Created: 2026-03-06 07:59:30
Updated: 2026-03-06 08:16:49

1. Fundamental groups

We first show that the loops with some fixed basepoint form a group under certain defined multiplication.

Let XX be a topological space and I=[0,1]I=[0,1], a continuous map α:IX\alpha:I\to X is called a path with initial point x0x_{0}, endpoint x1x_{1} if α(0)=x0,α(1)=x1\alpha(0)=x_{0},\alpha(1)=x_{1}. We call the path a loop if x0=x1x_{0}=x_{1} with base point x0x_{0}.

Let α,β:IX\alpha,\beta:I\to X be paths such that α(1)=β(0)\alpha(1)=\beta(0), then αβ\alpha*\beta is the product of two paths defined by

αβ(s)={α(2s),0s12β(2s1),12s1\begin{equation} \alpha*\beta(s)= \begin{cases} \alpha(2s), & 0\leq s\leq \frac{1}{2} \\ \beta(2s-1), & \frac{1}{2}\leq s\leq 1 \end{cases} \end{equation}

Let α:IX\alpha:I\to X be a path from x0x_{0} to x1x_{1}. The inverse α1\alpha ^{-1} is defined by

α1(s)α(1s),sI\begin{equation} \alpha ^{-1}(s)\equiv\alpha(1-s),s\in I \end{equation}

Let α,β:IX\alpha,\beta:I\to X be loops at x0x_{0}. They are homotopic, written as αβ\alpha\sim\beta, if there exists a continuous map F:I×IXF:I\times I\to X such that

F(s,0)=α(s),F(s,1)=β(s),F(0,t)=F(1,t)=x0\begin{equation}\begin{aligned} F(s,0)=\alpha(s),F(s,1)=\beta(s), \\ F(0,t)=F(1,t)=x_{0} \end{aligned} \end{equation}

The map FF is called a homotopy between α,β\alpha,\beta

In other words, if two paths can continuously deformed to each other, they are homotopic.

One can prove that homotopy is an equivalence relation; in other words, it satisfies the following three properties:

2.General properties of fundamental groups

Arcwise connected: there exists a path from any point to any point
Simply connected: any loop in X can be continuously shrunk to a point

2.1. Arcwise connectedness

Let XX be an arcwise connected topological space, x0,x1Xx_{0},x_{1}\in X. Then π1(X,x0)π1(X,x1)\pi_{1}(X,x_{0})\cong \pi_{1}(X,x_{1}) are isomorphic.

Consider η\eta a path from x0x_{0} to x1x_{1}. For [α]π1(X,x0)[\alpha]\in\pi_{1}(X,x_{0}), consider η1αη\eta ^{-1}*\alpha*\eta.

According to the theorem, for arcwise connected XX, we do not need t o specify the basepoint, and simply rewrite as π1(X)\pi_{1}(X).

2.2. Homotopic invariance of fundamental groups

The homotopic equivalence of paths and loops can be generalized to arbitary maps. If f,g:XYf,g:X\to Y are continuous maps, if there exists continuous map F:X×IYF:X\times I\to Y such that F(x,0)=f(x),F(x,1)=g(x)F(x,0)=f(x),F(x,1)=g(x), then ff is said to be homotopic to gg, denoted fgf\sim g. The map FF is called a homotopy between ff and gg.

X,YX,Y are topological spaces. They are of the same homotopy type, denoted as XYX\simeq Y, iff there exists f:XY,g:YXf:X\to Y,g:Y\to X such that fgidYf\circ g\sim id_{Y} and gfidXg\circ f\sim id_{X}. The map ff is called homotopy equivalence and gg its homotopy inverse.

"of the same homotopy type" is an equivalence relation in the set of topological spaces

  • XXX\simeq X, where idXid_{X} is a homotopy equivalence
  • If XYX\simeq Y with f,gf,g, then YXY\simeq X with function g,fg,f
  • XY,YZX\simeq Y,Y\simeq Z then XZX\simeq Z: if XYX\simeq Y with function f:XY,f:YXf:X\to Y,f':Y\to X, YZY\simeq Z with g:YZ,g:ZYg:Y\to Z,g':Z\to Y, thenXZX\simeq Z with functionfgf'\circ g' and gfg\circ f:

    (gf)(fg)=g(ff)ggidYg=ggidZ(fg)(gf)=f(gg)ffidYf=ffidX\begin{equation} \begin{aligned} (g\circ f )(f'\circ g') = g(f\circ f')g'\sim g\circ id_{Y}\circ g'=g\circ g'\sim id_{Z} \\ (f'\circ g')(g\circ f )=f'(g'\circ g)f \sim f'\circ id_{Y}\circ f= f'\circ f\sim id_{X} \end{aligned} \end{equation}

If X,YX,Y have the same homotopy type, f:XYf:X\to Y is a homotopy equivalence, then π1(X,x0)π1(Y,f(x0))\pi_{1}(X,x_{0})\cong \pi_{1}(Y,f(x_{0}))

A fundamental group is invariant under homeomorphisms and hence is a topological invariant.

Fundamental groups are a weak classification of topological spaces compared with homeomorphisms. If X and Y have different fundamental groups, they are not homeomorphic.

Let RR be a subspace of XX. If \exists continuous map f:XRf:X\to R, fR=idRf|_{R}=id_{R}, then RR is called a retract of XX and ff a retraction.

Let RR a subspace of XX. If \exists continuous H:X×IXH:X\times I\to X, such that

H(x,0)=x,H(x,1)R,xXH(x,t)=x,xR,tI\begin{equation} \begin{aligned} H(x,0)=x,H(x,1)\in R,\forall x\in X \\ H(x,t)=x,\forall x\in R,t\in I \end{aligned} \end{equation}

The spaceRR is a deformation retract of XX. HH is a homotopy between idxid_{x} and a retraction f:XRf:X\to R.


RR is a retract but not deformation retract of XX, since the hole in XX is an obstruction to continuous deformation of idxid_{x} to retraction.

Let X={eiθ,0θ<2π}X=\{e^{i\theta},0\leq\theta<2\pi\}, Y={reiθ,0θ<2π,12r23}Y=\left\{ re^{i\theta},0\leq \theta<2\pi, \frac{1}{2}\leq r \leq \frac{2}{3} \right\}
Define f:XY,f(eiθ)=eiθf:X\to Y,f(e^{i\theta})=e^{i\theta}, g:YX,g(reiθ)=eiθg:Y\to X, g(r e^{i\theta})=e^{i\theta}. Observe that fgidYf\circ g\sim id_{Y} and gf=idXg\circ f=id_{X}

If a point aXa\in X is a deformation retract of XX, XX is contractible

  • Rn\mathbb{R}^{n} is contractible to the origin: H:Rn×IRnH:\mathbb{R}^{n}\times I\to \mathbb{R}^{n} by H(x,t)=txH(x,t)=tx.
  • D2={(x,y)x2+y21}D^{2}=\{(x,y)\mid x^{2}+y^{2}\leq 1\}. Show that S1S^{1} is a deformation retract of D2{0}D^{2}-\{0\}
    • H:(D2{0})×IS1:(r,θ;t)(r=t+(1t)r,θ=θ)H:\left(D^{2}-\{0\}\right)\times I \to S^{1}: (r,\theta;t)\to(r'=t+(1-t)r,\theta'=\theta)
  • SnS^{n} is a deformation retract of Dn+1{0}D^{n+1}-\{0\}.

A fundamental group of a contractible space XX is trivial, π1(X,x0){e}\pi_{1}(X,x_{0})\cong \{e\}.

π1(S1)Z\pi_{1}(S^{1})\cong \mathbb{Z}

3 Fundamental group of torus

Let X,YX,Y be arcwise connected topological spaces. then

π1(X×Y,(x0,y0))π1(X,x0)π1(Y,y0)\begin{equation} \pi_{1}(X\times Y,(x_{0},y_{0}))\cong \pi_{1}(X,x_{0})\oplus \pi_{1}(Y,y_{0}) \end{equation}

T2=S1×S1T^{2}=S^{1}\times S^{1}, so π1(T2)ZZ\pi_{1}(T^{2})\cong \mathbb{Z}\oplus \mathbb{Z}
X=S1×RX=S^{1}\times \mathbb{R}, since π1(R)={e}\pi_{1}(\mathbb{R})=\{e\}, so π1(X)Z\pi_{1}(X)\cong \mathbb{Z}

4. Fundamental groups of polyhedra

Free groups: a subset of a group GG is a free set of generators if any gGg\in G, geg\neq e can be uniquely written as g=x1i1xning=x_{1}^{i_{1}}\dots x_{n}^{i_{n}}, nn finite and ikZi_{k}\in\mathbb{Z}. Adjacent x is not equal and x0x^{0} term should be dropped. If so GG is called free group.
Given a set of generators XX, we can construct a free group.The set of all reduced words form a well-defined free group denoted as F[X]F[X]. For example, F[{a}]=ZF[\{a\}]=\mathbb{Z}.
For a group GG generated by XX, any element gGg\in G can be written as g=x1i1xning=x_{1}^{i_{1}}\dots x_{n}^{i_{n}}. GG is not necessarily free. Let F[X]F[X] be the free group generated by XX, define a natural homomorphism φ:F[X]G\varphi:F[X]\to G,

x1i1xninx1i1xninG\begin{equation} x_{1}^{i_{1}}\dots x_{n}^{i_{n}}\to x_{1}^{i_{1}}\dots x_{n}^{i_{n}}\in G \end{equation}

LHS is not unique. so F[X]/kerφGF[X]/ker \varphi \cong G.

As noted earliear, the polyhedron K|K| is a nice approximation of a given topological space XX within a homeomorphism. Since fundamental groups are topological invariants, π1(X)=π1(K)\pi_{1}(X)=\pi_{1}(|K|). We assume that XX is arcwise connected space, drop base point. The problem becomes calculation of π1(K)\pi_{1}(|K|).

Define the edge path in simplicial complex KK as a sequence of vertices v0vivi+1vkv_{0}\dots v_{i}v_{i+1}\dots v_{k} of K|K|. Each adjacent point pairs are a 1-simplex. We allow that vi=vi+1v_{i}=v_{i+1}, in this case viviv_{i}v_{i} is a 0-simplex. If v0=vk=vv_{0}=v_{k}=v, then the path is a loop.

We classify loops into equivalence classes according to some equivalence relation.

Two edge loops α,β\alpha,\beta are equivalent iff one is obtained from the other by finite number of times of the following operations:

  • If u,v,wu,v,w span 2-simplex in KK, edge path uvwuvw may be replaced to uwuw, and vice versa.
  • If u=wu=w, the uvwuvw corresponds to traversing along uvuv first then go backwords. The uvuuvu can be replaced to uu.

Define the product operation of edge paths

{vu1uk1v}{vv1vi1v}={vu1uk1vv1vi1v}\begin{equation} \left\{v u_{1}\dots u_{k-1}v\right\}*\left\{v v_{1}\dots v_{i-1}v\right\}=\left\{v u_{1}\dots u_{k-1} v v_{1}\dots v_{i-1}v\right\} \end{equation}

The unit element is {v}\{v\} and inverse is defined as

{vu1ukv}1={vuku1v}\begin{equation} \left\{v u_{1}\dots u_{k}v\right\}^{-1} = \left\{v u_{k}\dots u_{1} v\right\} \end{equation}

The group is claled edge group of KK at vv, denoted as E(K;v)E(K;v).

E(K;v)E(K;v) is isomorphic to π1(K;v)\pi_{1}(|K|;v)

Given an arcwise-connected simplicial complex KK, there always exists subcomplex LL satisying:

Some more specific relations:

The rule is summarized as follows:

定义K(2)K^{(2)}KK的2-skeleton,如果其只包含KK的全部0,1,2-simplex. 我们有

π1(K)π1(K(2))\begin{equation} \pi_{1}(|K|) \cong \pi_{1}(|K^{(2)}|) \end{equation}

例如,K\left| K \right|是三维球D3D^{3}的多面体,其边界L|L|S2S^{2}的多面体。由于D3D^{3}可收缩(contractible),π1(K){e}\pi_{1}(|K|)\cong \{e\}. 因此π1(S2)π1(K(2))π1(K){e}\pi_{1}(S^{2})\cong\pi_{1}(K^{(2)})\cong \pi_{1}(|K|)\cong\{e\}. In general, n2n\geq 2, (n+1)(n+1)-simplex σn+1\sigma_{n+1} and σn+1\partial \sigma_{n+1} has the same 2-skeleton. If we find that σn+1\sigma_{n+1} is contractible, and σn+1\partial\sigma_{n+1} is a polyhedron of SnS^{n}, we have π1(Sn){e},n2\pi_{1}(S^{n})\cong\{e\},n\geq 2

K{v1,v2,v3,<v1v2>,<v2v3>,<v1v3>}K\equiv\left\{v_{1},v_{2},v_{3},\left<v_{1}v_{2}\right>,\left<v_{2}v_{3}\right>,\left<v_{1}v_{3}\right>\right\}S1S^{1}的simplicial complex, 令v1v_{1}为基点。最大树可以选做L={v1,v2,v3,<v1v2>,<v1v3>}L=\{v_{1},v_{2},v_{3},\left<v_{1}v_{2}\right>,\left<v_{1}v_{3}\right>\}, 于是KLK-L剩余的生成元只有g23g_{23}。由于KK没有2-simplex,条件集合为空集。从而

π1(S1)G(K,L)=(g23;)Z\begin{equation} \pi_{1}(S^{1})\cong G(K,L)=(g_{23};\emptyset)\cong\mathbb{Z} \end{equation}

nn-bouquet被定义为n个圈有一个点相连。
对于上图的3-bouquet,生成元为g12,g34,g56g_{12},g_{34},g_{56},规则集合为空,于是π1(3-bouquet)=G(K;L)=(x,y,z;)\pi_{1}(\text{3-bouquet})=G(K;L)=(x,y,z;\emptyset)。注意群非阿贝尔,因为αβα1β\alpha*\beta*\alpha ^{-1}\neq\beta.

环面的triangulation。下面阴影区域被选作subcomplex LL,可以验证其包含所有点,道路连通且单连通。一共有11个生成元(02,07,73,38,84,42,28,81,16,60,04)和10组关系(用下图字母表示)。取x=g02,y=g04x=g_{02},y=g_{04},则


于是 x1yx=g28,x1yx=yx ^{-1} y x=g_{28}, x ^{-1} y x=y,或 xyx1y1=1xyx ^{-1} y ^{-1}=1. 这说明G(K;L)G(K;L)由两个互易的生成元生成。从而G(K;L)=(x,y;xyx1y1)ZZG(K;L)=(x,y;xyx^{-1}y ^{-1}) \cong \mathbb{Z}\oplus \mathbb{Z}。考虑α=0120,β=0340\alpha=0\to1\to2\to0,\beta=0\to3\to4\to0α\alpha圈与x=g02x=g_{02}认同,因为g12=g01=g_{12}=g_{01}=,而圈β\betay=g04y=g_{04}认同。他们生成π1(T2)\pi_{1}(T^{2}),由于α,β\alpha,\beta为独立的非平凡圈,独立是因为他们的关系写作

αβα1β1cv\begin{equation} \alpha*\beta*\alpha ^{-1}*\beta ^{-1}\sim c_{v} \end{equation}

其中cvc_{v}为平凡的圈。一般的,令Σg\Sigma_{g}为亏格gg的环。则存在Σg\Sigma_{g}的基本群由 2g2gαi,βi\alpha_{i},\beta_{i}生成,1ig1\leq i\leq g.

triangulation of Klein bottle. 同样有11个生成元和10组关系。令x=g02,y=g04x=g_{02},y=g_{04},可以证明xyxy1=exyxy^{-1}=e,则

π1(Klein Bottle)(x,y;xyxy1)\begin{equation} \pi_{1}(\text{Klein Bottle})\cong (x,y;xyxy^{-1}) \end{equation}

: π1(RP2)(x;x2)Z2\pi_{1}(\mathbb{R}P^{2})\cong(x;x^{2})\cong \mathbb{Z}_{2}

Relation between H1(K)H_{1}(K) and π1(K)\pi_{1}(|K|)

Let KK be a connected simplicial complex, then H1(K)π1(K)/FH_{1}(K)\cong \pi_{1}(\left| K \right|)/ F, where FF is the commutator subgroup of π1(K)\pi_{1}(\left| K \right|).

Let G=(xi;rm)G=(x_{i};r_{m}), then the commutator subgroup is a gr pgenerated by elements of the form xixjxi1xj1x_{i}x_{j}x_{i}^{-1}x_{j}^{-1}. Thus, G/FG / F is the group generated by {xi}\{x_{i}\} with relation {rm}{xixjxi1xj1}\{r_{m}\}\cup\{x_{i}x_{j}x_{i}^{-1}x_{j}^{-1}\}.

For example, if π1(2-bouquet)=(x,y;)\pi_{1}(\text{2-bouquet})=(x,y;\emptyset), then H1(K)π1(2-bouquet)/F(x,y;xyx1y1)=ZZH_{1}(K)\cong \pi_{1}(\text{2-bouquet}) / F \cong (x,y;xyx^{-1}y^{-1})=\mathbb{Z}\oplus \mathbb{Z}

H1(Klein Bottle)π1(Klein Bottle)/F(x,y;xyxy1,xyx1y1)=(x,y;xyx1y1,x2)ZZ2\begin{equation} H_{1}(\text{Klein Bottle})\cong\pi_{1}(\text{Klein Bottle}) / F \cong (x,y;xyxy^{-1},xy x ^{-1} y ^{-1})= (x,y; xyx ^{-1} y ^{-1},x^{2})\cong \mathbb{Z}\oplus \mathbb{Z}_{2} \end{equation}

5. Higher homotopy groups

The fundamental group classifies the homotopy classes of loops in a topological space XX. We can also classify homotopy classes of the spheres in XX or homotopy classes of the tori in XX. It turns out that the homotopy classes of Sn(n2)S^{n}(n\geq 2) form a group similar to the fundamental group.

Let In/InI^{n} / \partial I^{n} be the cube InI^{n} whose boundary In\partial I^{n} is shrunk to a point, we have In/InSnI^{n} / \partial I^{n} \cong S^{n}, the continuous maps α:InX\alpha:I^{n}\to X is called nn-loop at x0x_{0} if In\partial I^{n} is mapped to x0x_{0}.

Let XX be a topological space, and α,β:InX\alpha ,\beta: I^{n}\to X be nn-loops at x0Xx_{0}\in X. The map α\alpha is homotopic to β\beta, denoted as αβ\alpha\sim\beta, if F:In×IX\exists F: I^{n}\times I\to X which is continuous and satisfies

F(s1,,sn,0)=α(s1,,sn)F(s1,,sn,1)=β(s1,,sn)F(s1,,sn,t)=x0 for (s1,,sn)In,tI\begin{equation} \begin{aligned} F(s_{1},\dots,s_{n},0) & =\alpha(s_{1},\dots,s_{n}) \\ F(s_{1},\dots,s_{n},1) & =\beta(s_{1},\dots,s_{n}) \\ F(s_{1},\dots,s_{n},t) & =x_{0} \text{ for }(s_{1},\dots,s_{n})\in \partial I^{n},t\in I \end{aligned} \end{equation}

FF is called a homotopy between α,β\alpha,\beta

Next we define group operations. αβ\alpha*\beta is defined by

αβ(s1,,sn)={α(2s1,,sn),0s112β(2s11,,sn),12s11\begin{equation} \alpha*\beta(s_{1},\dots,s_{n})=\begin{cases} \alpha(2s_{1},\dots,s_{n}), & 0\leq s_{1}\leq \frac{1}{2} \\ \beta(2s_{1}-1,\dots,s_{n}), & \frac{1}{2}\leq s_{1} \leq 1 \end{cases} \end{equation}

and

α1(s1,,sn)=α(1s1,,sn)\begin{equation} \alpha ^{-1}(s_{1},\dots,s_{n})=\alpha(1-s_{1},\dots,s_{n}) \end{equation}

We have α1αcx0\alpha ^{-1}*\alpha\sim c_{x_{0}}, where cx0c_{x_{0}} is a constant nn-loop, cx0(s1,,sn)=x0c_{x_{0}}(s_{1},\dots,s_{n})=x_{0}.

The nnth homotopy group at x0x_{0}, denoted πn(x,x0)\pi_{n}(x,x_{0}), is the set of homotopy classes of nn-loops (n1)(n\geq 1) at x0Xx_{0}\in X.

Let us consider π0(X,x0)\pi_{0}(X,x_{0}). In this case I0={0}I^{0}=\{0\} and I0=\partial I^{0}=\emptyset, Let α,β:{0}X\alpha,\beta:\{0\}\to X be such that α(0)=x,β(0)=y\alpha(0)=x,\beta(0)=y. We define αβ\alpha\sim\beta if there exists continuous map F:{0}×IXF:\{0\}\times I\to X such that F(0,0)=x,F(0,1)=yF(0,0)=x,F(0,1)=y. This shows that αβ\alpha\sim\beta iff x,yx,y are connected by a curve in XX. This equivalence relation is independe-nt of x0x_{0} and denotes the number of arcwise connected components of XX.

6. General properties of higher homotopy groups

  1. Abelian: αββα\alpha * \beta \sim \beta * \alpha, or [α][β]=[β][α][\alpha]*[\beta]=[\beta]*[\alpha]
  2. If a topological space XX is arcwise connected, πn(X,x0)πn(X,x1)\pi_{n}(X,x_{0})\cong \pi_{n}(X,x_{1}) for any x0,x1Xx_{0},x_{1}\in X
  3. If f:XYf:X\to Y is a homotopy equivalence, then πn(X,x0)πn(Y,f(x0))\pi_{n}(X,x_{0})\cong\pi_{n}(Y,f(x_{0})). If XX is contractible, then πn(X,x0)={e},n>1\pi_{n}(X,x_{0})=\{e\},n>1.
  4. Let X,YX,Y be arcwise connected topological spaces, then πn(X×Y)πn(X)πn(Y)\pi_{n}(X\times Y)\cong \pi_{n}(X)\oplus \pi_{n}(Y)
  5. Let X,YX,Y be arcwise connected topological spaces, thenπn(X×Y)πn(X)πn(Y)\pi_{n}(X\times Y) \cong \pi_{n}(X)\oplus \pi_{n}(Y)

X,X~X,\tilde{X}为连通的拓扑空间。(X~,p)(\tilde{X},p)对被称作XX的覆盖空间,如果存在连续映射p:X~Xp:\tilde{X}\to X使得

  • pp为满射
  • 对每个xXx\in X,存在连通开集UXU\subset X包含xx,使得p1(U)p^{-1}(U)X~\tilde{X}中开集的不交并集,每一个都被pp同胚地映到UU
    如果X~\tilde{X}还是单连通的,(X~,p)(\tilde{X},p)被称作universal covering space.

R\mathbb{R} is a universal covering space of S1S^{1}. Since S1S^{1} si identified with U(1)U(1), R\mathbb{R} is a universal covering group of U(1)U(1) if R\mathbb{R} is regarded as an additive group. The map p:RU(1)p:\mathbb{R}\to U(1) may be p:xei2πxp:x\to e^{i2\pi x}, pp is surjective and if U={e2πixx(x00.1,x0+0.1)}U=\{e^{2\pi ix}\mid x\in(x_{0}-0.1,x_{0}+0.1)\}, then

p1(U)=nZ(x00.1+n,x0+0.1+n)\begin{equation} p ^{-1} (U) = \bigcup_{n\in \mathbb{Z}} (x_{0}-0.1+n,x_{0}+0.1+n) \end{equation}

which is a disjoint union of open sets of R\mathbb{R}. pp is also a homomorphism w.r.t. (R,+)\left(\mathbb{R},+\right) and (S1,)(S^{1},\cdot). So (R,p)(\mathbb{R},p) is the universal covering group of S1S^{1}.

If (X~,p)(\tilde{X},p) is the universal covering space of a connected topological space XX, and x0X,p(x~0)=x0x_{0}\in X,p(\tilde{x}_{0})=x_{0}, then induced homomorphism

p:πn(X~,x~0)πn(X,x0)\begin{equation} p_{*}: \pi_{n}(\tilde{X},\tilde{x}_{0})\to \pi_{n}(X,x_{0}) \end{equation}

is an isomorphism for n2n\geq 2.
Note: when n=1n=1, π1(R)={e},π1(S1)=Z\pi_{1}(\mathbb{R})=\{e\},\pi_{1}(S^{1})=\mathbb{Z}

The real projective space RPn\mathbb{R} P^{n} is obtained from SnS^{n} by identifying (x,x)(x,-x). So it is easy to see SnS^{n} is a covering space of RPn\mathbb{R}P^{n} for n2n\geq 2. Since π1(Sn)={e}\pi_{1}(S^{n})=\{e\}, n2n\geq 2, we have

πn(RPm)πn(Sm)\begin{equation} \pi_{n}(\mathbb{R}P^{ m})\cong \pi_{n}(S^{m}) \end{equation}

Note that RP3\mathbb{R}P^{3} is identified with SO(3)SO(3). SO(3)SO(3) can be seen as a disc D3D^{3} of radius π\pi whose antipodal points on S2S^{2} identified. We may also express RP3\mathbb{R}P^{3} as nothern hemisphere D3D^{3} of S3S^{3}, whose antipodal points on the boundary S2S^{2} are identified.

S3S^{3} is identified with SU(2)SU(2) too: express gSU(2)g\in SU(2) as

g=(abˉbaˉ)\begin{equation} g=\left(\begin{matrix} a & -\bar{b} \\ b & \bar{a} \end{matrix}\right) \end{equation}

and a2+b2=1|a|^{2}+|b|^{2}=1. Rewrite a=u+iv,b=x+iya=u+iv,b=x+iy, this becomes S3S^{3}: u2+v2+x2+y2=1u^{2}+v^{2}+x^{2}+y^{2}=1

We find that

πn(Sn,x0)Z,n1\begin{equation} \pi_{n}(S^{n},x_{0})\cong\mathbb{Z},n\geq 1 \end{equation}

If α\alpha maps SnS^{n} to a point, [α][\alpha] is the unit element. If α\alpha maps In/InI^{n} / \partial I^{n} and SnS^{n} in the same sense of orientation, then [α][\alpha] is 1Z1\in \mathbb{Z}. Consider n=2n=2, I2/I2S2I^{2} / \partial I^{2}\cong S^{2}, use (θ,ϕ)(\theta,\phi) to represent points on them. α:(θ,ϕ)(θ,ϕ)\alpha:(\theta,\phi)\to(\theta',\phi'). If α(θ)=θ,α(ϕ)=ϕ\alpha(\theta)=\theta',\alpha(\phi)=\phi', the point sweeps S2S^{2} once, [α]=1Z[\alpha]=1\in \mathbb{Z}. If α(θ,ϕ)=(θ,2ϕ)\alpha(\theta,\phi)=(\theta,2\phi), the point sweeps S2S^{2} twice while (θ,ϕ)\left(\theta,\phi\right) scans I2I^{2} once. [α]=2Z[\alpha]=2\in \mathbb{Z}. So α(θ,ϕ)=(θ,kϕ)\alpha(\theta,\phi)=(\theta,k\phi) gives [α]=kZ[\alpha]=k\in \mathbb{Z}.

Since SnS^{n} is universally covering space of RPn\mathbb{R}P^{n}, πn(RPn)πn(Sn)Z,n2\pi_{n}(\mathbb{R}P^{n})\cong \pi_{n}(S^{n})\cong\mathbb{Z},n\geq 2

  • Since Spin(4)=SU(2)×SU(2)Spin(4)=SU(2)\times SU(2) universally covers SO(4)SO(4), πn(SO(4))=πn(SU(2))πn(SU(2)),n2\pi_{n}(SO(4))=\pi_{n}(SU(2))\oplus \pi _n(SU(2)),n\geq 2
  • JJ-homomorphism: πk(SO(n))πk+n(Sn)\pi_{k}(SO(n))\to \pi_{k+n}(S^{n}), so we have

    π1(SO(2))π3(S2)Zπ1(SO(3))π4(S3)π4(SU(2))π4(SO(3))Z2\begin{equation} \begin{aligned} \pi_{1}(SO(2))\cong \pi_{3}(S^{2})\cong\mathbb{Z} \\ \pi_{1}(SO(3))\cong\pi_{4}(S^{3})\cong \pi_{4}(SU(2))\cong \pi_{4}(SO(3))\cong\mathbb{Z}_{2} \end{aligned} \end{equation}

  • Bott periodicity theorem

    πk(U(n))πk(SU(n)){{e},k evenZ,k odd\begin{equation} \pi_{k}(U(n))\cong\pi_{k}(SU(n))\cong \begin{cases} \left\{e\right\}, & k \text{ even} \\ \mathbb{Z}, & k \text{ odd} \end{cases} \end{equation}

    for nk+12n\geq \frac{k+1}{2}

8. Orders in condensed matter system

1. Order parameter

Let HH be a Hamiltonian describing a condensed matter system. Assume HH is invariant under a certain symmetry operation. The ground state need not preserve the symmetry of HH. If so the system undergoes spontaneous symmetry breakdown.
Consider heisenberg hamiltonian

H=J(i,j)SiSj+hiSi\begin{equation} H=-J \sum_{(i,j)} \mathbf{S}_{i}\cdot \mathbf{S}_{j} + \mathbf{h}\cdot \sum_{i} \mathbf{S}_{i} \end{equation}

which describes NN ferromagnetic Heisenberg spins {Si}\{\mathbf{S}_{i}\}, where (i,j)(i,j) sums over nearest neighbours, hh is the external magnetic field. The free energy FF is defined by exp(βF)=Z\exp(-\beta F)=Z, and average magnetization (β?\beta?)

m=1Ni<Si>=1NβFh\begin{equation} \mathbf{m} = \frac{1}{N} \sum_{i}\left<\mathbf{S}_{i}\right> = \frac{1}{N\beta} \frac{ \partial F }{ \partial \mathbf{h} } \end{equation}

Consider h0\mathbf{h}\to0 limit. Although HH is invariant under SO(3)SO(3) rotations for all SiS_{i} in this limit, under large β\beta limit, the system does not observe SO(3)SO(3) symmetry. It is called spontaneous magnetization. The maximum temperature such that m0\mathbf{m}\neq 0 is called critical temperature.
m\mathbf{m} is the order parameter describing the phase transition between ordered state (m0\mathbf{m}\neq 0) and disordered state (m0\mathbf{m}\neq 0).

10 Textures in Superfluid He3-A

From NMR and other observations, the superfluid is in the spin-triplet, p-wave state. Instead of field operators Ψαβ(x)=Δμ(x)i(σμσ2)αβ\Psi_{\alpha\beta}(\vec{x})=\Delta ^{\mu}(\vec{x}) i(\sigma_{\mu}\cdot\sigma_{2})_{\alpha\beta}

The most general form of triplet superfluid order parameter is

<cα,kcβ,k>μ=13(iσ2σμ)αβdμ(k)\begin{equation} \left<c_{\alpha,\vec{k}}c_{\beta,-\vec{k}}\right> \propto \sum_{\mu=1}^{3} (i\sigma_{2}\sigma_{\mu}) _{\alpha\beta} d_{\mu}(\vec{k}) \end{equation}

where α,β\alpha,\beta are spin indices. The Cooper pair forms in the p-wave state, so dμ(k)Y1mkid_{\mu}(\vec{k})\propto Y_{1m}\sim k_{i},

dμ(k)=i=13Δ0Aμiki\begin{equation} d_{\mu}(\vec{k})= \sum_{i=1}^{3} \Delta_{0}A_{\mu i}k_{i} \end{equation}

Phase Diagram of He3
The A-phase order parameter takes the form Aμi=dμ(Δ1+iΔ2)iA_{\mu i}= d_{\mu}(\Delta_{1}+i\Delta_{2})_{i}, d\vec{d} is a unit vector along which the spin projection of Cooper pair vanishes. (Δ1,Δ2)(\Delta_{1},\Delta_{2}) is a pair of orthonormal unit vectors. d\vec{d} takes value from S2S^{2}. Let lΔ1×Δ2\vec{l}\equiv\Delta_{1}\times\Delta_{2}, the (Δ1,Δ2,l)(\Delta_{1},\Delta_{2},\vec{l}) forms an orthonormal frame at each point of medium. Since any orthonormal frame is obtained from fixed frame by a rotation, the order parameter of He-3A is S2×SO(3)S^{2}\times SO(3). The vector l\vec{l} is the axis of angular momentum of Cooper pair.

Neglect variation of d^\hat{d} for simplicity (in fact, d^\hat{d} is locked along l^\hat{l} due to dipole force). The order parameter assumes the form Ai=Δ0(Δ^1+Δ^2)iA_{i}=\Delta_{0}(\hat{\Delta}_{1}+\hat{\Delta}_{2})_{i}, where Δ^1,Δ^2,l^Δ^1×Δ^2\hat{\Delta}_{1},\hat{\Delta}_{2},\hat{l}\equiv\hat{\Delta}_{1}\times\hat{\Delta}_{2} forms an orthonormal frame at each point of the medium. Take a standard frame e1,e2,e3e_{1},e_{2},e_{3}, the frame is obtained by applying gSO(3)g\in SO(3) to (e1,e2,e3)(e_{1},e_{2},e_{3}). So (Δ^1(x),Δ^2(x),l^(x))(\hat{\Delta}_{1}(\vec{x}),\hat{\Delta}_{2}(\vec{x}),\hat{l}(\vec{x})) defines a map ψ:XSO(3)\psi:X\to SO(3), called texture of superfluid He-3. The relevant homotopy groups for classifying defects are πn(SO(3))\pi_{n}(SO(3)).

Brouwer fixed-point theorem: prove that there are no smooth function f:D2D2f:D^{2}\to D^{2} such that ff has no fixed points, namely f(p)pf(p)\neq p for any pD2p\in D^{2}.

Let S1=D2S^{1}=\partial D^{2} be its boundary. Suppose there exists such ff, consider a semi-line startng at pp through f(p)f(p) and it intersects S1S^{1} at point qq (this is well defined since f(p)pf(p)\neq p). Define f~:D2S1\tilde{f}:D^{2}\to S^{1} by f~(p)=q\tilde{f}(p)=q. f~\tilde{f} is also smooth. However, π1(S1)=Z\pi_{1}(S^{1})=\mathbb{Z} and π1(D2)={0}\pi_{1}(D^{2})=\{0\}, such smooth f~\tilde{f} does not exist and hence, ff must have fixed points.

Leave a Comment

captcha
Fontsize