连续对称变换
考虑连续对称场变换及其无穷小形式
$$x\to x’\qquad \Phi(x)\to\Phi’(x’)=\mathcal{F}(\Phi(x))$$
$$x’^\mu=x^\mu+\omega_a\frac{\delta x^\mu}{\delta\omega_a}\qquad \Phi’(x’)=\Phi(x)+\omega_a\frac{\delta \mathcal{F}}{\delta\omega_a}(x)$$
可定义变换生成元 $G_a$
$$\delta_\omega\Phi(x)=\Phi’(x)-\Phi(x)=-i\omega_aG_a\Phi(x)$$
$$iG_a\Phi=\frac{\delta x^\mu}{\delta\omega_a}\partial_\mu\Phi-\frac{\delta \mathcal{F}}{\delta\omega_a}$$
平移、缩放与旋转变换及其生成元形式如下
$$\begin{array}{l}&\Phi’(x+a)=\Phi(x)\qquad & P_\mu=-i\partial_\mu\\
\quad\\
&\Phi’(\lambda x)=\lambda^{-\Delta}\Phi(x)\qquad & D=-ix^\mu\partial_\mu-i\Delta\\
\quad\\
&\Phi’(\Lambda x)=U(\Lambda)\Phi(x)\qquad & L^{\mu\nu}=i(x^\mu\partial^\nu-x^\nu\partial^\mu)+S^{\mu\nu}\end{array}$$
Noether定理
Noether定理表述了作用量的每个连续对称变化都对应一个经典守恒流,考虑以下作用量及其由 $\omega_a G_a$ 生成的对称变化
$$\mathcal{S}=\int\text{d}^d x\;\mathcal{L}(\Phi,\partial_\mu\Phi)$$
$$S’=\int\text{d}^d x\left|\frac{\partial x’}{\partial x}\right|\;\mathcal{L}(\mathcal{F}(\Phi(x)),\frac{\partial x^\nu}{\partial x’^\mu}\partial_\nu\mathcal{F}(\Phi(x)))$$
其无穷小变化形式如下
$$S’=\int\text{d}^d x\left(1+\partial_\mu\left(\omega_a\frac{\delta x^\mu}{\delta\omega_a}\right)\right)\mathcal{L}\left(\Phi+\omega_a\frac{\delta\mathcal{F}}{\delta\omega_a},\left(\delta_\mu^\nu-\partial_\mu\left(\omega_a\frac{\delta x^\nu}{\delta\omega_a}\right)\right)\left(\partial_\nu\Phi+\partial_\nu\left(\omega_a\frac{\delta\mathcal{F}}{\delta\omega_a}\right)\right)\right)$$
由于作用量在刚性变换下对称,即 $\omega_a$ 为常数时 $\delta S=S’-S=0$,因此作用量一阶变分仅含 $\omega_a$ 导数项,可将其写为以下形式
$$\delta S=-\int \text{d}^d x\; j_a^\mu\partial_\mu\omega_a=\int \text{d}^d x\; \omega_a\partial_\mu j_a^\mu$$
与无穷小变化形式比较即得 $j_a^\mu$ 表达式
$$j_a^\mu=\left(\frac{\partial\mathcal{L}}{\partial(\partial_\mu\Phi)}\partial_\nu\Phi-\delta_\nu^\mu\mathcal{L}\right)\frac{\delta x^\nu}{\delta\omega_a}-\frac{\partial\mathcal{L}}{\partial(\partial_\mu\Phi)}\frac{\delta \mathcal{F}}{\delta\omega_a}$$
若场满足经典运动方程,则对任意无穷小参数 $\omega_a(x)$ 均有 $\delta S=0$,可知 $j_a^\mu$ 散度为零,即 $j_a^\mu$ 为守恒流,其对应守恒荷 $Q_a$
$$\partial_\mu j_a^\mu=0\qquad Q_a=\int\text{d}^{d-1}x\; j^0_a$$
$$\frac{\dd Q_a}{\dd t}=\int\text{d}^{d-1}x\; \partial_0 j^0_a=-\int\text{d}^{d-1}x\; \partial_i j^i_a=0$$
其中守恒流添加一项反对称张量的散度仍可满足连续性方程
$$j_a^\mu\to j_a^\mu+\partial_\nu B^{\nu\mu}_a\qquad B^{\nu\mu}_a=-B^{\mu\nu}_a$$
考虑平移变换,其守恒流称为能动张量,添加一项散度仍可满足连续性方程
$$T_c^{\mu\nu}=\frac{\partial\mathcal{L}}{\partial(\partial_\mu\Phi)}\partial^\nu\Phi-\eta^{\mu\nu}\mathcal{L}$$
$$T_B^{\mu\nu}=T_c^{\mu\nu}+\partial_\rho B^{\rho\mu\nu}\qquad B^{\rho\mu\nu}=-B^{\mu\rho\nu}$$
考虑Lorentz变换的守恒流
$$j^{\mu\nu\rho}=T_c^{\mu\nu}x^\rho-T_c^{\mu\rho}x^\nu+\frac{i}{2}\frac{\partial\mathcal{L}}{\partial(\partial_\mu\Phi)}S^{\nu\rho}\Phi$$
寻找 $B^{\mu\nu\rho}$ 使得 $j^{\mu\nu\rho}$ 具有以下形式,由连续性条件知 $T_B^{\mu\nu}$ 对称,称为Belinfante张量
$$j^{\mu\nu\rho}=T_B^{\mu\nu}x^\rho-T_B^{\mu\rho}x^\nu\qquad T_B^{\mu\nu}=T_B^{\nu\mu}$$
满足上述条件的 $B^{\mu\nu\rho}$ 并不唯一,可验证以下选择符合要求
$$B^{\mu\rho\nu}=\frac{i}{4}\left(\frac{\partial\mathcal{L}}{\partial(\partial_\mu\Phi)}S^{\nu\rho}\Phi+\frac{\partial\mathcal{L}}{\partial(\partial_\rho\Phi)}S^{\mu\nu}\Phi+\frac{\partial\mathcal{L}}{\partial(\partial_\nu\Phi)}S^{\mu\rho}\Phi\right)$$
$$T_c^{\rho\nu}-T_c^{\nu\rho}=-\partial_\mu\left(\frac{i}{2}\frac{\partial\mathcal{L}}{\partial(\partial_\mu\Phi)}S^{\nu\rho}\Phi\right)$$
$$\partial_\mu B^{\mu\rho\nu}-\partial_\mu B^{\mu\nu\rho}=\partial_\mu\left(\frac{i}{2}\frac{\partial\mathcal{L}}{\partial(\partial_\mu\Phi)}S^{\nu\rho}\Phi\right)$$
考虑缩放变换的守恒流
$$j_D^\mu=T^{\mu}_{c\;\nu}x^\nu+\frac{\partial\mathcal{L}}{\partial(\partial_\mu\Phi)}\Delta\Phi$$
同样可修改能动张量使其具有以下形式,并且仍保持 $T^{\mu\nu}$ 对称
$$j_D^\mu=T^{\mu}_{\;\;\nu}x^\nu\qquad T^{\mu}_{\;\;\mu}=0$$
因此下文中默认能动张量无迹且对称
关联函数
考虑以下关联函数,其中 $x_1^0\ge \cdots\ge x_n^0$ 已按时序排列,下文关联函数默认省略时序
$$\ev{\Phi(x_1)\cdots\Phi(x_n)}=\frac{1}{\mathcal{Z}}\int[\dd\Phi]\Phi(x_1)\cdots\Phi(x_n)\exp(-S[\Phi])$$
$$\mathcal{Z}=\int[\dd\Phi]\exp(-S[\Phi])$$
由对称变换下测度和作用量不变可得
$$\begin{aligned}\ev{\Phi(x’_1)\cdots\Phi(x’_n)}&=\frac{1}{\mathcal{Z}}\int[\dd\Phi]\Phi(x’_1)\cdots\Phi(x’_n)\exp(-S[\Phi])\\&=\frac{1}{\mathcal{Z}}\int[\dd\Phi’]\Phi’(x’_1)\cdots\Phi’(x’_n)\exp(-S[\Phi’])\\&=\frac{1}{\mathcal{Z}}\int[\dd\Phi]\mathcal{F}(\Phi(x_1))\cdots\mathcal{F}(\Phi(x_n))\exp(-S[\Phi])\\&=\ev{\mathcal{F}(\Phi(x_1))\cdots\mathcal{F}(\Phi(x_n))}\end{aligned}$$
由上式可知关联函数在平移、旋转与缩放下的变化形式
$$\ev{\Phi(x_1+a_1)\cdots\Phi(x_n+a_n)}=\ev{\Phi(x_1)\cdots\Phi(x_n)}$$
$$\ev{\Phi(\Lambda^\mu_{\;\;\nu}x_1^\nu)\cdots\Phi(\Lambda^\mu_{\;\;\nu}x_n^\nu)}=\ev{\Phi(x_1^\mu)\cdots\Phi(x_n^\mu)}$$
$$\ev{\phi_1(\lambda x_1)\cdots\phi_n(\lambda x_n)}=\lambda^{-\Delta_1}\cdots\lambda^{-\Delta_n}\ev{\phi_1(x_1)\cdots\phi_n(x_n)}$$
考虑两点关联函数,以上三种对称性要求其具有以下形式
$$\ev{\phi_1(x_1)\phi_2(x_2)}=\frac{C_{12}}{|x_1-x_2|^{\Delta_1+\Delta_2}}$$
另外还剩下一种共形变换,即SCT
$$x’^\mu=\frac{x^\mu-x^2 f^\mu}{1-2fx+f^2 x^2}$$
$$|x_i’-x_j’|=\frac{|x_i-x_j|}{(1-2fx_i+f^2x_i^2)^{1/2}(1-2fx_j+f^2x_j^2)^{1/2}}$$
由SCT对称性可定出缩放维数 $\Delta_1=\Delta_2$,即只有共形维数相等的准初级场是相关的
$$\frac{C_{12}}{|x_1-x_2|^{\Delta_1+\Delta_2}}=\frac{C_{12}}{\gamma_1^{\Delta_1}\gamma_2^{\Delta_2}}\frac{(\gamma_1\gamma_2)^{(\Delta_1+\Delta_2)/2}}{|x_1-x_2|^{\Delta_1+\Delta_2}}$$
$$\gamma_i=1-2fx_i+f^2x_i^2$$
综上两点关联函数具有以下形式
$$\ev{\phi_1(x_1)\phi_2(x_2)}=\left\lbrace\begin{array}.\displaystyle\frac{C_{12}}{|x_1-x_2|^{2\Delta_1}}& \text{if} & \Delta_1=\Delta_2\\ 0&\text{if}&\Delta_1\neq\Delta_2\end{array}\right.$$
同样可以定出三点及四点关联函数的形式,构造四点关联函数利用了交比共形不变的性质,其中 $x_{ij}=|x_i-x_j|$,$\Delta=\sum_{i=1}^4\Delta_i$
$$\ev{\phi_1(x_1)\phi_2(x_2)\phi_3(x_3)}=\frac{C_{123}}{x_{12}^{\Delta_1+\Delta_2-\Delta_3}x_{23}^{\Delta_2+\Delta_3-\Delta_1}x_{13}^{\Delta_3+\Delta_1-\Delta_2}}$$
$$\ev{\phi_1(x_1)\cdots\phi_4(x_4)}=f\left(\frac{x_{12}x_{34}}{x_{13}x_{24}},\frac{x_{12}x_{34}}{x_{23}x_{14}}\right)\prod_{i<j}^4 x_{ij}^{\Delta/3-\Delta_i-\Delta_j}$$
Ward-Takahashi等式
Noether定理的量子化版本称为Ward-Takahashi等式,考虑无穷小对称场变换
$$\Phi’(x)=\Phi(x)-i\omega_aG_a\Phi(x)$$
记 $X=\Phi(x_1)\cdots\Phi(x_n)$ ,已按时序排列,由关联函数不变可知
$$\ev{X}=\frac{1}{\mathcal{Z}}\int [\dd \Phi’](X+\delta X)\exp\left(-S[\Phi]-\int\text{d}^d x\;\omega_a(x)\partial_\mu j^\mu_a\right)$$
$$\ev{\delta X}=\int\text{d}^d x\;\omega_a(x)\partial_\mu \ev{j^\mu_a X}$$
由无穷小变换写出 $\delta X$ 具体形式
$$\begin{aligned}\delta X&=-i\sum_{i=1}^n\omega_a(x_i)\Phi(x_1)\cdots G_a\Phi(x_i)\cdots\Phi(x_n)\\&=-i\int\text{d}^d x\;\omega_a(x)\sum_{i=1}^n\delta(x-x_i)\Phi(x_1)\cdots G_a\Phi(x_i)\cdots\Phi(x_n)\end{aligned}$$
比较即得Ward-Takahashi等式
$$\frac{\partial}{\partial x^\mu} \ev{j^\mu_a(x) \Phi(x_1)\cdots\Phi(x_n)}=-i\sum_{i=1}^n\delta(x-x_i)\ev{\Phi(x_1)\cdots G_a\Phi(x_i)\cdots\Phi(x_n)}$$
令 $t=x_1^0$,$Y=\Phi(x_2)\cdots\Phi(x_n)$,对Ward-Takahashi等式在狭窄区间 $x^0\in[t_-,t_+],\quad t_-< t< t_+$ 内积分可得
$$\ev{Q_a(t_+)\Phi(x_1)Y}-\ev{Q_a(t_-)\Phi(x_1)Y}=-i\ev{G_a\Phi(x_1)Y}$$
取极限 $t_-\to t_+$,由 $Y$ 的任意性可知守恒荷即为无穷小对称变换生成元
$$[Q_a,\Phi]=-iG_a\Phi$$
共形Ward-Takahashi等式
对无穷小共形变换 $x^\mu\mapsto x^\mu+\epsilon^\mu(x)$,守恒流可写为以下形式,其中 $T^{\mu\nu}$ 对称无迹
$$j_a^\mu(x)\omega_a=T^{\mu\nu}\epsilon_\nu$$
代入ward-Takahashi等式可得下式,其中区域 $D$ 包含 $X$ 中所有场的坐标
$$\ev{\delta_\epsilon X}=\int_D\text{d}^2 x\;\partial_\mu\ev{T^{\mu\nu}(x)\epsilon_\nu(x)X}$$
将二维能动张量用复变量表示,$\displaystyle T_{\alpha\beta}=\frac{\partial x^\mu}{\partial x^\alpha}\frac{\partial x^\nu}{\partial x^\beta}T_{\mu\nu}$,其中$x^\alpha$对应$(z,\bar z)$,利用能动张量对称性,迹和散度为零可得
$$T_{zz}=\frac{1}{2}(T_{00}-iT_{10})\qquad T_{\bar z\bar z}=\frac{1}{2}(T_{00}+iT_{10})\qquad T_{z\bar z}=T_{\bar z z}=0$$
$$\partial_{\bar z} T_{zz}=0\qquad\partial_z T_{\bar z\bar z}=0$$
因此二维能动张量仅有全纯和共轭全纯两个非零分量,可定义
$$T(z)=-2\pi T_{zz}(z,\bar z)\qquad \overline T(\bar z)=-2\pi T_{\bar z\bar z}(z,\bar z)$$
$$\epsilon(z)=\epsilon^z\qquad\bar\epsilon(\bar z)=\epsilon^{\bar z}\qquad\epsilon^z=\epsilon^0+i\epsilon^1\qquad\epsilon^{\bar z}=\epsilon^0-i\epsilon^1$$
选取Euclidean度规,由Stokes公式可得以下共形Ward-Takahashi等式,其中积分路径 $C=\partial D$ 包含 $X$ 中所有场的坐标
$$\ev{\delta_{\epsilon,\bar\epsilon}X}=\frac{1}{2\pi i}\oint_C\left[-\text{d} z\;\epsilon(z)\ev{T(z)X}+\text{d} \bar z\;\bar\epsilon(\bar z)\ev{\overline T(\bar z)X}\right]$$
径序量子化
注意上文推导中关联函数均默认省略了时序,其不利于共形Ward-Takahashi等式中环路积分的计算,因此作以下映射,使得时序 $\mathcal{T}$ 成为径序 $\mathcal{R}$
$$w=e^z=e^{x^0+ix^1}$$
将共形Ward-Takahashi等式中 $z$ 替换为 $w$
$$\dd z=\frac{\dd w}{w}\qquad \epsilon(z)=\frac{\epsilon(w)}{w}\qquad T(z)=w^2T(w)$$
共形Ward-Takahashi等式形式在此映射下仍保持不变
$$\ev{\delta_{\epsilon,\bar\epsilon}X}=\frac{1}{2\pi i}\oint_C\left[-\text{d} w\;\epsilon(w)\ev{T(w)X}+\text{d} \bar w\;\bar\epsilon(\bar w)\ev{\overline T(\bar w)X}\right]$$
注意在 $w=e^z$ 映射下环路积分圈数可能增多,故应限制初始Euclidean平面上坐标$(x^0=it,x^1)$范围,使得 $x^1\in [0,2\pi]$,让平面卷曲成无穷长圆柱,即可保证共形Ward-Takahashi等式中环路积分圈数仍为$1$
Hamiltonian与动量算符分别为时间平移与空间平移的生成元,在此映射下成为 $w$ 平面上缩放与旋转的生成元,可将其写为
$$H=L_0+\overline L_0\qquad P=i(L_0-\overline L_0)$$
另外在径序下算符对易子可表示为环路积分的形式
$$[A,B]=\oint_{C(0)}\dd w\oint_{C(w)}\dd z\;a(z)b(w)$$
$$A=\oint_C a(z)\dd z\quad B=\oint_C b(w)\dd w$$
由Euclidean平面守恒荷 $Q_a=\displaystyle\int \text{d} x^1\;j_a^0\quad x^0=\text{const.}$,$\omega_a j_a^\mu=T^{\mu\nu}\epsilon_\nu$,可推出径序复平面中守恒荷表达式
$$Q_{\epsilon,\bar\epsilon}=\frac{1}{2\pi i}\oint_C\left(\text{d} w \;T(w)\epsilon(w)-\text{d} \bar w\;\overline T(\bar w)\bar \epsilon(\bar w)\right)$$
共形Ward-Takahashi等式可用守恒荷表示为以下形式
$$\delta_{\epsilon,\bar\epsilon}X=-[Q_{\epsilon,\bar\epsilon},X]$$
下文默认使用径序复平面,即 $z=e^{x^0+ix^1}$,$z$ 平面上原点对应负无穷时刻;下文关联函数与OPE均默认省略径序
初级场
若场 $\phi(z,\bar z)$ 在共形变换 $z\mapsto w(z)$ 下按以下规律变化,则称其为具有共形维数 $(h,\bar h)$ 的初级场;若此规律仅适用于 $w\in SL(2,\mathbb{C})/\mathbb{Z}_2$,则称其为准初级场
$$\phi’(w,\bar w)=\left(\frac{\text{d} w}{\text{d} z}\right)^{-h}\left(\frac{\text{d} \bar w}{\text{d} \bar z}\right)^{-\bar h}\phi(z,\bar z)$$
考虑初级场在无穷小共形变换 $w(z)=z+\epsilon(z)$ 下的变化规律,直接计算可得
$$\delta_{\epsilon,\bar\epsilon}\phi(z,\bar z)=-(h\partial_z\epsilon+\epsilon\partial_z+\bar h\partial_{\bar z}\bar\epsilon+\bar\epsilon\partial_{\bar z})\phi(z,\bar z)$$
将具有共形维数 $(h,\bar h)$ 的初级场在原点附近Laurent展开
$$\phi(z,\bar z)=\sum_{m,n\in\mathbb{Z}}z^{-m-h}\bar{z}^{-n-\bar h}\phi_{m,n}$$
将 $\phi(z,\bar z)$ 视为算符,要求其在原点处非奇异,入射态可写为
$$\ket{\phi_\text{in}}=\lim_{z,\bar z\to 0}\phi(z,\bar z)\ket{0}=\phi_{-h,-\bar h}\ket{0}$$
$$\phi_{m,n}\ket{0}=0\qquad(m>-h\;\;\text{or}\;\;n>-\bar h)$$
考虑 $\phi(z,\bar z)$ 的Hermitian共轭算符,由 $x^0=it$,取共轭,$(x^0,x^1)\mapsto (-x^0,x^1)$即 $z\mapsto 1/\bar z$,类比初级场的变化规律定义 $\phi^\dagger(z,\bar z)$
$$\phi^\dagger(z,\bar z)=\bar z^{-2h} z^{-2 \bar h}\phi(1/\bar z,1/z)$$
同样对 $\phi^\dagger(z,\bar z)$ Laurent展开
$$\phi^\dagger(z,\bar z)=\sum_{m,n\in\mathbb{Z}}\bar z^{m-h}z^{n-\bar h}\phi_{m,n}$$
对比 $\phi,\phi^\dagger$ 展开式可知
$$\phi^\dagger_{m, n}=\phi_{-m,- n}\qquad \bra{\phi_\text{out}}=\bra{0}\phi_{h,\bar h}=\ket{\phi_\text{in}}^\dagger$$
将入射态与出射态作内积
$$\begin{aligned}\bra{\phi_\text{out}}\ket{\phi_\text{in}}&=\lim_{z,\bar z,w,\bar w\to 0}\bra{0}\phi^\dagger(z,\bar z)\phi(w,\bar w)\ket{0}\\&=\lim_{z,\bar z,w,\bar w\to 0}\bar z^{-2h}z^{-2\bar h}\bra{0}\phi(1/\bar z,1/z)\phi(w,\bar w)\ket{0}\\&=\lim_{\xi,\bar \xi\to \infty}\bar \xi^{2h}\xi^{2\bar h}\bra{0}\phi(\bar \xi,\xi)\phi(0,0)\ket{0}\end{aligned}$$
算符乘积展开(OPE)
观察初级场变化规律,其中项可写为Chauchy积分形式
$$h\partial_w\epsilon(w)\phi(w,\bar w)=\frac{1}{2\pi i}\int_{C(w)}\dd z\;h\frac{\epsilon(z)}{(z-w)^2}\phi(w,\bar w)$$
$$\epsilon(w)\partial_w\phi(w,\bar w)=\frac{1}{2\pi i}\int_{C(w)}\dd z\;\frac{\epsilon(z)}{z-w}\partial_w\phi(w,\bar w)$$
考虑全纯初级场乘积 $X=\phi_1(w_1)\cdots\phi_n(w_n)$ ,与共形Ward-Takahashi等式比较可得下式,其中reg.表示全纯部分
$$\ev{T(z)X}=\sum_{i}\left(\frac{1}{z-w_i}\partial_{w_i}\ev{X}+\frac{h_i}{(z-w_i)^2}\ev{X}\right)+\text{reg.}$$
去掉 $\ev{}$ 与全纯部分,即得 $T(z)X$ 的算符乘积展开(OPE)
$$T(z)X\sim \sum_{i}\left(\frac{1}{z-w_i}\partial_{w_i}X+\frac{h_i}{(z-w_i)^2}X\right)$$
若 $X$ 由单个初级场 $\phi(w,\bar w)$ 构成,则下式可作为初级场的另一种定义
$$T(z)\phi(w,\bar w)\sim\frac{h}{(z-w)^2}\phi(w,\bar w)+\frac{1}{z-w}\partial_w\phi(w,\bar w)$$
$$\overline T(\bar z)\phi(w,\bar w)\sim\frac{\bar h}{(\bar z-\bar w)^2}\phi(w,\bar w)+\frac{1}{\bar z-\bar w}\partial_{\bar w}\phi(w,\bar w)$$
以下考虑两个全纯能动张量算符的OPE,先给出结果,再进行验证;共轭全纯算符的展开式类似,而 $T(z)\overline T(\bar w)$ 仅含非奇异项
$$T(z)T(w)\sim\frac{c/2}{(z-w)^4}+\frac{2T(w)}{(z-w)^2}+\frac{\partial_w T(w)}{z-w}$$
$$T(z)\overline T(\bar w)\sim 0$$
先将 $T(z)$ Laurent展开,计算守恒荷可知展开系数 $L_n$ 应为对应的 $\text{Vir}_c$ 中生成元
$$T(z)=\sum_{n\in\mathbb{Z}}z^{-n-2}L_n\qquad L_n=\frac{1}{2\pi i}\oint\dd z\;z^{n+1}T(z)$$
$$Q_\epsilon=\frac{1}{2\pi i}\oint\dd z\;T(z)\sum_{n\in\mathbb{Z}}\epsilon_n z^{n+1}=\sum_{n\in\mathbb{Z}}\epsilon_n L_n$$
只需验证 $TT$ OPE $L_n$ 满足Virasoro代数即可,以下计算说明上文OPE正确
$$\begin{aligned}\lbrack L_m,L_n\rbrack&=\oint\frac{\dd z}{2\pi i}\oint\frac{\dd w}{2\pi i}z^{m+1}w^{n+1}[T(z),T(w)]\\&=\oint_{C(0)}\frac{\dd w}{2\pi i}w^{n+1}\oint_{C(w)}\frac{\dd z}{2\pi i}z^{m+1}\mathcal{R}(T(z)T(w))\\&=\oint_{C(0)}\frac{\dd w}{2\pi i}w^{n+1}\oint_{C(w)}\frac{\dd z}{2\pi i}z^{m+1}\left(\frac{c/2}{(z-w)^4}+\frac{2T(w)}{(z-w)^2}+\frac{\partial_w T(w)}{z-w}\right)\\&=\oint_{C(0)}\frac{\dd w}{2\pi i}w^{n+1}\left(\frac{c(m+1)m(m-1)}{12}w^{m-2}+2(m+1)w^mT(w)+w^{m+1}\partial_w T(w)\right)\\&=(m-n)L_{m+n}+\frac{c}{12}(m^3-m)\delta_{m+n,0}\end{aligned}$$
$TT$ OPE说明仅在中心荷为零时 $T(z)$ 为初级场,考虑 $T(z)$ 的无穷小共形变换形式
$$\delta_\epsilon T(z)=\epsilon(z)\partial_z T(z)+2\partial_z\epsilon(z)T(z)+\frac{c}{12}\partial^3_z \epsilon(z)$$
在无穷小Mobius变换 $\epsilon(z)=\alpha+\beta z+\gamma z^2$ 下,$T(z)$ 满足共形维数为 $(2,0)$ 的准初级场变换规律,因而可将 $T(z)$ 在共形变换 $z\mapsto w(z)$ 下的变换规律写为以下形式
$$T’(w)=\left(\frac{\dd w}{\dd z}\right)^{-2} (T(z)-\frac{c}{12}\{w;z\})$$
其中 $\{w;z\}$ 称为Schwarzian导数,由共形变换结合律及 $T(z)$ 为准初级场及可知Schwarzian导数应满足以下性质
$$\{u;z\}=\left(\frac{\dd w}{\dd z}\right)^2 \{u;w\}+\{w;z\}$$
$$\{w;z\}=0\quad\text{for}\quad w\in SL(2,\mathbb{C})/\mathbb{Z}_2$$
结合无穷小共形变换形式可得Schwarzian导数具体形式
$$\{w;z\}=\frac{\text{d}^3 w/\text{d} z^3}{\text{d} w/\text{d} z}-\frac{3}{2}\left(\frac{\text{d}^2 w/\text{d} z^2}{\text{d} w/\text{d} z}\right)^2$$
实际上Schwarzian导数具有比第二条更强的性质
$$\{\Gamma w;z\}=\{w;z\}\quad\text{for}\quad \Gamma\in SL(2,\mathbb{C})/\mathbb{Z}_2$$
Verma Module
真空态 $\ket{0}$ 应在全局共形变换下保持不变,因此 $L_{-1},L_0,L_1$ 及其共轭全纯为真空湮灭算符;此性质也可从 $T(z)\ket{0},\overline T(\bar z)\ket{0}$ 在原点处良定义得出,其同时隐含能动张量真空期望为零的条件
$$L_n\ket{0}=0\quad \bar L_n\ket{0}=0\qquad\text{for}\quad n\ge -1$$
$$\bra{0}T(z)\ket{0}=\bra{0}\overline T(\bar z)\ket{0}=0$$
考虑初级场作用于真空得到的渐近态
$$\phi(0,0)\ket{0}=\ket{h,\bar h}$$
由 $\displaystyle [A,b(w)]=\oint \dd z\; a(z)b(w)$ 与 $T\phi$ OPE计算生成元与初级场的对易子
$$[L_n,\phi(w,\bar w)]=h(n+1)w^n\phi(w,\bar w)+w^{n+1}\partial_w\phi(w,\bar w)$$
$$[\bar L_n,\phi(w,\bar w)]=\bar h(n+1)\bar w^n\phi(w,\bar w)+\bar w^{n+1}\partial_w\phi(w,\bar w)$$
将对易子作用于真空可得 $L_n,\bar L_n$ 对渐近态的作用,由 $H=L_0+\bar L_0$ 可知 $\ket{h,\bar h}$ 为Hamiltonian本征态
$$L_0\ket{h,\bar h}=h\ket{h,\bar h}\qquad \bar L_0\ket{h,\bar h}=\bar h\ket{h,\bar h}$$
$$L_n\ket{h,\bar h}=0\quad \bar L_n\ket{h,\bar h}=0\qquad\text{for}\quad n> 0$$
假设对易子中初级场全纯,将其Laurent展开,可得以下对易关系
$$[L_n,\phi_m]=[n(h-1)-m]\phi_{n+m}$$
其中 $[L_0,\phi_m]=-m\phi_m$ 说明 $\phi_{-m}$ 为共形维数升 $m$ 算符;而由Virasoro代数可知,$L_{-m}\quad m>0$ 也为共形维数的升 $m$ 算符,具备更高共形维数的激发态可由一系列升算符作用于 $\ket{h}$ 得到
$$[L_0,L_{-m}]=mL_{-m}\qquad L_0L_{-m}\ket{h}=(h+m)L_{-m}\ket{h}$$
具有以下形式的激发态称为 $\ket{h}$ 的$N$级裔,升算符乱序作用于 $\ket{h}$ 得到的激发态均可由 $\ket{h}$ 裔的线性组合得到
$$L_{-k_1}L_{-k_2}\cdots L_{-k_n}\ket{h}\qquad 1\le k_1\le\cdots\le k_n$$
$$k_1+k_2+\cdots+k_n=N$$
$\ket{h}$ 及其裔在Virasoro代数下构成Hilbert空间的子空间,称为Verma Module