arXiv:2108.06067v1 [math.AT] 13 Aug 2021 h aea hs admitting those as same the ntr group unitary eeeaetoform two degenerate udes that so bundle moe ormv h word the remove to imposed oxcat rthe or chart, boux chart. Fact. ooooycaso h ypetcform symplectic of the cohomology of class cohomology hoe 1.1. Theorem of omtoso 2 a of formations o ftefr 2 form the of nor 3 ubrof number ..K1 n bevn h iia oa et ubr fsy of numbers Betti total minimal the observing and K41, e.g. oa et ubr fcmlxmnflswt epc oterdime Conjecture their formulated: to he respect fact, with above the complex to of and numbers Riemann Betti of total work the featuring realm four, complex and two dimension in except symplectic fl smin is ifold sahee yHp n aaiEkanmnflsi l vndimension even all in manifolds Calabi-Eckmann and Hopf n by achieved is 4 n iioeic[]poe hssaeeti athlsi the in holds fact in statement this proved Milivojevi´c [1] and iefr:fis rw ieryadte asot e lo[1, also See out. flats then t and gives thus linearly grows conjecture This first graph. form: linear nice nice a into fitting ilimply will 3 = Sp h ypetccniini ietyrflce ntetplg fclosed of the in reflected directly is condition symplectic The h group The 1 ncnrs,hwvr h oooia osqec ftecomp the of consequence topological the however, contrast, In h beto hsppri ocnr hscnetr xetwhen except conjecture this confirm to is paper this of object The ttrsotqiesrrsnl,ee huhtecnetr smad is conjecture the though even surprisingly, quite out turns It LOTCMLXMNFLSWT OA ET UBRTHREE NUMBER BETTI TOTAL WITH MANIFOLDS COMPLEX ALMOST codn oSlia,h a nprdb o nomto is information how by inspired was he Sullivan, to According (2 h iia oa et ubr(..tesmo l et ubr)o numbers) Betti all of sum the (i.e. number Betti total minimal The n, in2 sion epoeteol ipycnetdcoe ope aiodh complex closed plane. connected projective simply complex only the prove we Abstract. R M aiodis manifold { ) n ∩ n (Sullivan) Sp is 1 + should O ≥ CP J U Let (2 ≥ 2 (2 sfu,tu ofimn ojcueo ulvnecp for except Sullivan of conjecture a confirming thus four, is 8 S ( n , n n = n, 6 4 n eso h iia oa et ubro lsdams comp almost closed a of number Betti total minimal the show We dmninlra etrsaesaeacmo aia opc s compact maximal common a share space vector real –dimensional 4. l complex hsimplies: This . .Cneunl mohmnflsadmitting manifolds smooth Consequently ). } M = ) disn ope structure. complex no admits for − R . emro otesmlci el,Dni ulvnpooe osuyt study to proposes Sullivan Dennis realm, symplectic the to mirror be ω fsmlci rnfrain n h group the and transformations symplectic of ) id n . eaclosed a be hl nams ope tutr sa automorphism an is structure complex almost an while , l GL ihaygvnReana metric, Riemannian given any With . 1. + 1 ≥ h iia oa et ubro 2 a of number Betti total minimal The ( 0 u eutblwstlstelte aete etout. left they case latter the settles below result Our 10. odto htteNjnastensor Nijenhaus the that condition n, almost lotcomplex almost C ) ∩ lotcomplex almost hyae the are: They . O (2 n = ) 1. ω Introduction eeae urn hti smrhct h eRham de the to isomorphic is that subring a generates IHOHU JIAHAO U tutrs nams ypetcsrcuei non- a is structure symplectic almost An structures. ( n .Todffrn nerblt odtosaeusually are conditions integrability different Two ). ecrepniggahfrcmlxmnflsaparticularl a manifolds complex for graph corresponding he 1 aiodo ieso 2 dimension of manifold petcmnflsgo ierywt epc odimension to respect with linearly grow manifolds mplectic symplectic i.1]. Fig. euiul raie hog rpsi uddynamics, fluid in graphs through organized beautifully lotcomplex almost vn oa et ubrtrei the is three number Betti total aving ω condition N and n J dmninlclosed –dimensional e odto swdl unknown, widely is condition lex n for e lotsymplectic almost GL sos nmro correspondence mirror In nsions. iligalclholomorphic local a yielding 0 = ieso .Aogteway, the Along 6. dimension J s3 ent oa et number Betti total note We 3. is oar.Mtvtdb htthe that by Motivated Kodaira. ( eeae n nte nview in another one generates n, complex dω e aiodo dimen- of manifold lex n C aeoywhen category .As ofimto for confirmation a Also s. 2 a f ≥ aiod:ted Rham de the manifolds: fcmlxlna trans- linear complex of ) iligalclDar- local a yielding 0 = .Te h oa Betti total the Then 8. n aiod,Albanese manifolds, dmninlclosed –dimensional J ftetangent the of tutrsare structures complex ubgroup—the n eminimal he sneither is man- y , 2 JIAHAO HU

By Poincar´eduality, the total Betti number of a (nonempty positive dimensional) closed orientable manifold is at least two. In [1] it is shown (positive dimensional) almost complex manifolds having total Betti number exactly two must live in dimension 2 and 6. Theorem 1.1 is therefore achieved by studying almost complex manifolds with total Betti number 3. Our main theorem in this direction is:

Theorem 1.2. Let M be a positive dimensional closed almost complex manifold whose total Betti number is 3. Then dim M = 4 and M is complex cobordant to CP2. As consequences of Theorem 1.2, we immediately obtain: Theorem 1.3. Let M be a simply connected closed almost complex manifold whose total Betti number is 3. Then M is homeomorphic to CP2. Theorem 1.4. Let M be a simply connected closed complex manifold whose total Betti number is 3. Then M is biholomorphic to CP2. Proof of Theorem 1.3. The only zero–dimensional simply connected manifold is R0, whose total Betti number is one, hence dim M > 0. Then by Theorem 1.2, M is a 4–manifold. Since M is now assumed 2 to be simply connected, H1(M; Z) = 0. Thus by the universal coefficient theorem, H (M; Z) is free. Furthermore by Poincar´eduality and the total Betti number three assumption, H2(M; Z) is of rank one. So the intersection form of M is determined by its signature, which is equal to one by Theorem 1.2 (because signature is a cobordism invariant). Our theorem then follows from Freedman’s theorem [2, Theorem 1.5] that the homeomorphism type of a simply connected closed smooth 4–manifold is determined by its intersection form.  Proof of Theorem 1.4. This follows from Theorem 1.3 and Yau’s theorem [7, Theorem 5] that any compact complex surface homotopy equivalent to CP2 is biholomorphic to CP2.  It should be noted the simply-connectedness assumption is essential to both Theorem 1.3 and Theorem 1.4. There do exist non-simply-connected complex surfaces, so called fake projective planes (see e.g. [6]), having the same Betti numbers as CP2. The heart of our proof of Theorem 1.2 lies in the integrality of the signature and (especially) the Todd genus of an almost complex manifold. Another important ingredient is, on an almost complex manifold the Todd class coincides with the Aˆ class up to the exponential of half the first Chern class (see [3, pp. 197]): (1.1) Aˆ = ec1/2 Td . In particular if all the Chern numbers that involves the first Chern class are zero, then the Todd genus is equal to the Aˆ genus. Our major technical tool is therefore the intimate relation between the signature and the Aˆ genus, which we will establish in Section 2. In Section 3 we prove Theorem 1.2 and derive Theorem 1.1 as an easy corollary.

Acknowledgements Theorem 1.1 was also independently obtained by Zhixu Su and was announced by her in a conference in 2018; her argument differs from ours (see Remark 3.3 for a sketch of her proof). The author would like to thank Su for informing him on this matter and for sharing her proof. The author also wishes to thank Dennis Sullivan and Aleksandar Milivojevi´cfor their encouragement, helpful discussions and valuable comments on early versions of this paper.

2. Relation between the signature and Aˆ genus Let us introduce our notation. Throughout M is a nonempty closed oriented smooth manifold of dimension dim M = 2n > 0. The total Betti number of M is the sum of all Betti numbers of M, i th namely i 0 dim H (M; Q). By pi we mean the i Pontryagin class of M. If M is further almost P ≥ th complex, ci will be its i Chern class. The symbol M means pairing a cohomology class with the R ALMOST COMPLEX MANIFOLDS WITH TOTAL BETTI NUMBER THREE 3 fundamental class of M. The signature, Euler characteristic, Aˆ genus and Todd genus (if defined) of M will be denoted by σ(M), χ(M), Aˆ(M) and Td(M) respectively. Since the appearance of Hirzebruch’s celebrated work on multiplicative sequences and his signature theorem, σ(M) and Aˆ(M) are known to be rational linear combinations of Pontryagin numbers. That is,

ˆ (2.1) σ(M)= hi1,...,ir pi1 ··· pir , A(M)= ai1,...,ir pi1 ··· pir . X ZM X ZM

The coefficients of M pm can be obtained by applying the Cauchy formula to the characteristic power R ˆ √z √z/2 series associated to the signature and A genus, which are Qσ(z) = tanh √z and QAˆ(z) = sinh(√z/2) respectively (see [4, pp. 16,19], [3, pp.9-11]). The results are well-known (cf. [3, pp. 12-13]): 22m(22m 1 − 1) −B h = − B , a = m . m (2m)! m m 2(2m)! Here (2m)! · ζ(2m) (2.2) Bm = 2m 1 2m 2 − π is the mth nontrivial Bernoulli number without sign, ζ is the Riemann zeta function (see [4, pp. 129-131]). For instance 1 1 1 1 7 1 B = ,h = ,a = − ; B = ,h = ,a = − . 1 6 1 3 1 24 2 30 2 45 2 1440 Clearly hm and am are related by 2m+1 2m 1 (2.3) hm = −2 (2 − − 1)am. This immediately proves the following well-known lemma (cf. [4, pp. 90]).

Lemma 2.1. Let M be a manifold of dimension 4k with M pk being the only possibly nonzero Pontryagin number. Then R 2k+1 2k 1 σ(M)+2 (2 − − 1)Aˆ(M)=0. We will need a generalization of Lemma 2.1 in the following situation. 2 Lemma 2.2. Let M be a manifold of dimension 8k with M pk and M p2k being the only possibly nonzero Pontryagin numbers. Then R R 2 4k+1 4k 1 ˆ 4k 2k 2 Bk 2 σ(M)+2 (2 − − 1)A(M)=2 (2 − 1) pk. 2(2k)! ZM Proof. From Equation (2.1) we have

2 ˆ 2 σ(M)= h2k p2k + hk,k pk, A(M)= a2k p2k + ak,k pk. ZM ZM ZM ZM

By [3, Lemma 1.4.1] hk,k and ak,k can be expressed as 1 1 1 1 (2.4) h = h2 − h , a = a2 − a . k,k 2 k 2 2k k,k 2 k 2 2k Therefore we get 4k+1 4k 1 σ(M)+2 (2 − − 1)Aˆ(M)

4k+1 4k 1 4k+1 4k 1 2 = (h2k +2 (2 − − 1)a2k) p2k + (hk,k +2 (2 − − 1)ak,k) pk ZM ZM 4k 2k 2 2 2 =2 (2 − 1) ak pk ZM 2 4k 2k 2 Bk 2 =2 (2 − 1) pk. 2(2k)! ZM 4 JIAHAO HU

Here the second to last equality follows from Equation (2.3) and Equation (2.4).  Next we turn to almost complex manifolds.

2 Lemma 2.3. Let M be an almost complex manifold of dimension 8k with M c2k and M c4k being the only possibly nonzero Chern numbers. Then R R

4k+1 4k 1 2k+1 4k+1 2k 2 B2k 2 (2 − − 1)(1 − rk)+(3 − 2 ) · Td(M)=(1+ rk) · σ(M) − 2 (2 − 1) · χ(M), h i (4k)! 4k 2 where rk is defined by 2k Bk · rk = B2k.  Proof. Since M is assumed to be almost complex, its top Chern class coincides with its Euler class, therefore M c4k = χ(M). Also its Pontryagin numbers can be recovered from its Chern numbers. The only possiblyR nonzero Pontryagin numbers are:

2 k 2 2 2 pk = [(−1) 2c2k] =4 c2k, p2k = c2k +2 c4k. ZM ZM ZM ZM ZM ZM

Notice all the Chern numbers involving c1 vanish, so by Equation (1.1) we have Aˆ(M) = Td(M). Therefore from the proof of Lemma 2.2 we have

4k+1 4k 1 4k+2 2k 2 2 2 (2.5) σ(M)+2 (2 − − 1) Td(M)=2 (2 − 1) ak c2k. ZM On the other hand, we also have

ˆ 2 2 (2.6) Td(M)= A(M)= a2k p2k + ak,k pk = (a2k +4ak,k) c2k +2a2kχ(M). ZM ZM ZM 2 Combining Equation (2.5) and Equation (2.6) to eliminate M c2k, we get 4k+2 2k 2 2 R 4k+1 4k 1 (2.7) 2 (2 − 1) ak[Td(M) − 2a2kχ(M)] = (a2k +4ak,k)[σ(M)+2 (2 − − 1) Td(M)]. We note a2k +4ak,k 2 =1+ rk. 2ak 2 Then the desired identity is a straightforward consequence of Equation (2.7) by dividing by 2ak and reorganizing the terms.  We end this section by estimating the sizes and the relative sizes of the coefficients appearing in Lemma 2.3. For simplicity, let us denote

4k+1 4k 1 2k+1 4k+1 2k 2 B2k C =2 [(2 − − 1)(1 − r )+(3 − 2 )], C =1+ r , C =2 (2 − 1) . Td k σ k χ (4k)! Then the conclusion of Lemma 2.3 reads as

CTd · Td(M)= Cσ · σ(M) − Cχ · χ(M).

Lemma 2.4. We have the estimates

3 2 4k 3 8k 3 1 < C < , 0 < C < 8( ) , C ≥ · 2 − . σ 2 χ π Td 5 1 ζ(4k) Proof. By Equation (2.2) and the definition of rk, we see rk = 2 ζ(2k)2 . Since 1 1 1 ζ(2k)2 = > = ζ(4k), m2k m2k m4k  mX1  mX1  mX1 ≥ ≥ ≥ we have 3 1 < C =1+ r < . σ k 2 ALMOST COMPLEX MANIFOLDS WITH TOTAL BETTI NUMBER THREE 5

By Equation (2.2) and ζ(4k) < ζ(2) = π2/6 < 2 we get

2k 2 2 − 1 2 4k 0 < Cχ =4 ζ(4k) < 8( ) .  π2k  π Finally, for k ≥ 2

4k+1 1 4k 1 2k+1 4k+1 4k 2 2k+1 8k 2 3 8k 3 C > 2 [ (2 − − 1)+(3 − 2 )] > 2 [2 − − 2 ] ≥ 2 − > · 2 − . Td 2 5 1 3 5 As for k = 1, one can directly verify r1 = 5 and CTd = 5 · 2 . 

Corollary 2.5. We have the further estimates C C Td > 4k, σ ≥ 3k. Cσ + Cχ Cχ Proof. By Lemma 2.4 8k 3 8k 3 8k 3 CTd 3 · 2 − 3 · 2 − 2 − k > 4k ≥ 4 > > 4 , Cσ + Cχ 15/2+ 40(2/π) 15/2+ 40(2/π) 5 and C 1 π σ > ( )4k ≥ 3k for k ≥ 3. Cχ 8 2 For k =1, 2 one can directly verify Cσ =3, 15 respectively, so the desired estimate holds.  Cχ The estimates in Lemma 2.4 and Corollary 2.5 are not sharp. Nevertheless they will be sufficient for our applications.

3. Proof of main theorems In this section we prove Theorem 1.1 and Theorem 1.2. Recall the following theorem of Milnor (cf. [5]). Theorem 3.1 (Milnor). Two stably almost complex manifolds are complex cobordant if and only if they have the same Chern numbers.

2 Proposition 3.2. If M is an almost complex manifold of dimension 8k with M c2k and M c4k being the only possibly nonzero Chern numbers. Then either M is complex cobordantR to the emptyR manifold or the total Betti number of M is at least 3k. Proof. If Td(M) 6= 0, then since Td(M) is an integer, by Lemma 2.3 and Corollary 2.5 we get C Total Betti number ≥ max{|σ(M)|, |χ(M)|} > Td | Td(M)| > 4k. Cσ + Cχ If Td(M) = 0 but σ(M) 6= 0, then again by Lemma 2.3 and Corollary 2.5 we have C Total Betti number ≥ |χ(M)| = σ |σ(M)|≥ 3k. Cχ 2 If Td(M) = σ(M) = 0, then from Equation (2.5) and Equation (2.6) we have M c2k = 0 and R M c4k = χ(M) = 0. So all the Chern numbers of M are zero, hence M is complex cobordant to the Rempty manifold by Theorem 3.1. 

We are ready to prove our main theorems. Theorem 1.2. Let M be a positive dimensional closed almost complex manifold whose total Betti number is 3. Then dim M = 4 and M is complex cobordant to CP2. 6 JIAHAO HU

Proof. By [1, Theorem 2.2, Theorem 3.3, Remark 4.2] dim M is either 4 or divisible by 211. We must show the latter cannot happen. Assume otherwise dim M =8k for some k ≥ 2. Since the total Betti number of M is assumed to be three, by Poincar´eduality the only non-trivial rational cohomology groups are H0, H4k and H8k, each of which is one-dimensional. Therefore the only possibly nonzero 2 Chern numbers are M c2k and M c4k. Then by Proposition 3.2, either M is complex cobordant to R R the empty manifold which contradicts with M c4k = χ(M)=3 6= 0 (if it is complex cobordant to the empty manifold then all its Chern numbersR are zero by Theorem 3.1), or the total Betti number is at least 3k ≥ 9 which contradicts the total Betti number three assumption. This completes the proof of our first assertion.

To prove the second assertion, notice now M c2 = χ(M) = 3 and R p1 1 2 1 2 σ(M)= = (c1 − 2c2)= c1 − 2. ZM 3 ZM 3 3 ZM Meanwhile (cf. [3, pp. 14]),

1 2 1 2 1 Td(M)= (c1 + c2)= c1 + . ZM 12 12 ZM 4 It follows that σ(M)+3=4 · Td(M), so σ(M) ≡ 1 modulo 4. Similar as before, by Poincar´eduality and the total Betti number three assumption, the middle cohomology is one-dimensional. So σ(M) 2 must be one. This in turn implies M c1 = 9. We observe the Chern numbers of M are the same as those of CP2. So by Theorem 3.1 weR conclude M is complex cobordant to CP2.  Remark 3.3. We here communicate another proof of that 8k-manifolds with total Betti number three cannot admit almost complex structures, independently obtained by Zhixu Su2. Briefly, in such case 2 the Chern classes c1,...,c2k 1 are torsion, then from Hattori-Stong relations one can show c − M 2k is divisible by [(2k − 1)!]2. This extra divisibility combined with a careful 2-adic examinationR of the signature equation gives the result. Theorem 1.1. Let M be a closed almost complex manifold of dimension 2n ≥ 8. Then the total Betti number of M is ≥ 4. Proof. This follows from Theorem 1.2 and [1, Theorem 2.2] that almost complex manifolds with total Betti number two exist only in dimension 2 and 6.  References [1] Michael Albanese and Aleksandar Milivojevi´c. On the minimal sum of Betti numbers of an almost complex manifold. Differential Geom. Appl., 62:101–108, 2019. [2] Michael Hartley Freedman. The topology of four-dimensional manifolds. J. Differential Geometry, 17(3):357–453, 1982. [3] . Topological methods in algebraic geometry. Classics in . Springer-Verlag, Berlin, 1995. Translated from the German and Appendix One by R. L. E. Schwarzenberger, With a preface to the third English edition by the author and Schwarzenberger, Appendix Two by A. Borel, Reprint of the 1978 edition. [4] Friedrich Hirzebruch, Thomas Berger, and Rainer Jung. Manifolds and modular forms. Aspects of Mathematics, E20. Friedr. Vieweg & Sohn, Braunschweig, 1992. With appendices by Nils-Peter Skoruppa and by Paul Baum. [5] J. Milnor. On the cobordism ring Ω∗ and a complex analogue. I. Amer. J. Math., 82:505–521, 1960. [6] Gopal Prasad and Sai-Kee Yeung. Fake projective planes. Invent. Math., 168(2):321–370, 2007. [7] Shing Tung Yau. Calabi’s conjecture and some new results in algebraic geometry. Proc. Nat. Acad. Sci. U.S.A., 74(5):1798–1799, 1977.

Stony Brook University, Department of Mathematics, 100 Nicolls Road, 11794 Stony Brook Email address: [email protected]

2Our proof above was attained in summer 2019; later the author learned from Su that she had also proved this and announced it in a conference in 2018.