Harvard team uses fluorescent molecules to store data

A growing problem for the vast amount of digital information the world generates is how to store it efficiently and keep it intact and accessible.

A research team at Harvard University's Department of Chemistry and Chemical Biology has devised an ingenious new technique for storing data inside molecules of fluorescent dye mixtures that are deposited by an inkjet printer onto an epoxy surface. 

The new method works better and more efficiently for archival data storage than current data storage methods, such as optical disks, magnetic tape and flash drives, and is a less costly method than synthetically produced DNA which is also being considered for storage.

The researchers' work appears in the American Chemical Society Journal Central Science, Oct. 13, 2021.

The authors report that their approach "enables information storage with high density, fast read/write speeds, and multiple reads of a single set of molecules without loss of information, all at an acceptable cost."

Cost is an important factor, as the amount of digital information is increasing daily and is now hundreds of billions of gigabytes. 

Fluorescent molecule coding 

Current methods for storing digital information use a binary code like ASCII where strings of coded information are transferred using electrical or magnetic signals to media, most commonly tape, where they can be retrieved by decoding.

The new technology uses colored dots to encode ASCII information, and each tiny dot of information has a mixture of up to eight dyes encoded.

Different molecules, each with a specific fluorescent dye and specific wavelength, are deposited in an eight-bit string, using an inkjet printer, onto a stable epoxy substrate that has been chemically treated.

These strings are arrayed in a square pattern. 

A fluorescent microscope is used to read the information. Unlike other optical storage systems read by laser, there is no significant loss of data. The authors stress that the new method doesn't require the registration of individual molecules, and that the information can be read in parallel instead of sequentially.

The researchers used their new method to convert an article and a small JPEG photo of Michael Faraday into fluorescent dye patterns that were then read back with a "0.4 percent printing error."

Reading and writing molecules

Corresponding author George M. Whitesides, professor of chemistry, who conceived the idea of molecular information storage in mixtures of molecules, offered a description of the process.

"A mixture of molecules is important. When any molecule is present, that position in a binary string is occupied," he said. "Molecule A corresponds to position 1, molecule B corresponds to position 2, and then a mixture that contains one molecule or some molecules may be classified in a one (1) zero (0) type of system. It's a pretty straightforward type of system, a normal binary system very familiar to anyone who does information storage."

Whitesides emphasized that they take advantage of already developed devices. Inkjet printing is a highly optimized technology and readily accessible inexpensively, which makes this writing method attractive.

As for reading the information, he said, "This is done by exciting the molecules with lasers of various wavelengths and then seeing what colors are there. The colors can be read by using a microscope and the illumination. We tried a couple of methods and, basically, running a laser over the entire surface is a good way of doing this and fast."

A better method

In a previous paper, Whitesides said, the research team focused on measuring the different masses of molecules, but that had limitations, including slow readout. "It turned out that you get higher density and better separation of information if you use fluorescent molecules, so this is probably a better method," he said.

The fluorescent molecule method also has advantages over DNA storage:

"A key issue to consider in this problem is whether it would be possible to store information in sequences of artificially made DNA," Whitesides said. "And the answer is, yes, you can. But it's hard to make DNA. It's expensive, it's slow, [but] it's getting better and will continue to get better." 

The paper notes that DNA reading is so slow that it typically requires "multiple hours to decode a simple message," making it impractical for many applications.

Other storage methods that use tape need a climate-controlled environment, Whitesides said, which is costly. Also, the tapes need to be erased and rewritten periodically, "to make sure the information is stored in fresh and non-corrupted, as information does disappear all the time," he added.

In contrast, he said, "a molecule is a molecule, and if you choose the right molecule, they tend to last a very long time."

In general, Whitesides said, the same metric applies to any storage method.

It has to be easy to do, the rate at which you can store information has to be very high, and you have to realize that the world works with what's best and cheapest," he said. "We argue that this is another way of storing information that doesn't require as much in the way of energy as conventional methods and doesn't have the problem of code becoming obsolete." 

Taking an idea to development

Whitesides emphasized that a major motivation for the technology was to find a more efficient and less energy costly way of storing information.

"If you look at how energy is going to be used in the future, an amazingly large amount is going to go into electronics and storage of information," he said. "And there's so much large information--children's pictures, pictures of pets, family stuff, but also applications in national security and astronomy and other areas where there are large amounts of information to store. This is the principal motivation, to find ways of storing information that use as little energy as possible."

Whitesides said he is "very optimistic" about developing the technology because it works well for many different purposes.

His laboratory has started a company to develop the method, "because that's the most efficient way of doing it," Whitesides said, adding, "Which direction they'll develop depends on initial demonstrations of applications." 

He emphasized that "engineers in small companies can do things that you just can't do with graduate students, who just don't have the background or the training. Engineers in small companies can accomplish a kind of science development that's pretty hard to do in a university, particularly given the difficulties of raising money for research."

******

Amit A. Nagarkar, et al. "Storing and Reading Information in Mixtures of Fluorescent Molecules," ACS Central Science, Oct. 13, 2021. 

DOI: 10.1021/acscentsci.1c00728