This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.
Reactions FIRSTFirst REACTIONS
Molecular Storage, Data Writ Small Lisa Dhar Innovation and New Ventures Office and McCormick School of Engineering, Northwestern University, 1800 Sherman Avenue, Evanston, Illinois 60201, United States
number of research groups have achieved increasingly sophisticated demonstrations of data storage, moving from early systems where one bit was stored per base to more recently proposed architectures where short-defined sequences of DNA are enzymatically combined to represent patterns of bits.7 Other approaches under development shift focus from DNA to polymers composed from limited sets of monomers or, more futuristically, envision a digital basis set of millions of uniquely synthesized molecules8 with data readout undertaken by analytical techniques such as mass spectrometry or nuclear magnetic resonance. While the above approaches hold the theoretical promise of ultrahigh storage densities and aim to provide significant improvements in metrics such as archival life, tamper resistance, and energy consumption, they face scientifically interesting but perhaps technologically challenging complications. The writing of new data as “beads on a string”5 carries a heavy synthetic load. Methods to synthesize large molecules on demand can be susceptible to errors and, for many materials, are still in their infancy. Additionally, some of the envisioned formats for molecular storage in fluids or droplets are not easily compatible with the infrastructure underlying today’s highly developed storage industry. In this issue of ACS Central Science, our authors describe storage that takes advantage of chemistry without “having to do chemistry” and in formats that can translate to current storage architectures. The authors localize mixtures of low molecular-weight molecules with distinguishable signatures onto small spots arranged in an array. Each molecule represents a bit within a byte of information. By depositing mixtures of molecules, multiple bytes can be spatially multiplexed onto a single spot. Here, the molecules are chosen from a small, welldefined set of premade oligopeptides, resolvable by mass spectrometry. Choosing from a set of eight oligopeptides enables recording of a single byte in a spot, while expanding to a set of 32 oligopeptides yields 4 bytes. Capitalizing on the group’s expertise in self-assembled monolayers for matrixassisted laser desorption/ionization mass spectrometry
Downloaded via 95.85.70.33 on May 11, 2019 at 23:37:21 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Storing digital information in mixtures of small organic molecules
W
ith digitization reaching nearly every aspect of our society, data storage needs are growing at an accelerating rate. Most telling is the evolution in our use of prefixes to describe storage needs. From giga- to tera- to peta- to the now routinely contemplated exabytes, technologies will be soon be pushing us into the zettabyte and even yottabyte regimes.1 As illustrated in Figure 1,2 the amount of data being generated outpaces our ability to store it, placing tremendous pressure on the storage industry to find improvements in flash, magnetic, and tape memories and to look to new possibilities in fields ranging from molecular structures to nanoscience3 to quantum computing.4 In this issue of ACS Central Science, Mrksich and Whitesides and colleagues describe an elegant embodiment of molecular storage that circumvents many of the synthetic and implementation challenges of the field.5
In this issue of ACS Central Science, Mrksich and Whitesides and colleagues describe an elegant embodiment of molecular storage that circumvents many of the synthetic and implementation challenges of the field. Some of the most intriguing next generation storage strategies are those that exploit molecular or chemical structures and their signatures. For example, DNA, biology’s storage medium with its modular nucleotide code, lends itself to digital storage and computation and has inspired extrapolations to theoretical raw storage densities that approach hundreds of exabytes in only a gram of material.6 Leveraging advances in DNA synthesis and sequencing, a © XXXX American Chemical Society
A
DOI: 10.1021/acscentsci.9b00439 ACS Cent. Sci. XXXX, XXX, XXX−XXX
Reactions FIRSTFirst REACTIONS
ACS Central Science
Figure 1. Schematic of the growing disparity between the worldwide supply and demand for data storage. Adapted from ref 2 with permission. Copyright 2019 Statista.
Storage densities can be optimized by increasing the complexity of the molecular mixtures and decreasing both the spot sizes and the spot-to-spot distances in the arrays. Recording of data can be achieved by depositing the mixtures using high speed ink jet printers, and readout rates will increase with improvements in mass spectrometry techniques. Although not discussed in this paper, rewritability and computation can be envisioned by choosing the appropriate materials. With these many avenues for growth, this new approach of using small molecules for data storage presents a highly compelling opportunity for information technology.
(SAMDI), the mixtures are immobilized along an array plate with gold islands supporting self-assembled monolayers and then analyzed for readout. As an initial demonstration of feasibility, 400 kilobits of information in both text and image format were recorded and recovered with better than 99% accuracy at write speeds of 8 bits/s and read rates of 20 bits/s. This format affords wide flexibilitythe molecules need not be oligopeptides but only analytically differentiable and chosen based on large-scale commercial availability and long-term stability. The spatial distribution of the multiplexed bytes allows straightforward random access to the data similar to that of today’s disk storage, a feature that is not routinely achieved with DNA storage where strings of bits are recorded in solutions of the macromolecules requiring large-scale decoding to find embedded sections of data. The geometry of the arrays is consistent with today’s disklike storage, making this molecular storage approach potentially compatible with the handling and robotics of current archival systems.
Author Information
E-mail:
[email protected].
■
REFERENCES REFERENCES (1) Rizzatti, L. Digital Data Storage is Undergoing Mind-Boggling Growth. EE Times, September 14, 2016; https://www.eetimes.com/ author.asp?section_id=36&doc_id=1330462. (2) Data storage supply and demand worldwide, from 2009 to 2020 (in exabytes). Statistics Portal; Statista, 2019; https://www.statista. com/statistics/751749/worldwide-data-storage-capacity-anddemand/. (3) Zhang, J.; Gecevicius, M.; Beresna, M.; Kazansky, P. G. Seemingly unlimited lifetime data storage in nanostructured glass. Phys. Rev. Lett. 2014, 112 (3), No. 033901. (4) Zhong, T.; Kindem, J.; Bartholomew, J.; Rochman, J.; Craiciu, I.; et al. Nanophotonic rare-earth quantum memory with optically controlled retrieval. Science 2017, 357 (6358), 1392. (5) Cafferty, B. J.; Ten, A. S.; Fink, M. J.; Morey, S.; Preston, D. J.; Mrksich, M.; Whitesides, G. M. Storage of Information using Small Organic Molecules. ACS Cent. Sci. 2019, DOI: 10.1021/acscentsci.9b00210. (6) Rutten, M. G. T. A.; Vaandrager, F. W.; Elemans, J. A. A. W.; Nolte, R. J. M. Encoding information into polymers. Nat. Rev. 2018, 2, 365. (7) Molteni, M. The Rise of DNA Storage. Wired, June 26, 2018; https://www.wired.com/story/the-rise-of-dna-data-storage/. (8) Brown Researchers Aim to Store Data in Molecules. News from Brown; Brown University, 2018; https://www.brown.edu/news/ 2018-01-22/chemcpus.
The spatial distribution of the multiplexed bytes allows straightforward random access to the data similar to that of today’s disk storage, a feature that is not routinely achieved with DNA storage where strings of bits are recorded in solutions of the macromolecules requiring largescale decoding to find embedded sections of data. While still in the early stages of development, paths to performance improvements can be easily mapped out. B
DOI: 10.1021/acscentsci.9b00439 ACS Cent. Sci. XXXX, XXX, XXX−XXX
