A 2018 study showed that a self-assembled, two-dimensional molecular sheet could recognize one mirror-image form of a tryptophan derivative and largely ignore the other. One version of the sheet captured the L form with a reported 96% efficiency and showed no apparent affinity for the D form; the mirror-image sheet reversed that preference. Adding ammonium acetate released the captured molecules.
That result is best understood as chiral molecular recognition in a nanosheet dispersion, not as a finished industrial filter. The experiment did not demonstrate continuous, high-throughput processing, separation of every amino acid, or performance with unmodified free tryptophan. Later metal-organic and covalent-organic-framework membranes have demonstrated selective transport, but they are different materials, targets and mechanisms.
As an Amazon Associate I earn from qualifying purchases.
What the 2018 two-dimensional membrane actually was
The phrase “2D membrane sieves out amino acid enantiomers” describes a striking selectivity result, but the material was not a conventional polymer membrane with feed and permeate channels. Researchers designed chiral aromatic macrocycles that assemble into porous nanosheets in methanol. The sheets were dispersed in liquid, where their nanoscale cavities acted as recognition sites.
The molecular building block
Each building block combined a buckled aromatic macrocycle with an oligoether dendron. The macrocycle was prepared in two enantiomeric forms: each had the same atoms and connectivity as its mirror image, but its three-dimensional arrangement was opposite. That handedness was carried into the cavities formed when many molecules assembled.
#1 Best Overall
Dimer formation in methanol
In methanol, two macrocycles associated to form a dimer. Their ring-shaped faces pointed inward and sat at a relative angle of about 60 degrees. This orientation created a defined pocket rather than a flat, featureless opening. The oligoether dendrons projected outward and interacted favorably with solvent.
From dimers to a two-layer sheet
Intermolecular forces linked the dimers into extended sheets. Each layer had hexagonal symmetry, and the two layers were twisted by 60 degrees relative to one another. Together they produced a grid of homochiral nanocavities. Because the dendrons were attracted to solvent, they discouraged the sheets from simply stacking into a bulk crystal, helping maintain a dispersed, accessible nanosheet structure.
How a membrane can distinguish mirror-image molecules
Enantiomers have identical atom-to-atom connectivity and, in an achiral environment, many of the same bulk physical properties. A chiral cavity supplies the missing asymmetry. The cavity and the guest must match in three dimensions, much as a left-hand glove accepts a left hand more readily than a right hand.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchRecognition is not based on a single “L” or “D” label. It comes from the combined geometry of several contacts: the guest’s aromatic and polar groups must occupy compatible positions, and the surrounding cavity must allow favorable interactions without excessive strain. A mirror-image cavity presents the opposite arrangement of those contacts. It can therefore favor the opposite enantiomer while making the first guest difficult to bind.
Rank #2
In the 2018 sheets, the cavity was homochiral because every macrocycle had the same handedness. A sheet made from one macrocycle enantiomer favored one tryptophan-derivative enantiomer; a sheet made from the mirror-image macrocycle favored the other. This symmetry reversal is strong evidence that molecular handedness, rather than a nonspecific size or polarity effect, controlled uptake.
What happened when the tryptophan derivative was added
Selective capture from a racemic mixture
The researchers added a racemic mixture, containing equal mirror-image forms of a tryptophan derivative, to the nanosheet dispersion. One nanosheet enantiomer captured the L derivative with 96% efficiency in the reported experiment and showed no apparent affinity for the D derivative. Repeating the test with the mirror-image nanosheet switched the outcome: it captured D with the same reported efficiency and selectivity.
“Efficiency” here refers to the reported uptake result for that tested derivative under the study’s conditions. It is not a membrane throughput, a recovery yield for an industrial process, or a claim that 96% of any amino acid can be separated. The experiment also found no apparent affinity for derivatives of several other amino acids, so the demonstrated recognition was narrow rather than universal.
Recommended Free Tools
Salt-triggered release
Capture was reversible. When ammonium acetate was added, solvent molecules were drawn away from the oligoether dendrons. That changed the balance of interactions holding the assembly and caused the captured guest to be ejected. The sequence therefore had two chemically distinct stages: selective uptake by a chiral cavity, followed by release when the solvent environment was altered with salt.
Why this is not a free-tryptophan separation claim
The reported target was a tryptophan derivative, not unmodified free tryptophan. Substituting a different molecule can change size, charge distribution, hydrogen-bonding geometry and hydrophobic contacts. Those changes can determine whether a guest fits the cavity at all. The finding should therefore be stated precisely: the nanosheet recognized the tested derivative under the reported conditions.
Capture-and-release is different from a through-flow membrane
In ordinary filtration or nanofiltration, a feed solution moves continuously across a membrane and components emerge at different rates on the other side. The 2018 work instead demonstrated selective uptake from a dispersion and chemical release. A practical process based on that design would need additional steps to contact the feed, collect the nanosheets, trigger desorption, recover the product and regenerate the material.
Those engineering questions were not answered by the original report. The study did not establish long-duration cycling, membrane-area scale-up, pressure-driven flux, resistance to fouling, solvent recovery, or operation with a complex industrial feed. It also did not show that the nanosheets could be fabricated as a mechanically robust, continuous film. Calling the material a “sieve” is useful shorthand for molecular discrimination, but it should not be read as evidence of a commercial filtration module.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →The 2018 result in one view
| Stage | Reported detail | What it means |
|---|---|---|
| Building block | Buckled chiral aromatic macrocycle with an oligoether dendron | Provides a handed recognition surface and solvent-facing groups |
| Initial assembly | Methanol-induced dimers; inward-facing macrocycles at about 60° | Creates a defined molecular pocket |
| Sheet architecture | Two layers, hexagonal symmetry in each layer, 60° twist between layers | Forms a grid of homochiral nanocavities |
| Test feed | Racemic mixture of a tryptophan derivative | Challenges the cavity to distinguish mirror images |
| Preference | One sheet enantiomer captured L; the mirror sheet captured D | Handedness of the host determines handedness of the guest selected |
| Reported uptake | 96% efficiency for the favored enantiomer | Study-specific capture result, not a general membrane throughput |
| Release trigger | Ammonium acetate | Changes solvent–dendron interactions and ejects the guest |
| Scope limitation | No apparent affinity for derivatives of several other amino acids | Recognition was selective for the tested molecular shape |
What changed in later membrane research
Later studies moved from dispersed recognition sheets toward continuous porous membranes and, in some cases, selective transport. These results are related in theme but are not replications of the 2018 nanosheet experiment. Their materials, targets, operating modes and reported metrics differ.
Rank #4
Homochiral TAH-COF membrane for threonine
A 2025 Nature Communications study reported a homochiral covalent-organic-framework (COF) membrane with pores of approximately 0.6 nanometers. In a racemic threonine feed at 0.002 mol/L, operated for two hours at room temperature, the reported D-threonine flux was 1.42 mmol m−2 h−1 and the L-threonine flux was 0.03 mmol m−2 h−1. The resulting enantiomeric excess was 94.3% under those stated conditions.
At an initial concentration of 0.02 mol/L, the reported D-threonine flux increased to 3.62 mmol m−2 h−1. Flux is a transport rate per membrane area; it cannot be compared directly with the 2018 study’s 96% capture efficiency.
Why the COF membrane favors D transport
The COF authors attributed selectivity to stronger interactions between L-amino acids and the homochiral pores. L-threonine binds more strongly and is consequently retarded, while D-threonine moves through the channels more readily. Model-compound titrations reported binding constants of 372 M−1 for L-threonine and 137 M−1 for D-threonine. Calculations and simulations were used alongside the transport measurements to support this retarded-transport explanation.
Two-stage concentration experiment
The same paper described a two-stage cascade using a model mixture of four amino acids representing a hydrolysate. D-threonine rose from 20% to 85% after the first stage and to 99.5% after the second. The paper’s abstract also discusses extraction from an eight-component protein-hydrolysis complex; the detailed cascade percentages refer to the four-amino-acid model mixture and should not be conflated with the abstract’s broader description.
Best Value
Other 2025 membrane designs
A Chemical Communications report used a membrane incorporating copper ions and R- or S-tryptophan, reporting 99.38% enantiomeric excess. A separate COF nanochannel membrane study reported preferential D-isomer transport for racemic phenylalanine, with enantiomeric excess of up to 99.4%. Neither experiment used the 2018 aromatic-macrocycle nanosheet, and neither should be presented as a direct confirmation of its capture mechanism.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Comparing the architectures and operating principles
| Study | Architecture | Operation | Target and feed | Reported metric | Interpretation |
|---|---|---|---|---|---|
| 2018 nanosheet report | Self-assembled, two-layer homochiral aromatic-macrocycle sheet | Selective capture, then salt-triggered release | Racemic tryptophan derivative; other amino-acid derivatives showed no apparent affinity | 96% capture efficiency for the favored form | Molecular recognition in a dispersion; no demonstrated continuous flux |
| 2025 Nature Communications | Homochiral TAH-COF membrane; pores about 0.6 nm | Selective permeation with stronger L binding and retarded L transport | Racemic threonine; separate model-mixture cascade | 94.3% ee at specified feed and time; up to 99.5% D-threonine after two stages in the model mixture | Transport and staged enrichment, distinct from nanosheet capture |
| 2025 Chemical Communications | Copper-ion/R- or S-tryptophan MOF membrane | Membrane-based enantioselective separation | Tryptophan system | 99.38% ee | Different host chemistry and membrane architecture |
| 2025 Angewandte Chemie International Edition | COF nanochannel membrane | Preferential D-isomer transport | Racemic phenylalanine | Up to 99.4% ee | Different target and channel design |
These percentages are study-specific outcomes, not a league table. Enantiomeric excess, capture efficiency, flux and final product composition measure different things. Feed concentration, residence time, membrane area, stage count and mixture complexity all affect what a number means.
What the 2018 nanosheets prove—and what remains open
Established by the experiment
- Two mirror-image macrocycle building blocks can produce mirror-image porous nanosheets.
- The resulting cavities can distinguish enantiomers of a tested tryptophan derivative.
- The preferred guest can be released by changing the chemical environment with ammonium acetate.
- Switching the nanosheet handedness switches which guest enantiomer is captured.
Not established by the experiment
- That the nanosheets separate unmodified free tryptophan or a broad range of amino acids.
- That they operate as a pressure-driven, through-flow membrane.
- That they retain selectivity through many adsorption–release cycles.
- That they can be manufactured as large, mechanically stable membrane modules.
- That the reported uptake result translates into industrial productivity or process economics.
Myongsoo Lee, a researcher at Jilin University, expressed the appropriate level of caution in the 2018 Chemistry World report: “In terms of thinking about it as a true future technology for enantioselective separations – I wouldn’t go that far yet.” The statement fits the evidence: the work established an elegant recognition mechanism, while practical deployment would require substantial materials and process development.
Why this molecular-recognition strategy matters
Many chemical and biological products are chiral. Two enantiomers can have different biological activity even though they contain the same atoms. A separation method that recognizes handedness directly could, in principle, reduce reliance on repeated crystallization or chemically modified resolving agents. Porous organic materials are attractive because their cavity size, functional groups and chirality can be designed together.
The 2018 design also illustrates a useful separation concept: recognition and release can be independently tuned. The cavity supplies selectivity, while a salt-induced solvent change supplies a trigger for recovery. In a future process, that could be preferable to permanently binding the desired molecule. However, the trigger must be compatible with product stability, solvent recycling and material lifetime, issues not resolved in the nanosheet report.
How to interpret the headline accurately
“Sieve” should be read at the molecular scale. The nanosheet does not sift amino acids by letting small molecules pass through a macroscopic mesh. It presents homochiral cavities that bind one mirror-image guest more favorably than the other. The reported 96% figure describes that selective capture for one tryptophan derivative, followed by chemical release; it does not establish a commercial membrane that continuously filters every amino-acid enantiomer.
Later COF and MOF membranes show that chiral porous materials can also support selective permeation and high enantiomeric excess, including the reported threonine, tryptophan and phenylalanine results. Those advances broaden the field, but they should be evaluated on their own targets, architectures, fluxes and process configurations rather than folded into the 2018 result.
Free tools Windows power users keep installed
One-click scans. No signup required.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
