betaFeedback

WorksMathematics

On the homotopy automorphism group of simply-connected homology CP3\mathbb C \mathrm P^3

Abstract

In this paper, we compute the group of homotopy equivalences of closed, smooth, simply connected 66-manifolds whose integral homology groups are isomorphic to whose of the complex projective space CP3\mathbb C \mathrm P^3, building on computations of their mapping class groups.

Introduction

For a topological space XX, the group of homotopy equivalences haut⁡(X)\operatorname{haut}(X) is an object of significant interest in homotopy theory, and has been widely studied in the literature; see, for example, the survey article [2]. For a compact manifold MM, the group haut⁡(M)\operatorname{haut}(M) also plays an important role in manifold topology, particularly in the classification of manifolds and the study of group actions, see [3].

For simply connected 4- and 5-manifolds, the group haut⁡(M)\operatorname{haut}(M) has been computed by Milnor [13] and Baues-Buth [6], respectively. In dimension 6, however, the situation is less well understood. The group haut⁡(S2×S4)\operatorname{haut}(S^{2} \times S^{4}) was computed by Sawashita [17] as part of a broader study of the group of homotopy equivalences of products of spheres. Partial information on the group of homotopy equivalences of S2S^{2}-bundles over S4S^{4} has been obtained by Yamaguchi [21].

As a generalization of the results mentioned above, in this paper we compute the group haut⁡(M)\operatorname{haut}(M) for closed, smooth, simply connected 6-manifolds MM with H∗(M)≅H∗(CP3)H_{*}(M) \cong H_{*}(\mathbb{CP}^{3}) (referred to as simply connected homology CP3\mathbb{CP}^{3} for short). For this class of manifolds, classifications up to diffeomorphisms and up to homotopy equivalences have long been known. We briefly recall these results below.

By the theory of Wall ([19 Theorem 3]) and Jupp ([9 Theorem 1]), the diffeomorphism type of an oriented, simply connected 6-manifold MM with torsion-free homology is determined by the third Betti number b3(M)b_{3}(M), the cubic form

H2(M)×H2(M)×H2(M)→Z,(x,y,z)↦⟨x∪y∪z,[M]⟩H^{2}(M) \times H^{2}(M) \times H^{2}(M) \to\mathbb{Z}, \qquad(x,y,z) \mapsto\langle x \cup y \cup z,[M]\rangle

the second Stiefel-Whitney class w2(M)w_{2}(M) and the first Pontrjagin class p1(M)p_{1}(M).

When MM is a homology CP3\mathbb{CP}^{3}, the cubic form and p1(M)p_{1}(M) can be described in terms of numeric invariants as follows. Let x∈H2(M)x \in H^{2}(M), y∈H4(M)y \in H^{4}(M) be generators. Then there exists an integer dd such x2=dyx^{2}=dy. If MM is nonspin, then dd is even. By choosing the sign of yy, we may assume d≥0d \ge0. There exists an integer ll such that the first Pontrjagin class is given by

p1(M)=4(d+6l)y,if M is spin;p1(M)=(d+48l)y,if M is non-spin.p_{1}(M)=4(d+6l)y,\quad\text{if } M \text{ is spin};\qquad p_{1}(M)=(d+48l)y,\quad\text{if } M \text{ is non-spin}.

When d=0d=0, we choose the sign of yy so that l≥0l \ge0.

The homotopy classification of this class of 6-manifolds has been established by several authors (see [4 Proposition 1.9] and the discussion therein). We summarize these results in the theorem below.

Theorem 1.1. Let MM be a simply connected homology CP3\mathbb{C}P^{3}. Then MM admits an orientation-reversing diffeomorphism.

If MM is spin, there is a one-to-one correspondence between the set of diffeomorphism classes of such manifolds and the set

{(d,l)∈Z×Z∣d≥0, and l≥0 when d=0}.\{(d,l) \in\mathbb{Z} \times\mathbb{Z} \mid d \geq0,\ \text{and } l \geq0\text{ when }d=0\}.

If dd is odd, the homotopy type of MM is determined by dd; if dd is even, the homotopy type of MM is determined by dd together with d/2+l(mod2)d/2+l\pmod{2}.

If MM is non-spin, there is a one-to-one correspondence between the set of diffeomorphism classes and the set

{(d,l)∈2Z×Z∣d≥0, and l≥0 when d=0}.\{(d,l) \in2\mathbb{Z} \times\mathbb{Z} \mid d \geq0,\ \text{and } l \geq0\text{ when }d=0\}.

In this case, the homotopy type of MM is determined by dd.

We now give some example. In the spin case, the manifold with d=l=0d=l=0 is S2×S4S^{2}\times S^{4}, while the manifold with d=0d=0 and l=1l=1 is CP3\mathbb{C}P^{3}; in the non-spin case, the manifold with d=l=0d=l=0 is the nontrivial S4S^{4}-bundle over S2S^{2}, denoted by S4×~S2S^{4}\widetilde{\times}S^{2}.

Another class of geometrically interesting examples is given by the total spaces XnX_n of S2S^{2}-bundle over S4S^{4} with characteristic map n∈π3SO(3)≅Zn\in\pi_{3}SO(3)\cong\mathbb{Z}. These manifolds are spin and satisfy d=nd=n and l=0l=0. As a consequence of the homotopy classification, if MM is spin with dd odd, or with both dd and ll even, then MM is homotopy equivalent to the sphere bundle XdX_d.

A standard method in homotopy theory for studying the group of homotopy equivalences of an nn-dimensional simply connected Poincaré complex XX is provided by the exact sequence of Barcus and Barratt [5 Theorem 6.1] (see also [3, Proposition (1.3)])

Im⁡{πn(X(n−1))→πn(X)}→d∗haut⁡(X)→j∗haut⁡(X(n−1)).\operatorname{Im}\{\pi_n(X^{(n-1)}) \to\pi_n(X)\} \xrightarrow{d^*} \operatorname{haut}(X) \xrightarrow{j^*} \operatorname{haut}(X^{(n-1)}).

where X(n−1)X^{(n-1)} denotes the (n−1)(n-1)-skeleton of XX. The homomorphism

d∗:Im⁡{πn(X(n−1))→πn(X)}→haut⁡(X)d^*:\operatorname{Im}\{\pi_n(X^{(n-1)}) \to\pi_n(X)\}\to\operatorname{haut}(X)

is induced by the composition of the coaction μ:X→X∨Sn\mu:X\to X\vee S^n with a map Sn→XS^n\to X representing an element of Im⁡{πn(X(n−1))→πn(X)}\operatorname{Im}\{\pi_n(X^{(n-1)})\to\pi_n(X)\}. There are clear descriptions of Im⁡j∗\operatorname{Im}j^* and the action of Im⁡j∗\operatorname{Im}j^* on Im⁡d∗\operatorname{Im}d^*. The difficulty lies in the determination of the kernel of d∗d^*, which involves the computation of a certain homomorphism

π1(Map⁡∗(X(n−1),X),i)→πn(X).\pi_1(\operatorname{Map}_*(X^{(n-1)},X),i)\to\pi_n(X).

This method has been applied to the class of 6-manifolds under consideration by several authors. For X=S2×S4X=S^{2}\times S^{4}, Sawashita showed that this sequence extends to a split short exact sequence ([17 Theorem 4.3])

0→π6(S2×S4)→haut⁡(S2×S4)→(Z/2)3→1.0\to\pi_6(S^{2}\times S^{4})\to\operatorname{haut}(S^{2}\times S^{4})\to(\mathbb{Z}/2)^3\to1.

For X=XdX=X_d, Yamaguchi obtained an exact sequence ([21 Theorem 1.1])

π6(Xd)→haut⁡(Xd)→Gd→1(1)\pi_6(X_d)\to\operatorname{haut}(X_d)\to G_d\to1 \tag*{(1)}

with Gd≅Z/2G_d\cong\mathbb{Z}/2 when dd is odd, and Gd≅Z/2×Z/2G_d\cong\mathbb{Z}/2\times\mathbb{Z}/2 when dd is even. He also computed the group π6(Xd)\pi_6(X_d):

π6(Xd)≅{0if (d,6)=1,Z/3if (d,6)=3,Z/2×Z/(12,d/2)if d is even.(2)\pi_6(X_d)\cong \begin{cases} 0 & \text{if }(d,6)=1,\\ \mathbb{Z}/3 & \text{if }(d,6)=3,\\ \mathbb{Z}/2\times\mathbb{Z}/(12,d/2) & \text{if }d\text{ is even}. \end{cases} \tag*{(2)}

However, since the kernel of the homomorphism π6(Xd)→haut⁡(Xd)\pi_{6}(X_{d}) \to\operatorname{haut}(X_{d}) was not determined (except for the cases where π6(Xd)=0\pi_{6}(X_{d}) = 0), the problem of computing haut⁡(Xd)\operatorname{haut}(X_{d}) remained unsolved.

In this paper, we study the group of homotopy equivalences from a different perspective. In manifold topology, the mapping class group MCG⁡(M)\operatorname{MCG}(M) of a smooth manifold MM plays an analogous role to that of the group of homotopy equivalences in homotopy theory, and is itself an object of considerable interest. The group haut⁡(M)\operatorname{haut}(M) is related to group MCG⁡(M)\operatorname{MCG}(M) via the natural forgetful homomorphism ψ ⁣:MCG⁡(M)→haut⁡(M)\psi\colon\operatorname{MCG}(M) \to\operatorname{haut}(M).

As an example, for (n−1)(n-1)-connected 2n2n-dimensional almost parallelizable manifolds, both MCG⁡(M)\operatorname{MCG}(M) and haut⁡(M)\operatorname{haut}(M) have been computed by various authors, and a detailed comparison between them is given in [3 §10].

In this paper, we analyze the forgetful homomorphism ψ\psi for simply connected homology CP3\mathrm{CP}^{3}. We determine the kernel and cokernel of ψ ⁣:MCG⁡(M)→haut⁡(M)\psi\colon\operatorname{MCG}(M) \to\operatorname{haut}(M), and then use our knowledge of MCG⁡(M)\operatorname{MCG}(M), obtained in [12], to compute the group haut⁡(M)\operatorname{haut}(M). The manifold-theoretic perspective adopted in this paper also allows us to describe the generators of haut⁡(M)\operatorname{haut}(M) geometrically.

Before stating the main theorems, we present the construction of some self-homotopy equivalences of MM, which serve as generators of haut⁡(M)\operatorname{haut}(M). Let μ ⁣:M→M∨S6\mu\colon M \to M \vee S^{6} be the pinch map which collapses the boundary of an embedded 66-disk to a point. Let ξ ⁣:S6→M\xi\colon S^{6} \to M be the composite of maps S6→S2S^{6} \to S^{2} and S2→MS^{2} \to M representing a generator of π6(S2)≅Z/12\pi_{6}(S^{2}) \cong\mathbb{Z}/12 and π2(M)≅Z\pi_{2}(M) \cong\mathbb{Z}, respectively. When dd is even, it is shown in [21 Proposition 5.3] that π6(Xd)\pi_{6}(X_{d}) contains a direct summand which is mapped isomorphically to π6(S4)≅Z/2\pi_{6}(S^{4}) \cong\mathbb{Z}/2 under the bundle projection Xd→S4X_{d} \to S^{4}. When MM is spin with both dd and ℓ\ell even, by the homotopy classification, MM is homotopy equivalent to XdX_{d}. Let η ⁣:S6→M\eta\colon S^{6} \to M represent the corresponding non-trivial element in π6(M)\pi_{6}(M). Define homotopy equivalences fξf_{\xi} and fηf_{\eta}

fξ=(id⁡∨ξ)∘μ ⁣:M→M∨S6→M,fη=(id⁡∨η)∘μ ⁣:M→M∨S6→M.f_{\xi}=(\operatorname{id}\vee\xi)\circ\mu\colon M\to M\vee S^{6}\to M,\qquad f_{\eta}=(\operatorname{id}\vee\eta)\circ\mu\colon M\to M\vee S^{6}\to M.

In § 2.2, we define Dehn twists

tϵ, tρ ⁣:M→M,t_{\epsilon},\ t_{\rho}\colon M\to M,

which are diffeomorphisms of MM supported in a tubular neighborhood of an embedded S3S^{3} in MM.

Proposition 1.2. fξf_{\xi} is homotopic to tϵt_{\epsilon}; fηf_{\eta} is not homotopic to a diffeomorphism.

For integers mm and nn, we write (m,n)(m,n) for their greatest common divisor.

Theorem 1.3. Let MM be a simply connected spin homology CP3\mathrm{CP}^{3} with d≠0d\ne0. If dd is odd, then

haut⁡(M)≅Z/(3,d)⋊Z/2,\operatorname{haut}(M)\cong\mathbb{Z}/(3,d)\rtimes\mathbb{Z}/2,

the generator of Z/(3,d)\mathbb{Z}/(3,d) is represented by fξf_{\xi}, the generator of Z/2\mathbb{Z}/2 is represented by an orientation-reversing self-diffeomorphism of MM, and the action of Z/2\mathbb{Z}/2 on Z/(3,d)\mathbb{Z}/(3,d) is by the multiplication of −1-1.

If dd is even, then there is a short exact sequence

1⟶G⋊Z/2⟶haut⁡(M)⟶Z/2⟶1,(3)1\longrightarrow G\rtimes\mathbb{Z}/2\longrightarrow\operatorname{haut}(M)\longrightarrow\mathbb{Z}/2\longrightarrow1, \tag*{(3)}

where

G≅{Z/2×Z/(6,d/2)if d≢0(mod8) and l≢0(mod2),Z/2×Z/(12,d/2)otherwise.G \cong\begin{cases} \mathbb{Z}/2 \times\mathbb{Z}/(6,d/2) & \text{if } d \not\equiv0 \pmod{8}\text{ and } l \not\equiv0 \pmod{2},\\ \mathbb{Z}/2 \times\mathbb{Z}/(12,d/2) & \text{otherwise.} \end{cases}

In the group GG, the generator of Z/2\mathbb{Z}/2 is represented by tρt_{\rho}, the generator of the other direct summand is represented by fξf_{\xi}. The generator of factor Z/2\mathbb{Z}/2 in the semi-direct product is represented by an orientation-reversing smooth conjugation on MM. The action of Z/2\mathbb{Z}/2 on GG is by the multiplication by −1-1. If ll is even, then the exact sequence (3) splits, with a splitting given by fηf_{\eta}.

Remark 1.4. Together with Sawashita’s computation of haut⁡(S2×S4)\operatorname{haut}(S^{2} \times S^{4}), the above theorem provides a description of haut⁡(M)\operatorname{haut}(M) for all simply connected spin homology CP3\mathrm{CP}^{3}’s, except for the unique homotopy type with d=0d = 0 and ll odd (see Theorem 1.1). In the remaining case, we show that haut⁡(M)\operatorname{haut}(M) contains a subgroup of index 2 or 4 isomorphic to (Z/2×Z/6)⋊Z/2(\mathbb{Z}/2 \times\mathbb{Z}/6) \rtimes\mathbb{Z}/2; see Theorem 4.1 (5).

As a corollary, the exact sequence (1) extends to a split short exact sequence.

Corollary 1.5. Let XdX_d be the total space of the S2S^{2}-bundle over S4S^{4} with characteristic class d∈π3SO(3)d \in\pi_{3}SO(3). Assume that d≠0d \ne0. Then there is a split short exact sequence

0→π6(Xd)→haut⁡(Xd)→Gd→1.0 \to\pi_{6}(X_d) \to\operatorname{haut}(X_d) \to G_d \to1.

From this perspective, the generators of the corresponding groups are described as follows: when dd is odd, a generator of π6(Xd)\pi_{6}(X_d) is represented by the Dehn twist tϵt_{\epsilon}, and a splitting of Gd≅Z/2G_d \cong\mathbb{Z}/2 is given by the fiberwise antipodal map faf_a. When dd is even, a generator of the direct summand Z/(12,d/2)<π6(Xd)\mathbb{Z}/(12,d/2) < \pi_{6}(X_d) is represented by the Dehn twist tϵt_{\epsilon}, while a generator of the direct summand Z/2<π6(Xd)\mathbb{Z}/2 < \pi_{6}(X_d) is given by fηf_{\eta}. A splitting of Gd≅Z/2×Z/2G_d \cong\mathbb{Z}/2 \times\mathbb{Z}/2 is provided by faf_a and the Dehn twist tρt_{\rho}. The action of GdG_d on π6(Xd)\pi_{6}(X_d) is given by

fa⋅tϵ=−tϵ,tρ⋅tϵ=tϵ,fa⋅fη=−fa∗(η),tρ⋅fη=tρ∗(η),f_a \cdot t_{\epsilon} = -t_{\epsilon},\quad t_{\rho} \cdot t_{\epsilon} = t_{\epsilon},\quad f_a \cdot f_{\eta} = -f_{a*}(\eta),\quad t_{\rho} \cdot f_{\eta} = t_{\rho*}(\eta),

where fa∗f_{a*} and tρ∗t_{\rho*} stand for the induced isomorphism on π6(Xd)\pi_{6}(X_d).

Theorem 1.6. Let MM be a simply connected non-spin homology CP3\mathrm{CP}^{3}. If d≠0d \ne0, then there are short exact sequences

1→H→haut⁡(M)→Z/2→11 \to H \to\operatorname{haut}(M) \to\mathbb{Z}/2 \to1
1→G→H→Z/2→1.1 \to G \to H \to\mathbb{Z}/2 \to1.

The group GG is isomorphic to Z/4×Z/(6,d/2)\mathbb{Z}/4 \times\mathbb{Z}/(6,d/2), the action of Z/2\mathbb{Z}/2 on GG is given by multiplication by −1-1.

If d=0d = 0, then there is a short exact sequence

0→(Z/4×Z/6)⋊(Z/2×Z/2)→haut⁡(M)→Z/2→1.0 \to(\mathbb{Z}/4 \times\mathbb{Z}/6) \rtimes(\mathbb{Z}/2 \times\mathbb{Z}/2) \to\operatorname{haut}(M) \to\mathbb{Z}/2 \to1.

In the semi-direct product, the actions of (1,0)(1,0) and (0,1)(0,1) on Z/4×Z/6\mathbb{Z}/4 \times\mathbb{Z}/6 are both given by multiplication by −1-1.

The methods used in this paper are primarily drawn from manifold topology. In §2, we first recall the results on the Torelli group T(M)T(M) from [12], and then extend them to the full mapping class group MCG⁡(M)\operatorname{MCG}(M). In § 3, we study the forgetful homomorphism ψ0 ⁣:T(M)→haut⁡0(M)\psi_0\colon T(M) \to\operatorname{haut}_0(M) using the Browder–Novikov–Sullivan–Wall surgery exact sequence. Finally, in §4, we state and prove Theorem 4.1 and 4.2.

Acknowledgement. The second author is partially supported by NSFC 12471069.

The full mapping class group MCG⁡(M)\operatorname{MCG}(M)

The Torelli group T(M)\mathcal{T}(M)

The Torelli group T(M)\mathcal{T}(M). The Torelli group T(M)\mathcal{T}(M) is an important subgroup of the mapping class group MCG⁡(M)\operatorname{MCG}(M). It consists of isotopy classes of diffeomorphisms of MM that induce the identity on H∗(M)H_{*}(M). In [12], the Torelli groups of a more general class of 6-manifolds were computed, namely, simply connected 6-manifolds whose homology groups are isomorphic to H∗(CP3#r(S3×S3))H_{*}(\mathbb{C}P^{3}\#_{r}(S^{3}\times S^{3})). In this subsection, we review this computation in the case of simply connected homology CP3\mathbb{C}P^{3}’s. For further details, see [12 §6–8, §10].

Let MM be a simply connected homology CP3\mathbb{C}P^{3}, D⊂MD\subset M be a fixed embedded 6-disk. Let T(M,D)\mathcal{T}(M,D) be the relative Torelli group, i.e., the group of isotopy classes of diffeomorphisms f:M→Mf:M\to M such that f∣D=id⁡f|_{D}=\operatorname{id} and f∗=id⁡f_{*}=\operatorname{id} on H∗(M)H_{*}(M). It is shown in [12 Theorem 9.4, §10.3] that the homomorphism T(M,D)→T(M)\mathcal{T}(M,D)\to\mathcal{T}(M) is an isomorphism. Therefore we will identify these two groups in the sequel. There is an invariant S:T(M,D)→Z3/L(M)S:\mathcal{T}(M,D)\to\mathbb{Z}^{3}/L(M) constructed from the mapping torus Mf=M×[0,1]/(x,0)∼(f(x),1)M_{f}=M\times[0,1]/(x,0)\sim(f(x),1), which is shown to be an isomorphism.

The definitions of the invariant are slightly different for spin and nonspin manifolds. We start with spin manifolds. Let B=CP∞×K(Z,4)×BSpin⁡B=\mathbb{C}P^{\infty}\times K(\mathbb{Z},4)\times B\operatorname{Spin}, p:B→BSpin⁡→BSOp:B\to B\operatorname{Spin}\to BSO be the canonical projection. Then the normal Gauss map ν:Mf→BSO\nu:M_{f}\to BSO has a lifting g:Mf→Bg:M_{f}\to B constructed as follows. Fix an orientation of MM and a generator x∈H2(M)=H2(Mf)x\in H^{2}(M)=H^{2}(M_{f}), take a map Mf→CP∞M_{f}\to\mathbb{C}P^{\infty} representing xx. For d≠0d\ne0, let y=x2/dy=x^{2}/d be the generator of H4(M)H^{4}(M); for d=0d=0, l≠0l\ne0, let y=p1(M)/24l∈H4(M)y=p_{1}(M)/24l\in H^{4}(M); for d=l=0d=l=0, take y∈H4(M)y\in H^{4}(M) an arbitrary generator. Since ff induces the identity on H∗(M)H_{*}(M), the Wang sequence implies that H4(Mf)→H4(M)H^{4}(M_{f})\to H^{4}(M) is an isomorphism. Let y′∈H4(Mf)y'\in H^{4}(M_{f}) be the preimage of yy, then x2=dy′x^{2}=dy' and p1(Mf)=4(d+6l)y′p_{1}(M_{f})=4(d+6l)y'. Take a map Mf→K(Z,4)M_{f}\to K(\mathbb{Z},4) representing y′y'. Finally choose the spin structure on the stable normal bundle of MfM_{f}, which restricts to the trivial spin structure on the preferred embedding D×S1⊂MfD\times S^{1}\subset M_{f}. This induces a map Mf→BSpin⁡M_{f}\to B\operatorname{Spin}. Putting all these maps together we get g:Mf→Bg:M_{f}\to B which is a lift of the normal Gauss map Mf→BSOM_{f}\to BSO. Note that all maps involved are unique up to homotopy in the appropriate sense.

The pair (Mf,g)(M_{f},g) represents a bordism class in the BB-bordism group Ω7(B,p)=Ω7Spin⁡(CP∞×K(Z,4))\Omega_{7}(B,p)=\Omega_{7}^{\operatorname{Spin}}(\mathbb{C}P^{\infty}\times K(\mathbb{Z},4)). By a standard computation with the Atiyah-Hirzebruch spectral sequence one shows that this group is zero ([11 Theorem 6]). Therefore there exists an 8-dimensional normal BB-bordism G:W→BG:W\to B such that ∂(W,G)=(Mf,g)\partial(W,G)=(M_{f},g). After taking connected sum with quaternionic projective planes HP2\mathbb{H}P^{2}’s (or HP2\mathbb{H}P^{2}’s with opposite orientation) we may assume that the signature of WW is zero.

Let x‾∈H2(B)\overline{x}\in H^{2}(B) be the pull-back of the canonical generator of H2(CP∞)H^{2}(\mathbb{C}P^{\infty}), y‾∈H4(B)\overline{y}\in H^{4}(B) be the pull-back of the canonical generator of H4(K(Z,4))H^{4}(K(\mathbb{Z},4)) and p12∈H4(B)\frac{p_{1}}{2}\in H^{4}(B) be the pull-back of the universal spin Pontrjagin class in H4(BSpin⁡)H^{4}(B\operatorname{Spin}). Let

α=G∗(x‾2−dy‾),β=G∗(p12+2(d+6l)y‾)∈H4(W).\alpha=G^{*}(\overline{x}^{2}-d\overline{y}),\qquad\beta=G^{*}\left(\frac{p_{1}}{2}+2(d+6l)\overline{y}\right)\in H^{4}(W).

Then by construction, the restriction of α\alpha and β\beta on ∂W=Mf\partial W=M_{f} are zero, so we have preimages α‾,β‾∈H4(W,∂W)\overline{\alpha},\overline{\beta}\in H^{4}(W,\partial W) of α\alpha and β\beta respectively. Now define characteristic numbers

s1=⟨β‾2,[W,∂W]⟩,s2=⟨α‾2+α‾ β‾,[W,∂W]⟩,s3=⟨α‾ β‾,[W,∂W]⟩.\begin{aligned} s_{1}&=\langle\overline{\beta}^{2},[W,\partial W]\rangle,\\ s_{2}&=\langle\overline{\alpha}^{2}+\overline{\alpha}\,\overline{\beta},[W,\partial W]\rangle,\\ s_{3}&=\langle\overline{\alpha}\,\overline{\beta},[W,\partial W]\rangle. \end{aligned}

By comparing T(M,D)\mathcal{T}(M,D) and the mapping class group of (S2×D4,∂)(S^2 \times D^4,\partial), and taking consideration of the choices of the BB-bordism WW one has a homomorphism

S=[(s1/8,s2/2,s3)]:T(M,D)→Z3/L(M),S=[(s_1/8,s_2/2,s_3)]:\mathcal{T}(M,D)\to\mathbb{Z}^3/L(M),

where L(M)L(M) is the lattice generated by the characteristic numbers of elements in the subgroup of the 8-dimensional BB-bordism group Ω8(B)=Ω8spin(CP∞×K(Z,4))\Omega_8(B)=\Omega_8^{\mathrm{spin}}(\mathbb{CP}^{\infty}\times K(\mathbb{Z},4)), consisting of manifolds with zero signature. Using the Atiyah-Hirzebruch spectral sequence one can show that this group is isomorphic to Z6\mathbb{Z}^6 with explicit generators e1,⋯ ,e6e_1,\cdots,e_6 [11 Theorem 10]. The characteristic numbers of these generators are given in the following table, where k=(d+6l)k=(d+6l),

The lattice L(M)⊂Z3L(M)\subset\mathbb{Z}^3 is generated by column vectors. In [12 Theorem 6.1] it is shown that SS is an isomorphism, and the groups T(M)\mathcal{T}(M) are given there. In particular, there is a surjective homomorphism Z/28×Z/12×Z/2→T(M)\mathbb{Z}/28\times\mathbb{Z}/12\times\mathbb{Z}/2\to\mathcal{T}(M).

Remark 2.1. In the definition of the SS-invariant, we fix an orientation of MM and a generator x∈H2(M)x\in H^2(M). From the definition it is ready to see that if we reverse the orientation then the invariant changes by a sign, while if we use the generator −x-x then the invariant remains unchanged.

A diffeomorphism f:M→Mf:M\to M is disk supported if f=idf=\mathrm{id} outside an embedded closed 6-dimensional disk D⊂MD\subset M. Isotopy classes of disk supported diffeomorphisms form a subgroup K⊂T(M)K\subset\mathcal{T}(M). The quotient group T(M)/K\mathcal{T}(M)/K will be relevant in the computation of haut⁡(M)\operatorname{haut}(M) (see Proposition 3.2). Let MCG⁡(D6,∂)\operatorname{MCG}(D^6,\partial) be the mapping classes group of the 6-disk relative to boundary. Fix an embedding D6⊂MD^6\subset M, then there is a homomorphism MCG⁡(D6,∂)→T(M)\operatorname{MCG}(D^6,\partial)\to\mathcal{T}(M), extending a diffeomorphism of D6D^6 by the identity outside the embedded D6D^6 to a diffeomorphism of MM. By the disk theorem, the image of this isomorphism is exactly the subgroup KK. The group MCG⁡(D6,∂)\operatorname{MCG}(D^6,\partial) is isomorphic to the Kervaire-Milnor group Θ7\Theta_7 of 7-dimensional homotopy spheres, which is isomorphic to Z/28\mathbb{Z}/28, with a preferred generator 11 represented by the boundary of the E8E_8-plumbing. The SS-invariant of the image of 1∈Z/281\in\mathbb{Z}/28 in T(M)\mathcal{T}(M) equals [(1,0,0)][(1,0,0)]. Therefore from Table 2.1 we get the quotient group T(M)/K\mathcal{T}(M)/K.

e1e_1e2e_2e3e_3e4e_4e5e_5e6e_6
s1/8s_1/8−28-280000(k2−k)/2(k^2-k)/2k2k^200
s2/2s_2/200121200(d2−d(2k−1))/2(d^2-d(2k-1))/2d2−2dkd^2-2dkk−dk-d
s3s_3000022d(2k−1)d(2k-1)4dk4dk−2k-2k

Table 2.1.

Proposition 2.2. Let MM be a simply connected spin homology CP3\mathbb{CP}^3.

(1) If dd is odd,

T(M)/K≅{Z/3if d≡0(mod3),0if d≢0(mod3)\mathcal{T}(M)/K\cong \begin{cases} \mathbb{Z}/3 & \text{if } d\equiv0\pmod{3},\\ 0 & \text{if } d\not\equiv0\pmod{3} \end{cases}

(2) if d≡2(mod4)d\equiv2\pmod{4},

T(M)/K≅{Z/2×Z/3if d≡0(mod3),Z/2if d≢0(mod3)\mathcal{T}(M)/K\cong \begin{cases} \mathbb{Z}/2\times\mathbb{Z}/3 & \text{if } d\equiv0\pmod{3},\\ \mathbb{Z}/2 & \text{if } d\not\equiv0\pmod{3} \end{cases}

(3) if d≡4(mod8)d\equiv4\pmod{8},

T(M)/K≅{Z/2×Z/2×Z/3if d≡0(mod3),Z/2×Z/2if d≢0(mod3)\mathcal{T}(M)/K\cong \begin{cases} \mathbb{Z}/2\times\mathbb{Z}/2\times\mathbb{Z}/3 & \text{if } d\equiv0\pmod{3},\\ \mathbb{Z}/2\times\mathbb{Z}/2 & \text{if } d\not\equiv0\pmod{3} \end{cases}

(4) if d≡0(mod8)d \equiv0 \pmod{8}, T(M)/K\mathcal{T}(M)/K is isomorphic to

d≡0(mod3)d \equiv0 \pmod{3}d≢0(mod3)d \not\equiv0 \pmod{3}
ll evenZ/2×Z/4×Z/3\mathbb{Z}/2 \times\mathbb{Z}/4 \times\mathbb{Z}/3Z/2×Z/4\mathbb{Z}/2 \times\mathbb{Z}/4
ll oddZ/2×Z/2×Z/3\mathbb{Z}/2 \times\mathbb{Z}/2 \times\mathbb{Z}/3Z/2×Z/2\mathbb{Z}/2 \times\mathbb{Z}/2

Proof. From Table 2.1 we see that the finite abelian group T(M)/K\mathcal{T}(M)/K has only 2- and 3-torsions, with exponents at most 2 and 1, respectively. To determined the exponents we only need to take tensor product with Z/4\mathbb{Z}/4 or Z/3\mathbb{Z}/3, respectively. Then an elementary calculation shows this proposition. □

When MM is nonspin, we still take B=CP∞×K(Z,4)×BSpin⁡B = \mathbb{C}P^\infty\times K(\mathbb{Z}, 4) \times B\operatorname{Spin}, the fibration pp is

p:B=CP∞×K(Z,4)×BSpin⁡→BSO⁡×BSO⁡⟶⊕BSO⁡,p: B = \mathbb{C}P^\infty\times K(\mathbb{Z}, 4) \times B\operatorname{Spin} \to B\operatorname{SO} \times B\operatorname{SO} \overset{\oplus}{\longrightarrow} B\operatorname{SO},

where CP∞→BSO⁡\mathbb{C}P^\infty\to B\operatorname{SO} is the classifying map of the canonical complex line bundle η\eta over CP∞\mathbb{C}P^\infty and K(Z,4)×BSpin⁡→BSpin⁡→BSO⁡K(\mathbb{Z}, 4) \times B\operatorname{Spin} \to B\operatorname{Spin} \to B\operatorname{SO} is the composition of natural projections. Then similar to the spin case, for a diffeomorphism ff, we have a lifting g:Mf→Bg:M_f \to B of the normal Gauss map of MfM_f. The corresponding bordism group Ω∗(B,p)\Omega_*(B,p) is the twisted spin cobordism group Ω∗spin⁡(CP∞×K(Z,4);η)\Omega_*^{\operatorname{spin}}(\mathbb{C}P^\infty\times K(\mathbb{Z},4);\eta). This group vanishes ([11 Theorem 6]). Let xˉ∈H2(CP∞)\bar{x} \in H^2(\mathbb{C}P^\infty), yˉ∈H4(K(Z,4))\bar{y} \in H^4(K(\mathbb{Z},4)) and p12∈H4(BSpin⁡)\frac{p_1}{2} \in H^4(B\operatorname{Spin}) be as in the spin case. Let G:W8→BG:W^8 \to B be a coboundary of g:Mf→Bg:M_f \to B with signature zero, let α=G∗(xˉ2−dyˉ)\alpha= G^*(\bar{x}^2-d\bar{y}), β=G∗(p12+(d+24l)yˉ)∈H4(W)\beta= G^*(\frac{p_1}{2}+(d+24l)\bar{y}) \in H^4(W), then they have preimages αˉ,βˉ∈H4(W,∂W)\bar{\alpha},\bar{\beta} \in H^4(W,\partial W) respectively. Define characteristic numbers

σ1=⟨αˉ∪αˉ,[W,∂W]⟩σ2=⟨αˉ∪βˉ,[W,∂W]⟩σ3=⟨βˉ∪βˉ,[W,∂W]⟩.\begin{aligned} \sigma_1 &= \langle\bar{\alpha}\cup\bar{\alpha},[W,\partial W]\rangle\\ \sigma_2 &= \langle\bar{\alpha}\cup\bar{\beta},[W,\partial W]\rangle\\ \sigma_3 &= \langle\bar{\beta}\cup\bar{\beta},[W,\partial W]\rangle. \end{aligned}

Let λ=σ1+3σ2+σ3\lambda= \sigma_1+3\sigma_2+\sigma_3, μ=(6σ1+10σ2+3σ3)/4\mu=(6\sigma_1+10\sigma_2+3\sigma_3)/4, then the SS-invariant

S=[(μ,λ)]:T(M,D)→Z2/L(M)S=[(\mu,\lambda)]:\mathcal{T}(M,D)\to\mathbb{Z}^2/L(M)

is an isomorphism, where L(M)L(M) is the lattice generated by the column vectors of the following table.

μ\mu6⋅286\cdot2800−δ2−48δl+2δ+432l2−36l-\delta^2-48\delta l+2\delta+432l^2-36l−2δ2−96δl−864l2-2\delta^2-96\delta l-864l^23(δ−4l)3(\delta-4l)
λ\lambda00−4-4−4δ2−48δl+2δ+576l2−48l-4\delta^2-48\delta l+2\delta+576l^2-48l−2(4δ2+48δl−576l2)-2(4\delta^2+48\delta l-576l^2)6(δ−4l)6(\delta-4l)

The image of the preferred generator of MCG⁡(D6,∂)≅Z/28\operatorname{MCG}(D^6,\partial) \cong\mathbb{Z}/28 in T(M)\mathcal{T}(M) has SS-invariant [(−6,−12)][(-6,-12)]. From this one may compute the quotient group T(M)/K\mathcal{T}(M)/K.

Proposition 2.3. Let MM be a simply connected nonspin homology CP3\mathbb{C}P^3, T(M)/K\mathcal{T}(M)/K is given in the following table.

dd12∣d12\mid d3∣d,4∤d3\mid d,4\nmid d3∤d,4∣d3\nmid d,4\mid d3∤d,4∤d3\nmid d,4\nmid d
T(M)/K\mathcal{T}(M)/KZ/6×Z/4\mathbb{Z}/6\times\mathbb{Z}/4Z/3×Z/4\mathbb{Z}/3\times\mathbb{Z}/4Z/2×Z/4\mathbb{Z}/2\times\mathbb{Z}/4Z/4\mathbb{Z}/4

Dehn twists

Let i:S2→Mi:S^2\to M represent a generator of π2(M)≅Z\pi_2(M)\cong\mathbb{Z}, φ:S3→S2\varphi:S^3\to S^2 be the Hopf fiber map. Then i∘φ:S3→Mi\circ\varphi:S^3\to M represents a generator of π3(M)≅Z\pi_3(M)\cong\mathbb{Z}. Fix a thickened embedded 3-sphere S3×D3⊂MS^3\times D^3\subset M, such that S3×{0}S^3\times\{0\} represents this generator of π3(M)\pi_3(M). For any α∈π3SO⁡(4)\alpha\in\pi_3\operatorname{SO}(4), choose a representing smooth map f:D3→SO⁡(4)f:D^3\to\operatorname{SO}(4), which maps an open neighborhood of ∂D3\partial D^3 to II. The Dehn twist tα:M→Mt_\alpha:M\to M is defined as the diffeomorphism which maps the complement of S3×D3S^3\times D^3 by the identity and (x,y)∈S3×D3(x,y)\in S^3\times D^3 to (f(y)x,y)(f(y)x,y).

From the exact sequence of homotopy groups of fiber bundle

SO(3)→SO(4)→S3SO(3) \to SO(4) \to S^{3}

it can be shown that the group π3(SO(4))\pi_{3}(SO(4)) is isomorphic to Z⊕Z\mathbb{Z} \oplus\mathbb{Z}, with generators represented by ϵ:S3→SO(3)→SO(4)\epsilon:S^{3} \to SO(3) \to SO(4), which is the composite of the projection S3→SO(3)S^{3} \to SO(3) with the inclusion, and ρ:S3→SO(4)\rho:S^{3} \to SO(4), which is the inclusion of the unit quaternions S3S^{3} to SO(4)SO(4). The corresponding Dehn twists are denoted by tϵt_{\epsilon} and tρt_{\rho}, respectively.

Lemma 2.4. (1) The group T(M)/K\mathcal{T}(M)/K is generated by tϵt_{\epsilon} and tρt_{\rho}; the order of tρt_{\rho} is at most 2.

(2) Let J:π3SO(3)→π6(S3)J:\pi_{3}SO(3) \to\pi_{6}(S^{3}) be the JJ-homomorphism, ξ=i∘φ∘J(ϵ):S6→M\xi=i\circ\varphi\circ J(\epsilon):S^{6}\to M. Then the homotopy equivalence fξ=(id⁡∨ξ)∘μ:M→M∨S6→Mf_{\xi}=(\operatorname{id}\vee\xi)\circ\mu:M\to M\vee S^{6}\to M is homotopic to tϵt_{\epsilon}.

Proof. The statements in (1) is shown in [12 §7]. To show (2), consider the natural inclusion map SO(3)→G+(S3)SO(3)\to G^{+}(S^{3}), where G+(3)G^{+}(3) is the space of base point and orientation preserving homotopy equivalences of S3S^{3}. The induced homomorphism is the JJ-homomorphism

J:π3SO(3)→π3G+(3)≅π6(S3).J:\pi_{3}SO(3)\to\pi_{3}G^{+}(3)\cong\pi_{6}(S^{3}).

Therefore regarded as a diffeomorphism of S3×D3S^{3}\times D^{3}, the Dehn twist tϵt_{\epsilon} is homotopic to

(id⁡∨J(ϵ))∘μ:S3×D3→(S3×D3)∨S6→S3×D3,(\operatorname{id}\vee J(\epsilon))\circ\mu:S^{3}\times D^{3}\to(S^{3}\times D^{3})\vee S^{6}\to S^{3}\times D^{3},

where the map on S6S^{6} is the composition of J(ϵ):S6→S3×{0}J(\epsilon):S^{6}\to S^{3}\times\{0\} and the inclusion S3×{0}→S3×D3S^{3}\times\{0\}\to S^{3}\times D^{3}. As a diffeomorphism of MM, the Dehn twist tϵt_{\epsilon} is homotopic to (id⁡∨ξ)∘μ(\operatorname{id}\vee\xi)\circ\mu.

The full mapping class group MCG(M)

In this subsection we compute the full mapping class group MCG⁡(M)\operatorname{MCG}(M), in which orientation reversing diffeomorphisms are contained as well.

Theorem 2.5. Let MM be a simply connected homology CP3\mathbb{CP}^{3} which is not diffeomorphic to S2×S4S^{2}\times S^{4} or S4×~S2S^{4}\widetilde{\times}S^{2}. Then there is a short exact sequence

1→T(M)→MCG⁡(M)→Z/2→1,(4)1\to\mathcal{T}(M)\to\operatorname{MCG}(M)\to\mathbb{Z}/2\to1, \tag*{(4)}

where the action of Z/2\mathbb{Z}/2 on T(M)\mathcal{T}(M) is by −1-1. This short exact sequence splits when MM is spin, hence MCG⁡(M)=T(M)⋊Z/2\operatorname{MCG}(M)=\mathcal{T}(M)\rtimes\mathbb{Z}/2.

Proof. Denote by MCG⁡+(M)\operatorname{MCG}^{+}(M) the oriented mapping class group of MM. Then there is an exact sequence

1→MCG⁡+(M)→MCG⁡(M)→Z/2.1\to\operatorname{MCG}^{+}(M)\to\operatorname{MCG}(M)\to\mathbb{Z}/2.

It is assumed that MM is not diffeomorphic to S2×S4S^{2}\times S^{4} or S4×~S2S^{4}\widetilde{\times}S^{2}. Therefore either d≠0d\ne0 or l≠0l\ne0. If d≠0d\ne0, then clearly T(M)=MCG⁡+(M)\mathcal{T}(M)=\operatorname{MCG}^{+}(M). If d=0d=0 but l≠0l\ne0, then p1(M)≠0p_{1}(M)\ne0. In this case since a diffeomorphism preserves p1(M)p_{1}(M), it induces the identity on H4(M)H^{4}(M), therefore T(M)=MCG⁡+(M)\mathcal{T}(M)=\operatorname{MCG}^{+}(M). Moreover, by the classification theorem (Theorem 1.1), there exists an orientation-reversing self-diffeomorphism of MM, therefore we have a short exact sequence

1→T(M)→MCG⁡(M)→Z/2→1.1\to\mathcal{T}(M)\to\operatorname{MCG}(M)\to\mathbb{Z}/2\to1.

Since T(M)\mathcal{T}(M) is an abelian group, there is an action of Z/2\mathbb{Z}/2 on T(M)\mathcal{T}(M), given by conjugation: let ff be an orientation-reversing diffeomorphism, [g]∈T(M)[g]\in\mathcal{T}(M), then the action of 1∈Z/21\in\mathbb{Z}/2 on [g][g] is [f−1∘g∘f]∈T(M)[f^{-1}\circ g\circ f]\in\mathcal{T}(M). Observe that there is an orientation-preserving diffeomorphism between the mapping tori

Mf−1∘g∘f→≅Mg−1,[(x,t)]↦[(f(x),1−t)]M_{f^{-1}\circ g\circ f}\xrightarrow{\cong}M_{g^{-1}},\qquad[(x,t)]\mapsto[(f(x),1-t)]

This implies that f−1∘g∘ff^{-1}\circ g\circ f and g−1g^{-1} have the same SS-invariants, therefore [f−1∘g∘f]=[g−1]∈T(M)[f^{-1}\circ g\circ f]=[g^{-1}]\in\mathcal{T}(M). This shows that the action of Z/2\mathbb{Z}/2 on T(M)\mathcal{T}(M) is by −1-1.

Next we show that when MM is spin the exact sequence is split. If ll is even, it’s known that there exists a free involution on MM (see [14 Theorem 1.2] and the discussion there). By the Lefschetz fixed-point theorem, this free involution is orientation-reversing. This gives a splitting. If dd is even, then by [15 Corollary 1.5], there exists a conjugation on MM, which is a special type of involution. Such a conjugation has a 3-dimensional fixed-point set, hence must be orientation-reversing. This gives a splitting. If both dd and ll are odd, then from the tables in [12 Theorem 6.1] one sees that the Torelli group T(M)\mathcal{T}(M) has odd order. Therefore H2(Z/2;T(M))=0H^2(\mathbb{Z}/2;\mathcal{T}(M))=0. This implies that the sequence (4) splits. □

Theorem 2.6. Let M=S2×S4M=S^2\times S^4 or S4×~S2S^4\widetilde{\times}S^2, then

MCG⁡(M)=T(M)⋊(Z/2×Z/2)\operatorname{MCG}(M)=\mathcal{T}(M)\rtimes(\mathbb{Z}/2\times\mathbb{Z}/2)

where the actions of the standard generators of Z/2×Z/2\mathbb{Z}/2\times\mathbb{Z}/2 on T(M)\mathcal{T}(M) are by −1-1.

Proof. If M=S2×S4M=S^2\times S^4, then clearly we have a short exact sequence

1⟶T(M)⟶MCG⁡(M)⟶Z/2×Z/2⟶11\longrightarrow\mathcal{T}(M)\longrightarrow\operatorname{MCG}(M)\longrightarrow\mathbb{Z}/2\times\mathbb{Z}/2\longrightarrow1

where the homomorphism MCG⁡(M)→Z/2×Z/2\operatorname{MCG}(M)\to\mathbb{Z}/2\times\mathbb{Z}/2 is given by the actions on H2(M)H_2(M) and H4(M)H_4(M). Apparently the elements (1,0)(1,0) and (0,1)∈Z/2×Z/2(0,1)\in\mathbb{Z}/2\times\mathbb{Z}/2 have liftings (−id⁡,id⁡):S2×S4→S2×S4(-\operatorname{id},\operatorname{id}):S^2\times S^4\to S^2\times S^4 and (id⁡,−id⁡):S2×S4→S2×S4(\operatorname{id},-\operatorname{id}):S^2\times S^4\to S^2\times S^4 respectively. Since these are orientation reversing diffeomorphisms, as discussed in the proof of last theorem, their actions on T(M)\mathcal{T}(M) are by −1-1.

When M=S4×~S2M=S^4\widetilde{\times}S^2, the non-trivial linear S4S^4-bundle over S2S^2, we construct two smooth involutions as follows. The first involution f1f_1 is the fiberwise antipodal map, then f1f_1 induces the identity on H2H_2 and −id⁡-\operatorname{id} on H4H_4. To construct the second involution f2f_2, we use the decomposition S4×~S2=(S4×D2)1∪α(S4×D2)2S^4\widetilde{\times}S^2=(S^4\times D^2)_1\cup_\alpha(S^4\times D^2)_2, with α\alpha a map S1→SO(5)S^1\to SO(5) representing the nontrivial element in π1SO(5)≅Z/2\pi_1SO(5)\cong\mathbb{Z}/2. Consider the space

SO(5)Z/2={A∈SO(5)∣At=A}.SO(5)^{\mathbb{Z}/2}=\{A\in SO(5)\mid A^t=A\}.

Such matrices have eigenvalues ±1\pm1. It’s not hard to see that this space is a union of Grassmannians of even dimensional linear subspaces in R5\mathbb{R}^5. Especially we have an inclusion G4(R5)≅RP4⊂SO(5)Z/2⊂SO(5)G_4(\mathbb{R}^5)\cong\mathbb{R}P^4\subset SO(5)^{\mathbb{Z}/2}\subset SO(5), inducing isomorphism on π1\pi_1. We may choose α\alpha to be a loop in SO(5)Z/2SO(5)^{\mathbb{Z}/2}. Then we may define an involution f2:S4×~S2→S4×~S2f_2:S^4\widetilde{\times}S^2\to S^4\widetilde{\times}S^2, by interchanging the two copies S4×D2S^4\times D^2: (x,y)1↦(x,y)2(x,y)_1\mapsto(x,y)_2, (x,y)2↦(x,y)1(x,y)_2\mapsto(x,y)_1. Since for any y∈S1y\in S^1, α(y)2=I\alpha(y)^2=I, this is a well-defined diffeomorphism. Clearly f2f_2 induces −id⁡-\operatorname{id} on H2H_2 and the identity on H4H_4, and f1f_1 and f2f_2 commute. Therefore we have MCG⁡(S4×~S2)=T(S4×~S2)⋊(Z/2×Z/2)\operatorname{MCG}(S^4\widetilde{\times}S^2)=\mathcal{T}(S^4\widetilde{\times}S^2)\rtimes(\mathbb{Z}/2\times\mathbb{Z}/2), with the actions of (1,0)(1,0) and (0,1)(0,1) are by −1-1. □

The forgetful homomorphism

The full mapping class group MCG⁡(M)\operatorname{MCG}(M) and the group of self-equivalences haut⁡(M)\operatorname{haut}(M) are related by the forgetful homomorphism ψ:MCG⁡(M)→haut⁡(M)\psi:\operatorname{MCG}(M)\to\operatorname{haut}(M). Let haut⁡0(M)\operatorname{haut}_0(M) be the subgroup of haut⁡(M)\operatorname{haut}(M) consisting of homotopy equivalences which act trivially on H∗(M)H_*(M), then ψ0:T(M)→haut⁡0(M)\psi_0:\mathcal{T}(M)\to\operatorname{haut}_0(M). In this section we analyze this homomorphism by surgery theory.

Proposition 3.1. Let MM be a simply connected homology CP3\mathbb{CP}^3, ψ0:T(M)→haut⁡0(M)\psi_0:\mathcal{T}(M)\to\operatorname{haut}_0(M) be the forgetful homomorphism. When dd is odd, ψ0\psi_0 is surjective; when dd is even, the image of ψ0\psi_0 is an index 22 subgroup of haut⁡0(M)\operatorname{haut}_0(M).

Proof. Denote by S(M)\mathcal{S}(M) the smooth structure set of MM. Elements of S(M)\mathcal{S}(M) are equivalence classes of homotopy equivalences f:X→Mf:X\to M from a closed smooth manifold XX to MM. Two pairs (X1,f1)(X_1,f_1) and (X2,f2)(X_2,f_2) are equivalent if there exists a diffeomorphism ϕ:X1→X2\phi:X_1\to X_2 such that f1f_1 is homotopic to f2∘ϕf_2\circ\phi. There is a map α:haut⁡0(M)→S(M)\alpha:\operatorname{haut}_0(M)\to\mathcal{S}(M) sending a homotopy equivalence f:M→Mf:M\to M to [M,f]∈S(M)[M,f]\in\mathcal{S}(M). By definition, α(f)=α(g)\alpha(f)=\alpha(g) if and only if ff and gg are in the same coset of Im⁡ψ0\operatorname{Im}\psi_0. Therefore we have an injective map between sets

α‾:haut⁡0(M)/Im⁡ψ0→S(M)\overline{\alpha}:\operatorname{haut}_0(M)/\operatorname{Im}\psi_0\to\mathcal{S}(M)

The image of this map can be analyzed via the surgery exact sequence ([10 Theorem 14.8])

L7(Z)→S(M)→ηN(M)→θL6(Z).L_7(\mathbb{Z})\to\mathcal{S}(M)\xrightarrow{\eta}\mathcal{N}(M)\xrightarrow{\theta}L_6(\mathbb{Z}).

The surgery obstruction group L7(Z)L_7(\mathbb{Z}) is zero, hence η\eta is injective. Elements in N(M)\mathcal{N}(M) are represented by degree 11 normal maps

νN→fˉE↓↓N→fM\begin{CD} \nu N @>{\bar{f}}>> E \\ @VVV @VVV \\ N @>{f}>> M \end{CD}

More precisely, the set N(M)\mathcal{N}(M) can be described as

N(M)=⨁Ω6(M;E)1/Aut⁡(E),\mathcal{N}(M)=\bigoplus\Omega_6(M;E)_1/\operatorname{Aut}(E),

where the direct sum takes over all isomorphism classes of oriented stable vector bundles EE over MM, for the details see [10 §14.5].

We first analyze the group Aut⁡(E)\operatorname{Aut}(E), which is the group of homotopy classes of bundle isomorphisms of EE. Fix a kk-dimensional representative ξ\xi of EE with k≫0k\gg0, let P→MP\to M be the SO⁡(k)\operatorname{SO}(k)-principal bundle associated to ξ\xi. Then the group Aut⁡(ξ)\operatorname{Aut}(\xi) is identified with π0(G(P))\pi_0(\mathcal{G}(P)), where G(P)\mathcal{G}(P) is the gauge group of the principal bundle PP. It is known that the gauge group G(P)\mathcal{G}(P) is homotopy equivalent to the based loop space ΩMap⁡(M,BSO⁡(k))\Omega\operatorname{Map}(M,B\operatorname{SO}(k)) of the space of continuous maps from MM to BSO⁡(k)B\operatorname{SO}(k), with base point a classifying map φ:M→BSO⁡(k)\varphi:M\to B\operatorname{SO}(k) of ξ\xi (see [1]). Therefore Aut⁡(ξ)\operatorname{Aut}(\xi) is isomorphic to π1(Map⁡(M,BSO⁡(k)),φ)\pi_1(\operatorname{Map}(M,B\operatorname{SO}(k)),\varphi). From the fiber sequence

ΩBSO⁡(k)→Map⁡∗(M,BSO⁡(k))→Map⁡(M,BSO⁡(k))→ev⁡BSO⁡(k)\Omega B\operatorname{SO}(k)\to\operatorname{Map}_*(M,B\operatorname{SO}(k))\to\operatorname{Map}(M,B\operatorname{SO}(k))\xrightarrow{\operatorname{ev}}B\operatorname{SO}(k)

where Map⁡∗\operatorname{Map}_* denotes the space of based continuous maps, and ev⁡\operatorname{ev} is the evaluation map at the base point x0∈Mx_0\in M, one has an exact sequence

π1(Map⁡∗(M,BSO⁡(k)),φ)→π1(Map⁡(M,BSO⁡(k)),φ)→π0SO⁡(k).(5)\pi_1(\operatorname{Map}_*(M,B\operatorname{SO}(k)),\varphi)\to\pi_1(\operatorname{Map}(M,B\operatorname{SO}(k)),\varphi)\to\pi_0\operatorname{SO}(k). \tag*{(5)}

By the adjunction formula, the space Map⁡∗(S1,Map⁡∗(M,BSO(k))φ)\operatorname{Map}_*(S^1,\operatorname{Map}_*(M,BSO(k))_\varphi) is homeomorphic to the space of continuous maps from S1×M/S1×{x0}S^1 \times M/S^1 \times\{x_0\} to BSO(k)BSO(k) whose restriction on {s0}×M\{s_0\} \times M is φ\varphi, where s0∈S1s_0 \in S^1 is the base point. Notice that all the groups

Hi(S1×M/S1×{x0},{s0}×M;πiBSO(k))≅Hi(S1∧M;πiBSO(k))H^i(S^1 \times M/S^1 \times\{x_0\},\{s_0\} \times M;\pi_iBSO(k)) \cong H^i(S^1 \wedge M;\pi_iBSO(k))

vanish. Elementary obstruction theory shows that all such maps are homotopic relative to {s0}×M\{s_0\} \times M. Therefore the exact sequence (5) implies Aut⁡(ξ)=0\operatorname{Aut}(\xi)=0. By passing to the limit one has Aut⁡(E)=0\operatorname{Aut}(E)=0. Hence N(M)=⨁Ω6(M;E)1\mathcal{N}(M)=\bigoplus\Omega_6(M;E)_1.

Next we analyze the map η\eta. The element η([M,f])∈N(M)\eta([M,f]) \in\mathcal{N}(M) is represented by the degree 1 normal map (fˉ,f)(\bar{f},f)

νM→fˉg∗(νM)↓↓M→fM\begin{CD} \nu M @>{\bar{f}}>> g^*(\nu M) \\ @VVV @VVV \\ M @>{f}>> M \end{CD}

where gg is a homotopy inverse of ff. By the minimal cell structure theorem, MM is homotopy equivalent to Pd∪e6P_d \cup e^6, where Pd=S2∪de4P_d=S^2\cup_d e^4, η:S3→S2\eta:S^3\to S^2 the Hopf map. The Puppe sequence of the cofibration sequence S5→Pd→M→S6S^5\to P_d\to M\to S^6, together with the fact π5(BO)=π6(BO)=0\pi_5(BO)=\pi_6(BO)=0 implies that the restriction map [M,BO]→[Pd,BO][M,BO]\to[P_d,BO] is an isomorphism. A further analysis of the Puppe sequence of the cofibration sequence S3→S2→Pd→S4S^3\to S^2\to P_d\to S^4 shows that a stable vector bundle ξ\xi over PdP_d is determined by w2(ξ)w_2(\xi) and p1(ξ)p_1(\xi). Since gg acts trivially on homology, g∗(νM)g^*(\nu M) is isomorphic to νM\nu M. Therefore η([M,f])\eta([M,f]) is in the subset Ω6(M;νM)1\Omega_6(M;\nu M)_1 of N(M)\mathcal{N}(M).

The bordism group Ω6(M;νM)\Omega_6(M;\nu M) is identified by the Thom-Pontrjagin construction with the homotopy group π6(Th⁡(νM))\pi_6(\operatorname{Th}(\nu M)) of the Thom spectrum Th⁡(νM)\operatorname{Th}(\nu M), whose nn-th space is the Thom space of the normal bundle of M⊂R6+nM\subset\mathbb{R}^{6+n}. We compute the homotopy group of the spectrum by the Atiyah-Hirzebruch spectral sequence. The Ep,q2E^2_{p,q}-terms Ep,q2E^2_{p,q} are identified via the Thom isomorphism with Hp(M;πqS)H_p(M;\pi_q^S). The differential d2:E6,12→E4,22d_2:E^2_{6,1}\to E^2_{4,2} is dual to

Sq2+w2(M)∪−:H4(M;Z/2)→H6(M;Z/2);Sq^2+w_2(M)\cup-:H^4(M;\mathbb{Z}/2)\to H^6(M;\mathbb{Z}/2);

d2:E6,02→E4,12d_2:E^2_{6,0}\to E^2_{4,1} is the composition of the mod 2 reduction and the dual of Sq2+w2(M)∪−:H4(M;Z/2)→H6(M;Z/2)Sq^2+w_2(M)\cup-:H^4(M;\mathbb{Z}/2)\to H^6(M;\mathbb{Z}/2) ([18 Lemma]). Using the same idea in the proof of Lemma in [18], it can be shown that the differential d2:E4,22→E2,32d_2:E^2_{4,2}\to E^2_{2,3} is the composition of the dual of Sq2+w2(M)∪−:H2(M;Z/2)→H4(M;Z/2)Sq^2+w_2(M)\cup-:H^2(M;\mathbb{Z}/2)\to H^4(M;\mathbb{Z}/2) and the homomorphism j∗:H2(M;Z/2)→H2(M;Z/24)j_*:H_2(M;\mathbb{Z}/2)\to H_2(M;\mathbb{Z}/24) induced by the inclusion of coefficients. (The differential d2d_2 is a stable homology operation. Such operations are classified by

[HZ/2,Σ2HZ/24]≅H2(HZ/2;Z/24)≅Z/2{j∗∘(Sq2)∗},[H\mathbb{Z}/2,\Sigma^2H\mathbb{Z}/24]\cong H^2(H\mathbb{Z}/2;\mathbb{Z}/24)\cong\mathbb{Z}/2\{j_*\circ(Sq^2)^*\},

and the non-triviality of d2d_2 can be detected by CP2\mathbb{C}P^2, whose stable homotopy groups can be calculated using the cofiber sequence S2→CP2→S4S^2\to\mathbb{C}P^2\to S^4.) A straightforward calculation shows that d2:E6,12→E4,22d_2:E^2_{6,1}\to E^2_{4,2} and d2:E6,02→E4,12d_2:E^2_{6,0}\to E^2_{4,1} are trivial; d2:E4,22→E2,32d_2:E^2_{4,2}\to E^2_{2,3} is nontrivial when dd is odd, and trivial when dd is even.

Therefore, when dd is odd, we get a short exact sequence

0→Ω6fr→Ω6(M;νM)→H6(M)→0.0\to\Omega_6^{\mathrm{fr}}\to\Omega_6(M;\nu M)\to H_6(M)\to0.

The surgery obstruction map θ:Ω6fr→L6(Z)\theta:\Omega^{\mathrm{fr}}_{6}\to L_{6}(\mathbb{Z}) is an isomorphism. The normal map (fˉ,f)(\bar{f},f) has vanishing surgery obstruction and the degree of ff is 1. Therefore the image of η∘αˉ\eta\circ\bar{\alpha} consists of one element. Since η\eta is injective, this shows that haut⁡0(M)/Im⁡ψ0\operatorname{haut}_{0}(M)/\operatorname{Im}\psi_{0} consists of one element. When dd is even, there are short exact sequences

0→G→Ω6(M;νM)→H6(M)→00\to G\to\Omega_{6}(M;\nu M)\to H_{6}(M)\to0
0→Ω6fr→G→Z/2→0.0\to\Omega^{\mathrm{fr}}_{6}\to G\to\mathbb{Z}/2\to0.

Therefore there are two bordism classes of degree one normal maps with trivial surgery obstruction. Such an element is represented by

νN→gˉνM↓↓N→gM\begin{CD} \nu N @>{\bar{g}}>> \nu M \\ @VVV @VVV \\ N @>{g}>> M \end{CD}

with gg a homotopy equivalence. By the classification of homology CP3\mathbb{C}P^{3} (Theorem 1.1), NN is diffeomorphism to MM. Therefore, the image of η∘αˉ\eta\circ\bar{\alpha} consists of two elements. This shows that Im⁡ψ0\operatorname{Im}\psi_{0} is a subgroup of index 2 in haut⁡0(M)\operatorname{haut}_{0}(M). □

Proposition 3.2. The kernel of ψ0\psi_{0} is the subgroup of T(M)\mathcal{T}(M) consisting of disk-supported diffeomorphisms.

Proof. Clearly any disk-supported diffeomorphism is in the kernel of ψ0\psi_{0} by the Alexander’s trick. On the other hand, given a diffeomorphism ff which is homotopic to the identity, let h:M×[0,1]→M×[0,1]h:M\times[0,1]\to M\times[0,1] be a homotopy, with h0=id⁡h_{0}=\operatorname{id} and h1=fh_{1}=f, then [M×[0,1],h][M\times[0,1],h] can be viewed as an element in the relative structure set S∂(M×[0,1])\mathcal{S}_{\partial}(M\times[0,1]). Consider the surgery exact sequence

L8(Z)→S∂(M×[0,1])→[ΣM,G/O]→L7(Z).L_{8}(\mathbb{Z})\to\mathcal{S}_{\partial}(M\times[0,1])\to[\Sigma M,G/O]\to L_{7}(\mathbb{Z}).

The homomorphism L8(Z)→S∂(M×[0,1])L_{8}(\mathbb{Z})\to\mathcal{S}_{\partial}(M\times[0,1]) is given by taking connected sum of some 7-dimensional homotopy sphere Σ\Sigma with the base point [M×[0,1],id⁡][M\times[0,1],\operatorname{id}] in S∂(M×[0,1])\mathcal{S}_{\partial}(M\times[0,1]). By obstruction theory it’s easy to see that [ΣM,G/O]=0[\Sigma M,G/O]=0 (since πi(G/O)=0\pi_{i}(G/O)=0 for i=1,3,5,7i=1,3,5,7). Therefore there is a diffeomorphism g:M×[0,1]→(M×[0,1])#Σg:M\times[0,1]\to(M\times[0,1])\#\Sigma such that g∣M×0=id⁡g|_{M\times0}=\operatorname{id}, g∣M×1=fg|_{M\times1}=f. Composed with the diffeomorphism ϕ:(M×[0,1])#Σ→M×[0,1]\phi:(M\times[0,1])\#\Sigma\to M\times[0,1], where ϕ∣M×0\phi|_{M\times0} is the disk supported diffeomorphism corresponding to Σ\Sigma and ϕ∣M×1=id⁡\phi|_{M\times1}=\operatorname{id}, we have a diffeomorphism k:M×[0,1]→M×[0,1]k:M\times[0,1]\to M\times[0,1] such that k∣M×0k|_{M\times0} is the disk supported diffeomorphism corresponding to Σ\Sigma, and k∣M×1=fk|_{M\times1}=f. This shows that ff is pseudo-isotopic to a disk supported diffeomorphism. By the pseudo-isotopy theorem of Cerf [8], ff is isotopic to a disk supported diffeomorphism.

For a more general discussion of the homomorphism T(M)→haut⁡0(M)\mathcal{T}(M)\to\operatorname{haut}_{0}(M) in higher dimensions see [7 Proposition II.9]. □

Determination of haut⁡(M)\operatorname{haut}(M)

Let ψ:MCG⁡(M)→haut⁡(M)\psi:\operatorname{MCG}(M)\to\operatorname{haut}(M) be the forgetful homomorphism, haut⁡0(M)\operatorname{haut}_{0}(M) be the group of homotopy equivalences which acts trivially on homology.

Theorem 4.1. Let MM be a simply connected spin homology CP3\mathbb{C}P^{3}.

  1. If dd is odd, then ψ\psi is surjective,

haut⁡(M)=haut⁡0(M)×Z/2,haut⁡0(M)≅Z/(3,d),\operatorname{haut}(M)=\operatorname{haut}_0(M)\times\mathbb{Z}/2,\quad\operatorname{haut}_0(M)\cong\mathbb{Z}/(3,d),

the action of Z/2\mathbb{Z}/2 on haut⁡0(M)\operatorname{haut}_0(M) is by −1-1.

  1. If dd is even and d≠0d\ne0, then there is a short exact sequence

1→Im⁡ψ→haut⁡(M)→Z/2→1,1\to\operatorname{Im}\psi\to\operatorname{haut}(M)\to\mathbb{Z}/2\to1,

where Im⁡ψ\operatorname{Im}\psi is isomorphic to G×Z/2G\times\mathbb{Z}/2, and

G≅{Z/2×Z/(6,d/2)if d≠0(mod8) and l≠0(mod2)Z/2×Z/(12,d/2)otherwiseG\cong \begin{cases} \mathbb{Z}/2\times\mathbb{Z}/(6,d/2) & \text{if }d\ne0\pmod{8}\text{ and }l\ne0\pmod{2}\\ \mathbb{Z}/2\times\mathbb{Z}/(12,d/2) & \text{otherwise} \end{cases}

The action of Z/2\mathbb{Z}/2 on GG is by −1-1, and G⊂haut⁡0(M)G\subset\operatorname{haut}_0(M) is a subgroup of index 2.

  1. If d=l=0d=l=0, i.e., M=S2×S4M=S^2\times S^4, then there is a short exact sequence

1→Im⁡ψ→haut⁡(M)→Z/2→1,1\to\operatorname{Im}\psi\to\operatorname{haut}(M)\to\mathbb{Z}/2\to1,

where Im⁡ψ\operatorname{Im}\psi is isomorphic to G×(Z/2×Z/2)G\times(\mathbb{Z}/2\times\mathbb{Z}/2), G≅Z/2×Z/12G\cong\mathbb{Z}/2\times\mathbb{Z}/12, the actions of (1,0)(1,0) and (0,1)(0,1) on GG are by −1-1. G⊂haut⁡0(M)G\subset\operatorname{haut}_0(M) is a subgroup of index 2.

  1. If d=0d=0, ll is even and non-zero, then Im⁡ψ\operatorname{Im}\psi is a subgroup of index 4, isomorphic to G×Z/2G\times\mathbb{Z}/2, G≅Z/2×Z/12G\cong\mathbb{Z}/2\times\mathbb{Z}/12, the action of Z/2\mathbb{Z}/2 on GG is by −1-1. G⊂haut⁡0(M)G\subset\operatorname{haut}_0(M) is a subgroup of index 2.

  1. If d=0d=0 and ll is odd, then Im⁡ψ\operatorname{Im}\psi is a subgroup of index 2 or 4, isomorphic to (Z/2×Z/6)×Z/2(\mathbb{Z}/2\times\mathbb{Z}/6)\times\mathbb{Z}/2, the action of Z/2\mathbb{Z}/2 is by −1-1.

Theorem 4.2. Let MM be a simply connected nonspin homology CP3\mathbb{C}P^3.

  1. If d≠0d\ne0, then there are short exact sequences

1→Im⁡ψ→haut⁡(M)→Z/2→11\to\operatorname{Im}\psi\to\operatorname{haut}(M)\to\mathbb{Z}/2\to1
1→G→Im⁡ψ→Z/2→1,1\to G\to\operatorname{Im}\psi\to\mathbb{Z}/2\to1,

where GG is isomorphic to

12∣d12\mid d3∣d,4∤d3\mid d,4\nmid d3∤d,4∣d3\nmid d,4\mid dotherwise
GGZ/6×Z/4\mathbb{Z}/6\times\mathbb{Z}/4Z/3×Z/4\mathbb{Z}/3\times\mathbb{Z}/4Z/2×Z/4\mathbb{Z}/2\times\mathbb{Z}/4Z/4\mathbb{Z}/4

Table 1.

The action of Z/2\mathbb{Z}/2 on GG is by −1-1. G⊂haut⁡0(M)G\subset\operatorname{haut}_0(M) is a subgroup of index 2.

  1. If d=l=0d=l=0, i.e. M=S4×S2M=S^4\times S^2, then there is a short exact sequence

1→Im⁡ψ→haut⁡(M)→Z/2→1,1\to\operatorname{Im}\psi\to\operatorname{haut}(M)\to\mathbb{Z}/2\to1,

where Im⁡ψ\operatorname{Im}\psi is isomorphic to G×(Z/2×Z/2)G\times(\mathbb{Z}/2\times\mathbb{Z}/2), G≅Z/6×Z/4G\cong\mathbb{Z}/6\times\mathbb{Z}/4, the actions of (1,0)(1,0) and (0,1)(0,1) on GG are by −1-1. G⊂haut⁡0(M)G\subset\operatorname{haut}_0(M) is a subgroup of index 2.

  1. If d=0d=0 and l≠0l\ne0, then Im⁡ψ\operatorname{Im}\psi is a subgroup of index 4, and there is a short exact sequence

0→G→Im⁡ψ→Z/2→0,0\to G\to\operatorname{Im}\psi\to\mathbb{Z}/2\to0,

where G≅Z/6×Z/4G\cong\mathbb{Z}/6\times\mathbb{Z}/4, the action of Z/2\mathbb{Z}/2 on GG is by −1-1. G⊂haut⁡0(M)G\subset\operatorname{haut}_0(M) is a subgroup of index 2

Proof of Theorem 4.1 and Theorem 4.2. First assume d≠0d \ne0. Then by Theorem 2.5 we have the following commutative diagram, where rows and columns are exact sequences

ker⁡ψ0⟶ker⁡ψ↓↓0⟶T(M)⟶MCG⁡(M)⟶Z/2⟶1↓ψ0↓ψ↓=0⟶haut⁡0(M)⟶haut⁡(M)⟶Z/2⟶1\begin{aligned} \ker\psi_0 &\longrightarrow\ker\psi\\ \downarrow&& \downarrow\\ 0 &\longrightarrow\mathcal{T}(M) \longrightarrow\operatorname{MCG}(M) \longrightarrow\mathbb{Z}/2 \longrightarrow1 \\ \downarrow{\psi_0} && \downarrow{\psi} && \downarrow{=} \\ 0 &\longrightarrow\operatorname{haut}_0(M) \longrightarrow\operatorname{haut}(M) \longrightarrow\mathbb{Z}/2 \longrightarrow1 \end{aligned}

When MM is spin, by Theorem 2.5 the upper exact sequence splits. If dd is odd, by Proposition 3.1 ψ0\psi_0 is surjective. This implies that ψ\psi is surjective, and haut⁡(M)=haut⁡0(M)⋊Z/2\operatorname{haut}(M)=\operatorname{haut}_0(M)\rtimes\mathbb{Z}/2, with haut⁡0(M)=Im⁡ψ0\operatorname{haut}_0(M)=\operatorname{Im}\psi_0. If dd is even and d≠0d\ne0, by Proposition 3.1 Im⁡ψ0\operatorname{Im}\psi_0 is a subgroup of index 2 in haut⁡0(M)\operatorname{haut}_0(M). Hence Im⁡ψ\operatorname{Im}\psi is a subgroup of index 2 in haut⁡(M)\operatorname{haut}(M), and isomorphic to Im⁡ψ0⋊Z/2\operatorname{Im}\psi_0\rtimes\mathbb{Z}/2. By Proposition 3.2 Im⁡ψ0\operatorname{Im}\psi_0 is isomorphic to T(M)/K\mathcal{T}(M)/K, whose isomorphism class is given in Proposition 2.2.

When MM is non-spin, dd is even. By Proposition 3.1 Im⁡ψ0\operatorname{Im}\psi_0 is a subgroup of index 2 in haut⁡0(M)\operatorname{haut}_0(M). Hence Im⁡ψ\operatorname{Im}\psi is a subgroup of index 2 in haut⁡(M)\operatorname{haut}(M). There is a short exact sequence

1⟶Im⁡ψ0⟶Im⁡ψ⟶Z/2⟶11\longrightarrow\operatorname{Im}\psi_0\longrightarrow\operatorname{Im}\psi\longrightarrow\mathbb{Z}/2\longrightarrow1

with Im⁡ψ0≅T(M)/K\operatorname{Im}\psi_0\cong\mathcal{T}(M)/K given in Proposition 2.3.

Now consider the cases with d=0d=0.

First assume l=0l=0. Then M=S2×S4M=S^2\times S^4 or S4×~S2S^4\widetilde{\times}S^2. By Theorem 2.6 we have a similar commutative diagram,

ker⁡ψ0⟶ker⁡ψ↓↓0⟶T(M)⟶MCG⁡(M)⟶Z/2×Z/2⟶1↓ψ0↓ψ↓=0⟶haut⁡0(M)⟶haut⁡(M)⟶Z/2×Z/2⟶1\begin{aligned} \ker\psi_0 &\longrightarrow\ker\psi\\ \downarrow&& \downarrow\\ 0 &\longrightarrow\mathcal{T}(M) \longrightarrow\operatorname{MCG}(M) \longrightarrow\mathbb{Z}/2\times\mathbb{Z}/2 \longrightarrow1 \\ \downarrow{\psi_0} && \downarrow{\psi} && \downarrow{=} \\ 0 &\longrightarrow\operatorname{haut}_0(M) \longrightarrow\operatorname{haut}(M) \longrightarrow\mathbb{Z}/2\times\mathbb{Z}/2 \longrightarrow1 \end{aligned}

where the homomorphisms from MCG⁡(M)\operatorname{MCG}(M) or haut⁡(M)\operatorname{haut}(M) to Z/2×Z/2\mathbb{Z}/2\times\mathbb{Z}/2 are induced by the actions on H2(M)H_2(M) and H4(M)H_4(M), and the two rows are split short exact sequences. By Proposition 3.1 Im⁡ψ0\operatorname{Im}\psi_0 is a subgroup of index 2 in haut⁡0(M)\operatorname{haut}_0(M), hence Im⁡ψ\operatorname{Im}\psi is a subgroup of index 2 in haut⁡(M)\operatorname{haut}(M), isomorphic to Im⁡ψ0⋊(Z/2×Z/2)\operatorname{Im}\psi_0\rtimes(\mathbb{Z}/2\times\mathbb{Z}/2), with Im⁡ψ0≅Z/2×Z/12\operatorname{Im}\psi_0\cong\mathbb{Z}/2\times\mathbb{Z}/12 if M=S2×S4M=S^2\times S^4, or Im⁡ψ0≅Z/6×Z/4\operatorname{Im}\psi_0\cong\mathbb{Z}/6\times\mathbb{Z}/4 if M=S4×~S2M=S^4\widetilde{\times}S^2.

If MM is spin with ll even and l≠0l\ne0, or if MM is nonspin with l≠0l\ne0, then we have the following commutative diagram

ker⁡ψ0⟶ker⁡ψ↓↓0⟶T(M)⟶MCG⁡(M)⟶Z/2⟶1↓ψ0↓ψ↓0⟶haut⁡0(M)⟶haut⁡(M)⟶Z/2×Z/2⟶1(6)\begin{aligned} \ker\psi_0 &\longrightarrow\ker\psi\\ \downarrow&& \downarrow\\ 0 &\longrightarrow\mathcal{T}(M) \longrightarrow\operatorname{MCG}(M) \longrightarrow\mathbb{Z}/2 \longrightarrow1 \\ \downarrow{\psi_0} && \downarrow{\psi} && \downarrow\\ 0 &\longrightarrow\operatorname{haut}_0(M) \longrightarrow\operatorname{haut}(M) \longrightarrow\mathbb{Z}/2\times\mathbb{Z}/2 \longrightarrow1 \tag*{(6)} \end{aligned}

where the first row is by Theorem 2.5, the second row by the fact that MM is homotopy equivalent to S2×S4S^2 \times S^4 or S4×S2S^4 \times S^2, and the homomorphism Z/2→Z/2×Z/2\mathbb{Z}/2 \to\mathbb{Z}/2 \times\mathbb{Z}/2 maps 11 to (1,0)(1,0). This shows that Im⁡ψ\operatorname{Im}\psi has index 44 in haut⁡(M)\operatorname{haut}(M). When MM is spin, the upper exact sequence splits (Theorem 2.5), therefore Im⁡ψ\operatorname{Im}\psi is isomorphic to Im⁡ψ0⋊Z/2\operatorname{Im}\psi_0 \rtimes\mathbb{Z}/2, with Im⁡ψ0≅Z/2×Z/12\operatorname{Im}\psi_0 \cong\mathbb{Z}/2 \times\mathbb{Z}/12. When MM is nonspin, there is a short exact sequence

1→Im⁡ψ0→Im⁡ψ→Z/2→11 \to\operatorname{Im}\psi_0 \to\operatorname{Im}\psi\to\mathbb{Z}/2 \to1

with Im⁡ψ0≅Z/6\operatorname{Im}\psi_0 \cong\mathbb{Z}/6.

If MM is spin and ll is odd, it is not known to the authors whether the image of the homomorphism haut⁡(M)→Z/2×Z/2\operatorname{haut}(M) \to\mathbb{Z}/2 \times\mathbb{Z}/2 in diagram (6) is Z/2\mathbb{Z}/2 or Z/2×Z/2\mathbb{Z}/2 \times\mathbb{Z}/2. Therefore Im⁡ψ\operatorname{Im}\psi is a subgroup of index 22 or 44 in haut⁡(M)\operatorname{haut}(M).

Proof of Proposition 1.2. The first statement is shown in Lemma 2.4. The second statement is shown by examining the codimension 44 splitting invariant. It suffices to consider the case M=XdM = X_d, the S2S^2-bundle over S4S^4 with characteristic map d∈π3SO(3)d \in\pi_3SO(3). Let ff be a smooth map, transverse to a fiber S2⊂XdS^2 \subset X_d, and homotopic to fηf_\eta. Then the preimage F=f−1(S2)F = f^{-1}(S^2) is a framed surface in XdX_d. By inspecting the construction of fηf_\eta one sees that FF is a disjoint union of the fiber S2S^2 and g−1(x0)g^{-1}(x_0), where g:S6→S4g:S^6 \to S^4 is a smooth map representing the non-trivial element of π6(S4)\pi_6(S^4) and x0∈S4x_0 \in S^4 is a regular value of gg. By the Thom-Pontrjagin construction, the set of concordance classes of framed surfaces in XdX_d is isomorphic to the homotopy set [Xd,S4][X_d,S^4]. Therefore in order to show that ff is not homotopic to a diffeomorphism, it suffices to show that the map π6(S4)→[Xd,S4]\pi_6(S^4) \to[X_d,S^4] induced by the collapsing map Xd→S6X_d \to S^6 is injective. From the cofiber sequence

S5→bPd→Xd→S6→ΣbΣPdS^5 \xrightarrow{b} P_d \to X_d \to S^6 \xrightarrow{\Sigma b} \Sigma P_d

one has an exact sequence [ΣPd,S4]→π6(S4)→[Xd,S4][\Sigma P_d,S^4] \to\pi_6(S^4) \to[X_d,S^4]. When dd is odd, ΣPd\Sigma P_d is homotopy equivalent to ΣCP2\Sigma\mathbb{C}P^2 and it is easy to show [ΣCP2,S4]=0[\Sigma\mathbb{C}P^2,S^4] = 0. When dd is even, it is known that Σb\Sigma b is dd-multiple of a certain element in π6(ΣPd)\pi_6(\Sigma P_d) ([21 Lemma 4.5]). Since π6(S4)≅Z/2\pi_6(S^4) \cong\mathbb{Z}/2, the map [ΣPd,S4]→π6(S4)[\Sigma P_d,S^4] \to\pi_6(S^4) must be trivial.

Proof of Theorem 1.3 and 1.6. The statements are rephrasing of Theorem 4.1 and 4.2. It remains to show that when dd and ll are even there is a splitting. Since fηf_\eta represents an element of order 22 in haut⁡(M)\operatorname{haut}(M), and by Proposition 1.2 fηf_\eta is not homotopic to a diffeomorphism, it provides a splitting of the exact sequence

1→MCG⁡(M)→haut⁡(M)→Z/2→1.1 \to\operatorname{MCG}(M) \to\operatorname{haut}(M) \to\mathbb{Z}/2 \to1.

Proof of Corollary 1.5. By comparing the cardinalities of the group GG in Theorem 1.3 and the group π6(Xd)\pi_6(X_d) in formula (2), one sees that the map π6(Xd)→haut⁡(Xd)\pi_6(X_d) \to\operatorname{haut}(X_d) is injective. Therefore exact sequence (1) is extended to a short exact sequence

0→π6(Xd)→haut⁡(Xd)→Gd→1.0 \to\pi_6(X_d) \to\operatorname{haut}(X_d) \to G_d \to1.

Note that homotopy equivalences constructed from π6(Xd)\pi_6(X_d) are all orientation-preserving. When dd is odd, Gd≅Z/2G_d \cong\mathbb{Z}/2, a splitting is given by the fiberwise antipodal map faf_a, which is orientation-reversing. When dd is even, Gd≅Z/2×Z/2G_d \cong\mathbb{Z}/2 \times\mathbb{Z}/2, a preimage of (1,0)∈Z/2×Z/2(1,0) \in\mathbb{Z}/2 \times\mathbb{Z}/2 in haut⁡(Xd)\operatorname{haut}(X_d) is given by faf_a. Since in this case Im⁡ψ\operatorname{Im}\psi is isomorphic to Z/2×Z/(12,d/2)\mathbb{Z}/2 \times\mathbb{Z}/(12,d/2), the order of the Dehn twist tρt_\rho must be 2 (Lemma 2.4), and is not in the image of π6(Xd)\pi_6(X_d). It is a preimage of (0,1)∈Z/2×Z/2(0,1) \in\mathbb{Z}/2 \times\mathbb{Z}/2. □

Tool Use. In this paper no instrument, software or AI was use.

References

Mathematisches Institut, Universität Bonn

Email address: [email protected]

HLM, School of Mathematics and Systems Science, Chinese Academy of Sciences; School of Mathematical Sciences, University of Chinese Academy of Sciences, Beijing 100190, China

Email address: [email protected]

School of Mathematical Sciences, University of Chinese Academy of Sciences, Beijing 100049, China

Email address: [email protected]

References

  1. [1]M. Atiyah, R. Bott, The Yang–Mills equations over Riemann surfaces, Philos. Trans. R Soc. Lond. Ser. A Math. Phys. Eng. Sci. 308 (1983), 523-615.
  2. [2]M. Arkowitz, The group of self-homotopy equivalences—a survey, Lecture Notes in Math., vol. 1425, Springer, 1990, pp. 170-203.DOI
  3. [3]H. Baues, On the group of homotopy equivalences of a manifold, Trans. A.M.S., vol. 348, No. 12, December 1996, 4737-4773.DOI
  4. [4]H. Baues, The degree of maps between certain 6-manifolds, Compositio Mathematica 110: 51-64, 1998.DOI
  5. [5]W. D. Barcus, M. G. Barratt, On the homotopy classification of the extension a fixed map, Trans. Amer. Math. Soc. 88 (1958), 57-74.DOI
  6. [6]H. Baues, J. Buth, On the group of homotopy equivalences of simply connected five manifolds, Math. Z., 222, 1996, 573-614.DOI
  7. [7]G. Brumfiel, Homotopy equivalences of almost smooth manifolds, Comment. Math. Helv. 46 (1971), 381-407.
  8. [8]J. Cerf, La strafification naturelle des espaces de fonctions différentiables réelles et le théorème de la pseudo-isotopie, Inst. Hautes Etudes Sci. Publ. Math. No. 39 (1970) 5-173.DOI
  9. [9]P. E. Jupp, Classification of certain 6-manifolds, Proc. Camb. Phil. Soc. (1973), 73, 293-300.
  10. [10]M. Kreck and W. Lück, The Novikov Conjecture: Geometry and Algebra, Birkhäuser Verlag, 2005.DOI
  11. [11]M. Kreck, On the classification of 1-connected 7-manifolds with torsion free second homology, J. Topology 11 (2018) No. 3, 720-738.DOI
  12. [12]M. Kreck and Y. Su, Mapping class group of manifolds which look like 3-dimensional complete intersections, Duke Math. J. 174 (2025), no. 3, 501-574.
  13. [13]J. W. Milnor, On simply-connected 4-manifolds, Symposium International de Topologia Algebraica, Mexico (1958), 122-128.
  14. [14]M. Olbermann, Involutions on S^6 with 3-dimensional fixed point set, Algebr. Geom. Topol. 10 (2010), no. 4, 1905-1932.DOI
  15. [15]M. Olbermann, Conjugations on 6-manifolds with free integral homology, Math. Ann. 353 (2012), no. 1, 65-93.arxiv.org/abs/1001.0911
  16. [17]N. Sawashita, On the group of self-equivalences of the product of spheres, Horishima Math. J. 5 (1975), 69-86.DOI
  17. [18]P. Teichner, On the signature of four-manifolds with universal cover spin, Math. Ann. 295, 745-759 (1993).DOI
  18. [19]C. T. C. Wall, Classification problems in differential topology V, Invent. math. 1, 355-374 (1966).DOI
  19. [21]K. Yamaguchi, The group of self-homotopy equivalences of S²-bundles over S⁴. I, Kodai Math. J. 9(1986), 308-326.DOI

Paper details

vlnc:W_t7df9snx
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents