· 2D Materials
Every modern chip runs on pairs of transistors, one carrying electrons and one carrying holes. Atom-thin materials have only ever been good at one of them.
A Tokyo team just grew a wafer-sized atom-thin film that finally makes the other half work.
Researchers led by Vincent Tung at the University of Tokyo grew wafer-sized, atom-thin films of boron carbon nitride (BCN) by replacing some nitrogen atoms in boron nitride with carbon, which leaves holes that carry current in p-type transistors. A two-ingredient vapor growth method kept carbon from clumping, and the films were grown even on patterned silicon chips. Of 224 transistors tested under everyday conditions, "on" currents were 100 million times larger than "off" currents, most devices surpassed targets the chip industry aims to reach by 2031, and the material stayed stable in air. Remaining hurdles include unoptimized metal contacts, shrinking the devices, and lowering the 1,000°C growth temperature that could damage underlying components. Published in Nature.
Each carbon atom that replaces nitrogen in the honeycomb leaves behind a hole, the charge carrier of a p-type transistor.
The Missing Half of the Atom-Thin Chip
If computer chips could be built from semiconductors just one atom thick, they could pack far more transistors into a smaller space while using less power.
That promise has been held back by a stubborn gap. Atom-thin materials have proven good at one kind of transistor and poor at the other, and a chip needs both.
In a new study in Nature, researchers led by Vincent Tung at the University of Tokyo show that atom-thin sheets of boron carbon nitride (BCN) could fill that gap.
The material supports "p-type" transistors, the weaker half of the pair, and it does so with strong currents, clean switching, and stability in air.
It is still a laboratory result with real hurdles ahead, but it addresses what the source describes as a bottleneck for the whole technology.
Why Chips Need Two Kinds of Transistors
Today's chips rely on two types of transistors working in pairs.
| Type | What carries the current |
|---|---|
| n-type | Negatively charged electrons |
| p-type | Holes: vacancies left where electrons are missing, which move like positive charges |
The two types are paired because together they let a chip switch cleanly while wasting little power. Remove one half and the design falls apart.
That is why the weakness of p-type atom-thin devices matters so much. It is not a niche problem but half of the circuit.
The p-Type Bottleneck in Atom-Thin Materials
Researchers have found several atom-thin materials that carry electrons very well. Finding counterparts that carry holes just as well has proven far harder.
That imbalance is the bottleneck. Good n-type materials alone are not enough to build the paired circuits modern chips depend on.
Many candidate materials so far have involved trade-offs. Some conduct poorly, some break down in air or heat, and some are difficult to make at large scale.
Each failure mode is serious on its own. A material that works only in a vacuum, or only as a tiny flake, cannot become a manufacturing technology.
So the target was demanding: a hole-carrying atom-thin material that conducts well, survives ordinary conditions, and can be grown at wafer scale.
Swapping Nitrogen for Carbon in Boron Nitride
Tung's team started from boron nitride, an atom-thin insulator with a honeycomb structure.
They replaced some of its nitrogen atoms with carbon, creating boron carbon nitride, or BCN.
The logic comes down to electrons. Carbon has one fewer outer electron than nitrogen, so each swap leaves a hole behind.
The key condition is that the honeycomb stays intact. If the swap distorts the structure, the holes cannot move freely, and the material loses its advantage.
Growing Wafer-Sized Films Without Clumping
A second problem is that carbon atoms tend to clump together instead of spreading out, which would spoil the structure.
To prevent that, the researchers used two vapor-based ingredients. One steadily supplied boron and nitrogen, and the other released its carbon slowly.
This approach is called epitaxy, a method of growing a thin crystal layer in an ordered way. It let the team grow wafer-sized films of BCN, including on patterned silicon chips.
Wafer-sized is the phrase that separates a manufacturing path from a curiosity. Many exciting exist only as tiny flakes that cannot be scaled.
Growing directly onto patterned silicon chips also hints at compatibility with existing chip technology, though the source does not claim a production process.
What 224 Transistors Showed
After examining the films' structure with atomic-resolution microscopy, the team built and tested 224 transistors under everyday conditions.
The microscopy showed that carbon atoms were evenly distributed throughout the material, primarily replacing nitrogen. The honeycomb structure stayed almost undisturbed, so the holes could flow freely.
| Result | Reported |
|---|---|
| Transistors built and tested | 224, under everyday conditions |
| On/off current ratio | "On" currents 100 million times larger than "off" currents |
| Versus industry targets | Most devices surpassed targets the chip industry aims to reach by 2031 |
| Stability | Stayed stable in air |
A 100-million-fold ratio between on and off currents means clean switching, which is what makes a transistor useful as a digital switch.
The source does not say which specific industry targets were surpassed, nor does it provide hole mobility, device sizes, or how the devices compared head-to-head with other p-type candidates. This article does not add numbers beyond what is reported.
The Hurdles: Contacts, Size, and a 1,000°C Growth Temperature
The researchers hope the result will help pave the way for chips that pair atom-thin n-type and p-type transistors, potentially stacked in layers.
Several steps remain. The metal contacts that connect the BCN to surrounding wiring were not yet optimized, so improving them could boost performance further.
The transistors also need to shrink, which is part of why atom-thin materials are attractive in the first place.
If those challenges can be overcome, the source says the technology could become scalable enough for commercial rollout. That is a conditional, not a forecast, and no timeline is given.
Manohar Insight
What I like here is that the study attacks the right problem. Atom-thin n-type devices were never the issue. The missing half was the p-type side, and that is what BCN addresses.
The chemistry is also pleasingly direct. Carbon has one fewer electron than nitrogen, so every swap makes a hole. The material design follows from a simple fact about the periodic table.
I would keep two numbers in context. The 100-million on/off ratio is impressive, but the 224 devices were lab-built, and "most" beat the 2031 targets, not all. This is encouraging evidence, not a finished process.
The 1,000°C growth temperature is the detail I would watch most closely. A material that must be grown hot enough to damage the chip beneath it needs a workaround, such as growing elsewhere and transferring, before it can enter a real fab.
I also note what the source leaves out: no comparison table with rival p-type materials and no cost data. Those will decide whether BCN becomes the answer or one candidate among several.
The real story is not a new transistor record. It is that the weaker half of the atom-thin chip may finally have a credible partner, which is the precondition for building complete circuits from these materials.
Join the Signal
If atom-thin chips need a material that must be grown at 1,000°C, do you think the better fix is cooler growth or a transfer step onto finished chips? Share your take in the comments.
Discussion