Science News Jan 18 (Patreon)

[This is a transcript with references]

Welcome everyone to this week’s science news. Today we’ll talk about climate engineering, quantum computers, how to tell a nuclear bomb from TNT, what an atom really looks like, random keys from cosmic rays, who is filing the most patents and on what, climate labels for food, a tractor beam that didn’t quite live up to my expectations. And of course, the telephone will ring.

A small start-up in the US is pushing forward with climate engineering by injecting chemicals into the atmosphere before anyone can stop them.

We know that injecting certain substances into the atmosphere at high altitudes reflects sunlight back into space and cools the planet. We know that this works because volcanos sometimes blast dust so high up into the air that it continues to linger at tens of kilometres height for years. This happened for example in the 1991 eruption of Mount Pinatubo in the Philippines, which pumped out twenty million tons of sulphur dioxide. It caused the global temperature average to drop by half a degree for several years.

The company Make Sunsets wants to do basically the same, just without the volcano. They want to put sulphur dioxide into a weather balloon and release the stuff at high altitude. The company’s name refers to a side effect of injecting substances into the upper atmosphere. Because they scatter the last sunlight of the day, they make sunsets more colourful.

The company did a test run already in December and they are planning to launch at least three balloons in January, from a site in Southern Baja, Mexico.  Don’t get too alarmed, these weather balloons can’t carry a lot of weight, so the amount of sulphur dioxide they can inject into the atmosphere is just about a pound per each run.

According to a study from Rutgers university from 2018, it’d take about 5 million tons a year to cool the planet by one degree Celsius. That’s about a quarter of what Mount Pinatubo spat out. But those 5 million tons must be injected accurately in the right places, preferably near the equator, and it would probably have to be done by a fleet of aircraft.

What you *should be alarmed about is there aren’t currently any laws regulating either climate or weather modification. At the moment anyone can buy a weather balloon and inject shit into the atmosphere.

If it was done at large scale, injecting sulphur dioxide into the stratosphere would almost certainly indeed reduce the average global temperature, but it wouldn’t reverse the effects of carbon dioxide on our planet because the carbon dioxide would remain in the atmosphere. And carbon dioxide does more than heating up the planet, it also leads to ocean acidification. Moreover, injecting stuff into the stratosphere might change rainfall patterns in unexpected ways, and it would have to be repeated so long as carbon dioxide levels remain high, maybe for thousands of years.

I wish Americans start-ups would stick with social engineering, at least we’re used to that.

According to a press release from the quantum computing start-up Diraq: “Australian engineers have discovered a new way of precisely controlling single electrons nestled in quantum dots to run logic gates.” This may sound like what they’ve discovered is primarily a new way of confusing people with technical terms, and that may be part of it,but that’s not all there is to it.

Building a quantum computer requires several ingredients. First you need the quantum bits, the qubits. Then you need a way to control the qubits. You need an algorithm to do your calculation. And then you need a way to read out the result.

The new research, which was just published in Nature Nanotechnology, improves the control of silicon-based qubits. Silicon-based qubits are popular in some companies because they can be mass-produced with the existing technology for standard computers. The two states of those qubits are spin-states, either of an electron or of an atomic nucleus.

These silicon qubits do not work with superconductivity, but they still have to be cooled to about 20 milli-Kelvin to keep the spin states stable. This is a “much higher” temperature than superconducting qubits which must be cooled to a temperature of a few milli-Kelvin. Though I’d argue that 20 milli-Kelvin isn’t exactly warm either.

In any case, the new research tackles the question of how to control those qubits. Researchers usually control spin-qubits with microscopic-magnets. But the tricky part is to manipulate the spins of electrons without disturbing their neighbours, kind of like trying to raise your hand in a crowded lecture hall without knocking the phone out of your neighbours’ hand.

The team had been experimenting with different magnetic fields and materials to increase the accuracy with which the qubits could be operated. And they succeeded, but not in the way they expected. By coincidence or maybe good fate, the microwave antenna emitted a more powerful electrical field than planned, and they noticed that they could actually use an electric field to manipulate the spins by exploiting a different type of spin coupling.

The good thing about this is that controlling qubits with electric fields is simpler and the equipment takes up less space, so this rather unexpected spin-off, excuse the pun, might come in handy for building bigger quantum computers.

You might think you’d know if a nuclear bomb went off, what with the mushroom cloud and the twitter outrage and all that. But explode a weapon underground, and it’s easy to mistake a nuke for a few megatons of TNT, a problem that we’re all familiar with. Luckily new research from the Pacific Northwest National Laboratory can help you out.

While it’s relatively simple to pinpoint the time and location of such explosion with seismometers, those aren’t of much help to figure out whether its origin was chemical or nuclear. Traditionally, detection of a nuclear incident, whether a bomb or a reactor explosion, comes from sniffing for radioactive particles - particularly radioactive xenon, which can travel very long distances. Those are tracked by an International Monitoring System and analysts then sift through this data to tell apartpotential nuclear bomb tests from background noise for example from the production of medical isotopes.

A better way to do it would be to combine seismic data with the emission of radioactive particles. This is more difficult than it sounds because you need to know how long it takes for the gas to seep through the cracks in the rocks to the surface. To find out what happens, the researchers did not test a nuclear bomb, but instead blew up a small part of New Mexico. They set off chemical explosions underground via boreholes, and used electrical currents to image the rock structures before and after the blasts.

These models will make it easier to estimate how long it’d take for gases to leak out and better identify underground nuclear explosions.

Hey Joe,

Ah, who hasn’t misplaced a few classified documents.

Whatever you talk about, don’t mention nuclear power. It really splits the audience.

You’re welcome.

We often draw atoms like this. But would you believe it, it’s not what they look like. Electrons don’t zip around the atomic nucleus like little planets, they rather hover around it in clouds. But would you believe it, they don’t really look like clouds either.

Just what an atom looks like is so difficult to say because it doesn’t look like anything we *can see. The electrons around an atomic nucleus are described by a wave-function, usually denoted Psi, that takes on complex values. But how does that look like?

One way to take an image of an atom that physicists have come up with is to use light to kick off the electrons, and then measure their properties. Researchers from Waseda University in Japan have now taken the most detailed image of a neon atom using attosecond light pulses. An attosecond is a billionth of a billionth of a second, and about the attention span of the average twitter user. They use these pulses to kick off the electrons, and then measure their momentum, that is, a combination of their mass and velocity.

According to their paper that was just published in PRA, the researchers used two different pulses that kick the electrons off in two different ways. The electrons can then interfere, which reveals information about the phase of the electron wave-function. The result is stunning. In those images on the right you see the measurement results and on the left the computational reconstruction. These are the wave-functions in momentum space, the colour encodes the phase of the wave-function and the intensity of the colour shows the amplitude, that is, how likely the electron is to have this value. I find this fascinating. It actually gives you a feeling for what an atom is.

Now, electron orbits in a neon atom aren’t exactly cutting edge research. But this method can be applied to more complex molecules and materials and find applications in nanotechnology, quantum chemistry, and molecular biology. But nothing will stop us from drawing atoms like this.

Embracing cosmic randomness sounds like old advice from new agers, but maybe there’s something to it. Hiroyuki Tanaka from the University of Tokyo suggests that we use the random arrival of particles from outer space to securely share random keys.

Web commerce relies on random numbers for its security, by inserting them into encryption keys to make sure transactions are uncrackable. But generating truly random numbers is hard for computers which were designed to *not do random things, and sharing random numbers without getting intersected is even harder.

But cosmic rays are a natural source of randomness. Cosmic rays are usually atomic nuclei stripped of their electrons that come from outer space and hit the upper atmosphere. Some come from our Sun, some from other stars in the Milky Way, some even from other galaxies. When they hit our atmosphere, they create a shower of secondary particles known as a “cosmic ray shower”. A lot of the particles in those cosmic ray showers are muons, that are heavier versions of electrons. They can be fairly easily detected. It’s these muons that the new cryptographic system relies on.

Muons arrive on the ground at random times, so the exact arrival time can serve as a random number. But these muons have so much energy they don’t get stopped in the detector; they go through. This means they can be measured a second time. Suppose you have a muon passing through two detectors of which you know the exact distance. Then you can, at the second detector, reconstruct the arrival time at the first one. It’s a shared random key. But anyone in between who does not know the exact distance cannot recreate the random number.

The issue is that for this method to work, it needs to be sufficiently likely to have a muon going through both detectors and you need to be able to determine the distance between the detectors extremely precisely. In the experiment whose results were just published, the two devices sharing a random key were 70 centimetres apart. I’m not a security expert, but maybe at that distance you could just show them your credit card. The author believes that the method can be extended up to 10 meters, and then relay stations can be used.

A new report about trends in US patents from IFI CLAIMS  comes with a few surprises. The number of patent applications reached an all-time high in 2022 with more than four hundred thousand. Long-time champion IBM has been knocked from top position by the South-Korean company Samsung. Note that the vertical axis of this graph starts at 4k, not at zero.

IBM had topped the chart for the number of patents registered in the US each year ever since 1993. And this isn’t the only change. While most of the US patents are granted to companies in the US, the US share is dropping, while the share of patents going to Asia is increasing. The largest increase of patents this year comes from China.

The fastest growing technology sectors are partly what you’d expect but also hold a few surprises. On Place number 10 we have Breathing Masks. Amazing what a little virus can do to the world. Keep in mind that this is a list of fastest growing sectors, not necessarily large sectors. Number 9 is Machine Learning. More evidence that the world will be taken over by artificially intelligent robots soon.

Number 8 is biological 3D cultures, that’s lab grown cells. It’s currently mostly used for medical studies and most of the patents come from academic institutions, but maybe one day it’ll come in handy for artificial meat.

On place 7 we have cigarettes and cigars, and most of the patents come from Philip Morris supposedly to benefit “harm reduction”. Place 6 is held by quantum computers. Most of the quantum computing patents presenty come from IBM, followed by Google. On place 5 we have more smoking, but this time of the electric type. Also led by Philip Morris. Some things don’t change.

Number four are computer models for biology, that’s 3D models of blood vessels and organs and things like this. The leader in this area is also IBM. On Place 3 there’s drills and everything related to drilling. Mostly for oil and gas, but also for the exploration of geothermal energy. We’re working on a video about that.

On Place 2 we have an area with the catchy name “Electric Digital Data Processing” that’s everything from 3d printing to automated sensors. Boeing is leading the pack. And one place 1 we have, drums please, autonomous vehicles, aka self-driving cars, and Toyota is presently filing the most patents on that.

IFI Claims which put out this report is part of Digital Science, a technology company that is quietly working to revolutionize science and that itself deserves having an eye one. They own, among others, Altmetrics, ReadCube, and Overleaf.

Hi Elon,

Yeah, the Biden joke was pretty lame, sorry. Some things weren’t meant to fly, pigs, Biden jokes, Richard Branson’s rockets.

I knew you’d like that one. Talk to you soon.

If you’ve ever looked at a packet of Peruvian strawberries and marvelled at the globe-spanning supply chain it took to bring them to your plate, your marvel might soon get a colour code.

Scientists at Johns Hopkins Bloomberg School of Public Health recruited 5000 Americans, to test food labels that told them how environmentally friendly the item was. Participants were randomly assigned to three groups. They were all shown a fast-food menu and asked to pick one item they’d like to order for dinner. But depending on which group they were in, the menu contained different labels: a neutral control label; a green climate impact label on environmentally friendly foods; or red label on climate unfriendly food like red meat and other stuff you’re supposed to feel bad about.

The participants who were told that their choice positively or negatively contributed to climate change, were almost 25 percent more likely to choose something associated with lower carbon emissions. Women were more likely than men to respond to the labelling.

People who made a more sustainable food choice also thought it was healthier, which I think is really the main problem. Someone needs to come with carbon-neutral fast food that’s really unhealthy, it might just save the planet.

Speaking of food and climate change, a new study by researchers from Cornell University says that as temperature rise, Broccoli will begin to look more like cauliflower. If you still had doubts that climate change is bad, this should settle it. And if the coming summer will indeed be as hot as they say, then that’ll be a great opportunity to study how European heads will begin to look more like tomatoes.

The tractor beam is a staple of science fiction, frequently seen pulling the Starship Enterprise into trouble. But in reality, while they do kind of exist, they’re more likely to be moving dust than spaceships. Researchers from QingDao University of Science and Technology in China have now set a new record in using what they call a tractor beam to move the biggest object so far, that’s a thin plate about 5 centimetres long.

A tractor beam is made of radiation, either light or sound. You can trap an object with it and move it around, which includes pulling it toward you. How do you pull something by firing at it? Doesn’t that violate a law of nature or go down at least as noncompliance with some federal regulation?

Not necessarily. Because a tractor beam can stimulate the object to emit radiation into other directions which creates momentum, and that can actually move the object toward you. It’s like if you yell at someone and they come to hit you in the face, it’s the backlash that makes it happen.

In this new research, they used a laser to heat a specially prepared plate made from graphene and silicon dioxide. They put it in a box with air at very low pressure, about 20 thousand times less than normal atmospheric pressure, and illuminated the plate with a laser. The plate conducts the heat to the side facing away from the light source. That way, more energy is transferred to the gas molecules on the back side of the plate, so the plate is pushed toward the laser.

The researchers say that the method might have the potential to one day steer vehicles or aircraft on Mars. I think it might be easier to just use the laser to turn on an engine but maybe I lack imagination.