Tomographic Phase Imaging with Randomized Probe Imaging
This paper demonstrates the first application of randomized probe imaging (RPI) to tomography, enabling rapid 3D phase imaging of cubic gold nanoparticles with sub-100 nm resolution from single far-field diffraction patterns per projection, thereby significantly accelerating data collection compared to traditional ptychography-based methods while maintaining high consistency in reconstructed volumes.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine trying to see the inside of a tiny, invisible world, like the gears inside a watch that is smaller than a grain of sand. Scientists have a special tool for this called X-rays, which can pass through things our eyes can't see. But X-rays are tricky; they often just pass right through without leaving a clear picture, especially if the object doesn't absorb much light. To solve this, scientists use a clever trick called "phase imaging." Think of it like this: if you shine a flashlight through a clear glass window, you can't see the glass itself, but if you look at how the light bends and shifts as it goes through, you can figure out the shape of the window. That's what phase imaging does—it maps out how the light waves wiggle and shift to reveal the hidden 3D structure of tiny objects.
Usually, to get a perfect 3D picture of something this small, scientists have to take hundreds of photos from different angles, like a CT scan at the doctor's office. The problem is that the best way to get these high-quality photos is incredibly slow. It's like trying to paint a masterpiece by moving a single brushstroke at a time, scanning back and forth over the same spot for hours. This slowness makes it impossible to catch fast-moving events, like a crystal forming or a fuel cell working in real-time. The big question in this field has been: "Can we get a fast, 3D picture of the nanoscale world without spending all day waiting for the camera to finish its scan?"
This paper tells the story of a new, speedy shortcut the researchers at Argonne National Laboratory discovered to answer that question. They tested a method called "Randomized Probe Imaging" (RPI) on a cluster of tiny, cube-shaped gold nanoparticles. To understand their trick, imagine you are trying to figure out what a mysterious object looks like by shining a flashlight on it. The old, slow way (called ptychography) is like moving the flashlight in a tiny, precise grid pattern, taking a photo at every single step to build a perfect map. It's accurate but takes forever.
The new method, RPI, is more like throwing a handful of glitter at the object in a chaotic, random pattern and taking just one snapshot of the messy reflection. The researchers used a special lens called a "randomized zone plate" to create this chaotic, glitter-like light pattern. Even though the light hitting the object was random and messy, they used a smart computer algorithm to untangle the mess and reconstruct a clear 2D picture from that single snapshot. They did this for many different angles, just like a standard CT scan, but instead of spending hours scanning each angle, they grabbed one quick snapshot per angle.
The results were impressive. When they compared the 3D volume made from their fast, "one-shot" method against the slow, "perfect" method, the two looked very similar. The fast method couldn't quite see the tiniest details (like the tiny gaps between the gold cubes), resolving details down to about 62 nanometers, but it captured all the big shapes perfectly. The best part? The speed. While the traditional method took about four hours to scan the sample, their new method could have done the same job in just 73 seconds if they used a standard step-by-step scan, or even 7.3 seconds with a super-fast "fly scan."
The authors suggest that this technique is a game-changer for watching fast processes in real-time. Because they can now get a 3D picture of nanoscale objects in seconds instead of hours, they might finally be able to film things like crystals growing or fuel cells operating as they happen. While the resolution isn't quite as sharp as the slow method, the trade-off is worth it for speed. The paper confirms that this "one-shot" approach works reliably, proving that we don't always need to move the camera slowly and carefully to get a good 3D picture; sometimes, a little bit of chaos and a fast shutter speed can do the trick just as well.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.