The silicon has done its job; the chemistry has not kept pace.
At Harvard, a silicon chip once designed to listen to neurons has been quietly repurposed to speak the language of life itself—synthesizing 64 distinct DNA sequences simultaneously using water and enzymes rather than the harsh solvents that have long defined the field. Published in Nature Electronics, the work by Donhee Ham's team marks a meaningful step toward DNA manufacturing that is safer, more accessible, and less tethered to the centralized industrial facilities that currently gatekeep this essential biotechnology. The chip does not merely improve upon prior methods; it reframes what is possible, suggesting that the tools of computation and the tools of biology may share a deeper kinship than we once imagined.
- DNA synthesis has long been bottlenecked by hazardous chemical solvents and expensive centralized labs, limiting who can access this foundational biotechnology.
- The Harvard chip shatters a key benchmark, synthesizing 64 DNA sequences at once—five times more than previous enzymatic attempts—using only water-based reactions that mirror how living cells naturally build DNA.
- Precision electrical currents create and contain microscopic acidic zones at each of the chip's 64 synthesis sites, allowing parallel construction of unique sequences without chemical cross-contamination.
- The team even encoded 169 bytes of text into DNA, gesturing toward a future where the chip could underpin large-scale biological data storage.
- Progress now stalls not at the silicon but at the chemistry: intermediate molecules generated during deprotection drift between sites, and researchers must develop faster, more direct chemistry to match the chip's precision.
Silicon chips have spent decades processing information for computers, but a Harvard team led by Donhee Ham has handed them a second life—one built from biology rather than electronics. Their chip, described in Nature Electronics, can synthesize 64 different DNA sequences at once using water and enzymes instead of the hazardous organic solvents that have dominated DNA manufacturing for decades. That shift in method is as significant as the number itself, pointing toward synthesis that is safer, smaller, and potentially available far beyond today's specialized facilities.
The chip works with elegant precision. DNA grows one nucleotide at a time, and after each addition a chemical brake must be released by briefly lowering the local pH. The chip achieves this through concentric ring electrodes at each of its 64 synthesis sites: an inner ring generates protons to acidify the immediate area, while an outer ring simultaneously neutralizes any protons that drift outward, keeping the reaction tightly confined. Cycling through this process builds 64 unique sequences in parallel.
The path to this discovery wound through neuroscience. Jeffrey Abbott, a former graduate student in Ham's lab, originally designed the chip's precision current control to record electrical signals from neurons. The team recognized that the same underlying capability—controlling conditions at exact locations—could be redirected toward chemistry. The chip also demonstrated that its sequences could encode 169 bytes of text, hinting at a future role in DNA-based data storage.
Yet scaling up revealed an unexpected obstacle. When synthesis sites were packed more tightly together, the chip itself performed perfectly—but intermediate molecules generated during the deprotection step drifted into neighboring sites, blurring reactions the electrodes had kept chemically separate. The silicon has outpaced the chemistry. The clear next challenge is developing a more direct deprotection method that can match the precision the chip already delivers.
Silicon chips have spent decades as the quiet backbone of computing, processing information in machines we carry in our pockets and sit at on desks. But researchers at Harvard have now handed these chips a second life—one that has nothing to do with electrons and everything to do with the molecules that make up living things. A team led by Donhee Ham, a professor at Harvard's engineering school, has demonstrated that a silicon chip can do something that seemed unlikely just years ago: manufacture DNA.
The breakthrough, published in Nature Electronics, centers on a chip capable of synthesizing 64 different DNA sequences simultaneously. What makes this achievement significant is not just the number, but the method. Instead of relying on the harsh chemical solvents that have dominated DNA manufacturing for decades, the chip uses water and enzymes—a process that mirrors how cells naturally build DNA. The shift matters because it opens a path toward DNA synthesis that is safer, smaller, and potentially available far beyond the specialized laboratories where it happens today.
Making synthetic DNA has become essential to modern medicine and science. Researchers use it for diagnostics, for editing genomes, for understanding cancer. But the standard approach—phosphoramidite chemistry—requires hazardous organic solvents and expensive centralized facilities. Scientists have long known that enzymatic synthesis, which uses only water, could be gentler and more accessible. The problem was scale. Previous attempts at enzymatic DNA synthesis managed to produce only about a dozen sequences at once. The Harvard chip quintupled that number, establishing a new benchmark for the technology.
The mechanism is elegant in its precision. DNA is built one nucleotide at a time, like stacking blocks. After each block is added, a chemical brake prevents the strand from growing further. To add the next block, that brake must be removed by lowering the pH—making the environment more acidic. On the Harvard chip, this happens through tiny electrical currents. The chip's surface holds 64 synthesis sites, each surrounded by two concentric ring electrodes. When one site is activated, the inner ring generates protons that lower the pH locally, allowing the DNA to grow. The outer ring simultaneously removes protons spreading outward, keeping the acidic zone confined to that single spot. By cycling through this process repeatedly, the chip builds 64 unique sequences in parallel across its surface.
The path to this discovery was unexpected. Jeffrey Abbott, a former graduate student in Ham's lab, originally designed the silicon electronics to record electrical activity from neurons—to listen, in effect, to the brain's chatter. The team realized that the same precision current control they had developed for neuroscience could be repurposed. Instead of permeabilizing cell membranes, the electrodes could localize pH for chemical reactions. The insight came from recognizing that the underlying principle—precise control of conditions at specific locations—transcended the original application.
The researchers also demonstrated that the 64 sequences could encode text, storing 169 bytes of information in DNA. This hints at a future application: DNA as a data storage medium. Such a system would require manufacturing DNA at scales far beyond what exists today, but the team believes water-based enzymatic synthesis could eventually make that economically and environmentally feasible. As production volumes grow, the reduction in solvent use could significantly lower the environmental footprint of large-scale DNA manufacturing.
Yet the path forward is not without obstacles. When the team tried to pack synthesis sites closer together to increase the number of sequences produced simultaneously, the experiment failed—but not in the way they expected. The chip itself performed flawlessly, confining low pH exactly where it was supposed to. The limitation came from the chemistry itself. During deprotection, the low pH generates intermediate molecules that perform the actual removal of the blocking groups. These intermediates drift into neighboring sites, blurring the separation between reactions even though the pH remains tightly controlled. The silicon has done its job; the chemistry has not kept pace. The next step is clear: develop a more direct deprotection chemistry that can work at the speed and precision the chip now enables.
Notable Quotes
A defining feature of the chip was precision current injection, which we used to permeabilize neuronal membranes for intracellular access. At a certain point, we wondered whether that same current control could be redirected from cells to molecules.— Donhee Ham, Harvard professor of engineering
The chip did what we asked it to do: it localized low pH at selected sites. The limitation came from the deprotection chemistry, not from the silicon.— Han Sae Jung, co-first author of the study
The Hearth Conversation Another angle on the story
Why does it matter that this uses water instead of organic solvents?
Because solvents are toxic and expensive. They require specialized facilities with ventilation, safety equipment, waste disposal. Water-based synthesis could eventually happen in smaller labs, even clinics. It's safer for the people doing the work and for the environment.
But enzymatic synthesis has been around for a while, hasn't it? What's new here?
The scale. Previous attempts managed about a dozen sequences at once. This chip does 64. That's a real jump—it suggests the approach might actually be practical someday, not just a laboratory curiosity.
How did they get from studying neurons to making DNA?
The original chip was designed to record electrical activity inside neurons. They realized the same precision current control could lower pH in specific spots on a surface. Instead of opening cell membranes, the electrodes could trigger chemical reactions. It's a beautiful example of technology finding a new purpose.
What's stopping them from scaling up further?
Chemistry, not engineering. The chip can confine the acidic conditions perfectly. But the intermediate molecules created during the deprotection step drift into neighboring sites. The silicon is ready; the chemistry needs to catch up.
What would DNA data storage actually mean?
Storing information in DNA sequences instead of on hard drives. It's incredibly dense—theoretically, you could store the entire Library of Congress in a few grams. But it requires manufacturing DNA at enormous scale, which is why enzymatic synthesis in water becomes attractive. It's cleaner and could be cheaper at volume.