Papex

Open problem · Mathematics · Open

Smooth Poincaré conjecture

In which dimensions is every smooth manifold that is topologically a sphere also smoothly the standard sphere?

Summary

Exotic spheres make the statement false in many dimensions, starting with 7, while it holds in dimensions such as 1, 2, 3, 5, 6, 12, 56 and 61. Dimension four is the main open case, and it is not settled whether the statement holds in any even dimension of 140 or more.

Why it is open

A homotopy n-sphere is a smooth n-dimensional manifold that is homotopy equivalent to the n-sphere [5]. The topological Poincaré conjecture, that such a manifold is homeomorphic to the sphere, is true in every dimension: Michael Freedman proved it in dimension four in 1982 and Perelman in dimension three [3, 4]. The smooth version asks whether such a manifold is also diffeomorphic to the sphere, that is, equivalent as a smooth manifold [4]. One that is homeomorphic but not diffeomorphic to the sphere is called exotic [12]. John Milnor describes the smooth version as "true in some dimensions and false in others, while remaining totally mysterious in dimension 4" [3].

The problem has two parts. Above dimension four, whether exotic spheres exist reduces to the stable homotopy groups of spheres, and the answer is known in every odd dimension and in every even dimension below 140 [4, 5]. Even dimensions of 140 and above are not covered by these results [5].

Dimension four, the smooth four-dimensional Poincaré conjecture (SPC4), is the harder part. The group of homotopy 4-spheres up to h-cobordism is trivial, but Milnor calls the question whether they are all diffeomorphic to the sphere "the great unsolved problem" [3]. The h-cobordism theorem, the usual tool for building diffeomorphisms, fails for smooth four-dimensional h-cobordisms [2]. There are no known diffeomorphism invariants for homotopy spheres [8], and Freedman, Robert Gompf, Scott Morrison and Kevin Walker explain that gauge-theoretic slice obstructions cannot tell an exotic four-ball from the standard one [2].

No prize sets the bar. A resolution of the four-dimensional case would be a proof that every homotopy 4-sphere is diffeomorphic to the 4-sphere, or a homotopy 4-sphere shown not to be.

What is known

Exotic spheres (1956–1963). Milnor constructed an exotic smooth structure on the 7-sphere in 1956 [4]. Michel Kervaire and Milnor then defined a group of homotopy n-spheres, which classifies the smooth structures on the n-sphere for n ≥ 5, and related it to the stable homotopy groups of spheres [4]. These groups are finite except possibly in dimension four, and there are 28 homotopy 7-spheres up to oriented diffeomorphism [3]. In dimension three the answer is yes: by a theorem of Moise every closed 3-manifold has a unique smooth structure [4].

Higher dimensions (1963–2024). In dimensions 4k + 3 with k ≥ 1 the sphere always has exotic structures [4]. In dimensions 4k + 1 the answer depends on the Kervaire invariant, which Michael Hill, Michael Hopkins and Douglas Ravenel showed can be non-zero only in dimensions 2, 6, 14, 30, 62 and possibly 126 [4]; in a 2024 preprint, not peer reviewed, Weinan Lin, Guozhen Wang and Zhouli Xu report that 126 occurs as well [10]. In 2017 Wang and Xu proved that the 61-sphere has a unique smooth structure and that the only odd-dimensional spheres with this property are those of dimensions 1, 3, 5 and 61 [4]. Daniel Isaksen's corrected computations showed that the 56-sphere also has a unique smooth structure [4]. Mark Behrens, Hill, Hopkins and Mark Mahowald then showed that the only dimensions below 140 in which the sphere has a unique smooth structure are 1, 2, 3, 5, 6, 12, 56 and 61, and perhaps 4 [5]. Using motivic homotopy theory and machine computations, Isaksen, Wang and Xu determined the groups of homotopy spheres through dimension 90, except in dimension four [6]. Wang and Xu conjecture that above dimension four the only spheres with a unique smooth structure are those of dimensions 5, 6, 12, 56 and 61 [4]; Behrens and colleagues write that the general belief is that there are finitely many such dimensions [5].

Dimension four (1982–2025). Freedman's topological classification and Donaldson's gauge theory together showed that four-dimensional Euclidean space has uncountably many smooth structures, while in every other dimension it has only one [3]. Exotic versions of four-space can be built from knots that are topologically but not smoothly slice [9]. Candidate exotic 4-spheres come from constructions of Cappell and Shaneson in 1976, from Gluck twists, which cut out and reglue a neighbourhood of a knotted 2-sphere, and from presentations of the trivial group [2]. Some have been shown to be standard. In 2010 Selman Akbulut proved this for an infinite family of Cappell–Shaneson spheres [1], and Kazunori Iwaki added 145 further infinite families in 2025 [11]. Gluck twists are known to give the standard sphere for many 2-knots, such as twist-spun knots [2]. In a paper posted in 2009, Freedman and colleagues wrote that researchers' opinion "tends to lean in the direction that the conjecture is false", and that Akbulut's work "makes SPC4 seem more plausible" [2].

Recent claim (2023). In 2023 Akio Kawauchi published a paper in WSEAS Transactions on Mathematics stating that every smooth homotopy 4-sphere is diffeomorphic to the 4-sphere [7]. Papers published since, by Iwaki and by Sergei Gukov and colleagues, still treat the conjecture as open [11, 12].

Main approaches

Standardising candidates. Show by handle calculus that each candidate exotic sphere is diffeomorphic to the 4-sphere [1, 11]. Strongest evidence: the Cappell–Shaneson families above [1, 11]. Main difficulty: Gompf's conjecture on Cappell–Shaneson matrices, which would make all these spheres standard, is confirmed only for some ranges of the matrix trace [11], and Gluck twists are not known to be standard in general [2].

Slicing knots. Find a homotopy 4-sphere and a knot that bounds a smooth disc in it, minus a ball, but not in the standard 4-ball; such a sphere would be exotic [8]. Rasmussen's s-invariant, from Khovanov homology, can show that a knot bounds no disc in the standard ball [2]. Strongest evidence: Ciprian Manolescu and Lisa Piccirillo built candidates from pairs of knots with the same zero-surgery, found five knots that would each give an exotic 4-sphere if slice, and write that the strategy "seems to be the only presently tractable approach" [8]. Main difficulty: Freedman and colleagues found s = 0 after 766 hours of computation [2]. According to Gukov and colleagues, Nakamura later showed that s does not help for Manolescu and Piccirillo's family of 3,375 pairs; after algebraic obstructions and a machine-learning search for ribbon discs, only two pairs could still give counterexamples, which the authors consider unlikely [12]. They describe the ribbon search as a harder problem of the same type as the unknotting problem [12].

Khovanov-type invariants of 4-manifolds. Build invariants of the whole 4-manifold from Khovanov homology, as Freedman and colleagues proposed with a skein module [2]. Strongest evidence: in a preprint, Qiuyu Ren and Michael Willis show that the Khovanov skein lasagna module distinguishes an exotic pair of 4-manifolds, the first such proof without analysis [9]. Main difficulty: over the rationals it cannot detect homotopy 4-spheres that become standard after adding copies of the complex projective plane, which includes all Gluck twists [9].

Positive arguments. Freedman and colleagues suggest a symplectic route through a theorem of Gromov, and show that the conjecture is equivalent to a generalisation of Gabai's Property R about surgery on knots [2]. Strongest evidence: Gabai's theorem that in the 3-sphere only the unknot yields S¹ × S² under surgery, the case that the generalisation would extend [2]. Main difficulty: the symplectic argument "would have to be special to homotopy spheres" [2].

Open sub-questions

  1. The last two pairs. In the pairs KB/G(−2, 0, 0, −1, 2, −1) and KB/G(−1, 0, −1, −1, 2, −1), in Manolescu and Piccirillo's notation, is one knot slice and the other not, which would give an exotic 4-sphere [8, 12]?
  2. Gompf's conjecture at trace 70. Is every Cappell–Shaneson matrix of trace 70, outside the ranges confirmed by Kim and Yamada, Gompf equivalent to the matrix A0 [11]?
  3. Lasagna modules over finite fields. Over a finite field, does the Khovanov skein lasagna module also fail to detect homotopy 4-spheres that become standard after adding copies of the complex projective plane [9]?
  4. Even dimensions from 140. Is there an even n ≥ 140 for which the cokernel of the J-homomorphism in dimension n has no non-zero element of Kervaire invariant zero, giving a unique smooth n-sphere [4, 5]?

Work on this in Papex

If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It is not peer review and does not mean the hypothesis is correct.

Key references

  1. Akbulut, S. (2010). Cappell–Shaneson homotopy spheres are standard. Annals of Mathematics 171(3), 2171–2175. 10.4007/annals.2010.171.2171 · Checked 16 Oct 2026
  2. Freedman, M., Gompf, R., Morrison, S. & Walker, K. (2010). Man and machine thinking about the smooth 4-dimensional Poincaré conjecture. Quantum Topology 1(2), 171–208. 10.4171/QT/5 · Checked 16 Oct 2026
  3. Milnor, J. (2011). Differential topology forty-six years later. Notices of the American Mathematical Society 58(6), 804–809. Source · Checked 16 Oct 2026
  4. Wang, G. & Xu, Z. (2017). The triviality of the 61-stem in the stable homotopy groups of spheres. Annals of Mathematics 186(2), 501–580. 10.4007/annals.2017.186.2.3 · Checked 16 Oct 2026
  5. Behrens, M., Hill, M., Hopkins, M. J. & Mahowald, M. (2020). Detecting exotic spheres in low dimensions using coker J. Journal of the London Mathematical Society 101(3), 1173–1218. 10.1112/jlms.12301 · Checked 16 Oct 2026
  6. Isaksen, D. C., Wang, G. & Xu, Z. (2020). Stable homotopy groups of spheres. Proceedings of the National Academy of Sciences 117(40), 24757–24763. 10.1073/pnas.2012335117 · Checked 16 Oct 2026
  7. Kawauchi, A. (2023). Smooth homotopy 4-sphere. WSEAS Transactions on Mathematics 22, 690–701. 10.37394/23206.2023.22.76 · Checked 16 Oct 2026
  8. Manolescu, C. & Piccirillo, L. (2023). From zero surgeries to candidates for exotic definite four-manifolds. Journal of the London Mathematical Society 108(5), 2001–2036. 10.1112/jlms.12800 · Checked 16 Oct 2026
  9. Ren, Q. & Willis, M. (2024). Khovanov homology and exotic 4-manifolds. Preprint, arXiv:2402.10452 (version 3, 18 December 2025), not peer reviewed. Source · Checked 16 Oct 2026
  10. Lin, W., Wang, G. & Xu, Z. (2024). On the last Kervaire invariant problem. Preprint, arXiv:2412.10879 (version 2, 22 February 2025), not peer reviewed. Source · Checked 16 Oct 2026
  11. Iwaki, K. (2025). Infinite families of standard Cappell–Shaneson homotopy 4-spheres. Topology and its Applications 366, 109293. 10.1016/j.topol.2025.109293 · Checked 16 Oct 2026
  12. Gukov, S., Halverson, J., Manolescu, C. & Ruehle, F. (2025). Searching for ribbons with machine learning. Machine Learning: Science and Technology 6(2), 025065. 10.1088/2632-2153/ade362 · Checked 16 Oct 2026

Page reviewed 16 Oct 2026

Source

Open, as listed by Wikipedia · Unsolved problems in mathematics. Checked 08 Oct 2026.

The status is the source’s. Papex does not decide whether a problem is solved.

Records on this problem

No public record addresses this problem yet.

More in Mathematics