ArXiv preprint proposes nuclear-spin laser method for >90% single-pass isotope enrichment

A newly posted arXiv preprint claims a laser-based isotope separation method built around nuclear spin, rather than the tiny isotope shifts used in conventional approaches, could theoretically deliver more than 90% single-pass enrichment from natural feed. In the abstract, the paper says “selectivity arises from nuclear spin rather than isotope shifts” and that “single-pass enrichment exceeding 90% from natural feed is achievable without cascading.”

The claim comes with major caveats. The paper, “Separation of even-even from even-odd isotopes using ultrafast lasers,” was posted to arXiv as arXiv:2605.00959v1 and, according to the arXiv entry, was submitted May 1, 2026. The author listed on the abstract page is Jacob Levitt. The public record describes it as a theoretical proposal supported by density-matrix simulations, not an experimental report, and there is no evidence it has been peer reviewed or published in a journal. No independent experimental verification is reported in the available material.

In plain terms, the proposal depends on a basic nuclear-physics distinction. Even-even isotopes — those with even numbers of protons and neutrons — generally have nuclear spin of zero. Many odd-mass isotopes have nonzero nuclear spin, which creates hyperfine splitting, a small internal structure in their energy levels. The preprint proposes using a Ramsey pulse sequence, a standard two-pulse technique, on paramagnetic molecular isotopologues, meaning molecules that differ only by isotope. In the paper’s account, isotopologues with spin 0 rephase cleanly and return to the ground state, while those with nonzero spin do not fully “echo” back because hyperfine splitting disrupts the rephasing, leaving some population in an excited state that could then be separated.

The abstract gives specific simulated numbers for that effect. For even-even isotopologues with nuclear spin I = 0, it reports a trapped excited-state fraction Pm = 0. For even-odd isotopologues with I > 0, it reports Pm of about 0.23 to 0.47 in representative systems. The examples named in the abstract are uranium-235, strontium-87 and iron-57. The arXiv record lists the paper as four pages with two figures.

What makes the proposal notable is how sharply it differs from traditional laser isotope separation. Existing laser schemes generally rely on narrowband lasers tuned to the tiny frequency differences, or isotope shifts, between isotopes. This preprint instead argues that hyperfine structure linked to nuclear spin could provide the needed selectivity, potentially allowing broadband ultrafast lasers to do the job. If that idea were ever validated in the lab and scaled, it could matter both for industrial isotope production and for nonproliferation policy, since laser enrichment methods have long drawn export-control scrutiny. That concern is especially relevant in uranium-related work because compact optical systems can be difficult to monitor and many of their components have legitimate civilian uses. Past debate over SILEX, a highly secretive laser enrichment technology, is the clearest precedent.

Still, laser enrichment has a long and sensitive history, and promising concepts on paper often run into major engineering limits when researchers try to turn them into practical systems. This preprint does not report such a demonstration. For now, the headline claim — very high single-pass enrichment using ultrafast laser pulses and nuclear-spin effects — remains a simulation-based, unreviewed result that has not been independently verified outside the arXiv posting.