Aaron
Celestian

I didn't fall in love with minerals because they're beautiful. I fell in love with them because they remember. And because what they remember turns out to matter — sometimes enormously, for people who have never thought about a mineral in their lives.

Two minerals. Two emergencies. One underlying principle. A microporous titanium silicate framework is now the cleanup material for radioactive cesium contamination at Fukushima Daiichi. A zirconium silicate, geometrically related but structurally distinct, is the active ingredient in Lokelma — an FDA-approved drug that pulls lethal excess potassium from the bloodstream of patients in cardiac crisis. We characterized the ion-selectivity mechanism that makes microporous frameworks capable of this kind of precision — and described it before anyone knew what it was going to be used for. That is what basic science does when you give it enough time.

"Minerals are the planet's oldest archivists. Most of us walk past them without knowing what they're holding."

That's the logic this research program runs on. Every crystal that has ever grown — in a volcanic vent three kilometers beneath the ocean floor, in the hypersaline pools of a California desert, inside the body of a person who never knew it was forming — carries a chemical record of the conditions that made it. Temperature. Pressure. What was dissolved in the water. Whether anything alive was nearby. Minerals are the planet's oldest archivists, and most of us walk past them without knowing what they're holding. My job is to read those records. And then — this is the part that keeps me here — to ask what we can do with what they say.

I'm the Curator of Mineral Sciences at the Natural History Museum of Los Angeles County — steward of one of the finest mineral collections in the world: more than 150,000 specimens, some of them billions of years old, gathered from every geologically significant corner of the planet. Most people think of a museum collection as a display case. It isn't. It's a research instrument — a library where every book was written by the Earth itself, and where the newest chapters are still being written in geothermal brines, in the bones of marine organisms, in the walls of craters on Mars.

That comparative depth — 150,000 specimens, six continents, 4.5 billion years of Earth history — changes what questions you can ask. When I pull a halite crystal from a California salt lake and find something unexpected in its fluid inclusions, I can set it against material from a dozen other hypersaline systems in the collection. That's a question a university laboratory with no collection simply cannot ask.

"A natural history museum is not where science goes to be preserved. It's where science goes to be done."

There's something in the weight of a crystal, in the way light moves through it, that gets me every time, even now. I think that's the thing that actually drives the research: not the applications, not the papers, but the persistent suspicion that whatever the mineral is doing at the atomic level is stranger and more interesting than anything we've thought to look for yet. So far, that suspicion has been right every time.

The questions don't start in the laboratory. They start in the field, where the ground tells you something the textbook didn't anticipate — and where the thing you weren't looking for turns out to be the most important thing you found.

Salt lakes where the water is more chemistry than water. Desert playas where a billion years of continental runoff has been concentrated into a crust you can crack with your boot. Hydrothermal systems where the temperature gradient from vent to seafloor compresses the entire history of mineral formation into a few vertical meters. I work at those margins deliberately — not for the scenery, but because extreme environments are where Earth's processes run fast enough, and concentrate their signatures strongly enough, that you can actually see what's happening. In a temperate river system, the geochemistry is subtle. At the edge of a hypersaline lake in Western Australia, it is not. The record is right there, and it is legible, if you know what you're reading.

What comes back from those places — specimens, fluid samples, drill core, field measurements — feeds directly into the museum's research infrastructure: X-ray diffraction, Raman spectroscopy, high-resolution electron imaging, synchrotron access at national facilities. The field generates the questions. The laboratory reads the answers. The collection holds the evidence for the next person who needs to ask something we haven't thought of yet.

Every research program I run follows the same logic: find the record, read the record, ask what the record means for something alive.

Sometimes that means examining halite crystals from Searles Lake looking for the chemical signatures of microbial life preserved inside fluid inclusions — tiny pockets of ancient water, sealed inside a crystal, sitting in the dark for tens of thousands of years. The water is still there. The chemistry is still there. If something was living in that water when the crystal formed, the evidence may still be there too. For decades, the standard assumption was that these inclusions were geochemically inert — ancient water, sealed off, irrelevant to biology. That assumption is wrong. We are learning to read what the inclusions actually hold. And what we learn here has a direct bearing on how we design the next generation of Mars rovers — because the evaporite sequences at Searles Lake are among the best geological analogs for the sedimentary environments Mars orbiters keep finding evidence of. We are practicing the method on Earth so we know what to look for there.

Sometimes it means recognizing that nature solved an engineering problem millions of years before we did — and that the solution is sitting in the crystal structure of a mineral we already had in the collection. The titanium silicate framework of sitinakite has a cage geometry that selects for potassium ions with extraordinary precision. That same geometry, it turns out, is what biological potassium ion channels use to regulate electrochemical gradients across cell membranes. Evolution and mineralogy arrived at the same solution independently, for the same reason: the geometry works. Understanding that mechanism — characterizing exactly why this arrangement of atoms selects for cesium and potassium so precisely — led directly to two applications nobody planned: the emergency cleanup of radioactive wastewater at Fukushima Daiichi, and Lokelma, an FDA-approved drug that pulls excess potassium from the bloodstream of patients in cardiac crisis. Same crystal. Same mechanism. I described the mechanism that made both possible. That is what basic science does when you let it run long enough.

"Same crystal geometry. Same mechanism. One saves a contaminated bay. One saves a life in an emergency room."

It turns out that minerals and living systems have been collaborating on structural problems far longer than we have. Stromatolites — the layered mineral structures built by microbial mats in shallow seas — are the oldest macroscopic evidence of life on Earth, some of them 3.5 billion years old. They are not fossils in the conventional sense. They are architecture: the physical record of bacteria directing mineral precipitation to build structures that served biological purposes. Minerals and microbes, co-constructing something neither could build alone.

That same logic appears in one of the most medically significant findings this laboratory has produced. When we imaged kidney stones at synchrotron resolution, we found bacterial biofilm woven into the calcium oxalate matrix itself — not sitting on the surface, not contamination introduced during handling, but structurally integrated into the crystal architecture, as if the stone grew around the bacteria deliberately. Published in PNAS in 2026, this finding means kidney stone disease may be partly microbial in origin — opening a direct path to antibiotic and microbiome-targeted therapies for a condition affecting one in eleven people worldwide. From stromatolites to oncology: the mineral didn't just record the biology. The mineral is the biology.

"The mineral didn't just record the biology. The mineral is the biology."

Continue — lithium and the Salton Sea ›

When the International Mineralogical Association approves a new mineral species, it means something very specific: a unique atomic arrangement, found in nature, that has never been formally characterized. A new combination of elements, locked into a geometry the universe invented on its own, that now has a name.

I have named seventeen of them. People sometimes assume that mineral discovery is a collector's pursuit — a trophy, a postage stamp in a catalogue. But every new mineral is a new material. And new materials are where new science begins.

Take rowleyite. I found it at a mine in the Arizona desert — structurally unlike anything in the existing database. That distinction isn't merely academic. The same cage-like framework geometry that makes rowleyite crystallographically unique may make it capable of something extraordinary: carrying a drug molecule through the bloodstream and releasing it only when it reaches a tumor. We are investigating that possibility now. A mineral no one had ever heard of, potentially deployed in an oncology ward.

Most mineral discoverers stop at the name. The discovery earns a place in the catalogue and the work moves on. I don't stop there — because the catalogue is not the point. The point is what comes next: asking what this new arrangement of atoms can do, and whether anyone, anywhere, has a problem that this geometry might solve. Basic science advancing applied science — not the other way around.

"Seventeen species. Each one a door that didn't exist before we opened it."

Science belongs to everyone. Not as a destination you earn through credentials, but as a way of seeing — one that anyone can learn, and that changes what you notice about the world for the rest of your life. That's the obligation the museum floor carries. That's what Pocketful of χtals is for.

Pocketful of χtals treats mineralogy the way it deserves: as one of the most consequential and under-told stories in science. Each piece starts from a single mineral and follows it wherever it goes — into nuclear physics, into medicine, into the history of pigments, into the search for life on other planets. The goal is never simplification. It's translation.

Unearthed: Raw Beauty, running at NHMLAC through 2027, is the same argument made physical. Some of the largest and rarest uncut mineral specimens ever displayed — not as decoration, but as evidence: this is what the Earth has been building for four billion years, and we are only beginning to understand what it means. These specimens formed in the dark, underground, with no audience. They didn't need one. But now we get to see them — and that changes things.