power. Four others, including the largest, Philadelphia Energy Solutions’ 330 000 barrel-a-day refinery, were operating at reduced levels.

·      Utility companies across the affected area have collectively mobilized tens of thousands of out-of-state workers and contractors. 

PHOTO: Betty Flowers. Taken on 29 October near the Wachapreague Marina in Virginia.