next up previous
Next: Interaction matrix elements (second Up: qrpa16w Previous: Conclusion


QRPA equation

The QRPA equations are the small-oscillations limit of the time-dependent Hartree-Fock-Bogoliubov approximation, see, e.g., [8,13]. In the canonical basis the most general equations take the form

\begin{displaymath}
\sum_{L<L^\prime}
\left(
\begin{array}{cc}
A_{KK^\prime,LL^\...
...ime}\\
{Y}^k_{KK^\prime}
\end{array}\right),
\ \ K<K^\prime ,
\end{displaymath} (10)


$\displaystyle A_{KK^\prime,LL^\prime}$ $\textstyle =$ $\displaystyle E_{KL}\delta_{K^\prime L^\prime}
-E_{K^\prime L}\delta_{K L^\prime}
-E_{KL^\prime}\delta_{K^\prime L}
+E_{K^\prime L^\prime}\delta_{KL}$  
    $\displaystyle -\bar{V}^{\rm ph}_{K \bar{L} \bar{K}^\prime L^\prime}u_{L^\prime}...
...V}^{\rm ph}_{K^\prime \bar{L} \bar{K} L^\prime}u_{L^\prime}v_L u_{K^\prime} v_K$  
    $\displaystyle +\bar{V}^{\rm ph}_{K \bar{L}^\prime \bar{K}^\prime L}u_{L}v_{L^\p...
...}^{\rm ph}_{K^\prime \bar{L}^\prime \bar{K} L}u_L v_{L^\prime} u_{K^\prime} v_K$  
    $\displaystyle -\bar{V}^{\rm pp}_{\bar{L} \bar{L}^\prime \bar{K}^\prime \bar{K}}...
...
-\bar{V}^{\rm pp}_{K K^\prime {L}^\prime L}u_K u_{K^\prime} u_{L} u_{L^\prime}$  
    $\displaystyle -\bar{V}^{\rm 3p1h}_{\bar{L} \bar{L}^\prime {K} \bar{K}^\prime}v_...
..._{\bar{L} \bar{L}^\prime {K}^\prime \bar{K}}v_L v_{L^\prime} u_{K^\prime} v_{K}$  
    $\displaystyle -\bar{V}^{\rm 3p1h}_{{K} {K}^\prime \bar{L} {L}^\prime}u_{K}u_{K^...
... 3p1h}_{{K} {K}^\prime \bar{L}^\prime {L}}u_{K} u_{K^\prime} u_{L} v_{L^\prime}$  
    $\displaystyle -\bar{V}^{\rm 1p3h}_{\bar{L} {L}^\prime \bar{K}^\prime \bar{K}}u_...
...\bar{L}^\prime {L} \bar{K}^\prime \bar{K}}u_{L} v_{L^\prime} v_{K^\prime} v_{K}$  
    $\displaystyle -\bar{V}^{\rm 1p3h}_{{K} \bar{K}^\prime {L}^\prime {L}}u_{K}v_{K^...
...1p3h}_{{K}^\prime \bar{K} {L}^\prime {L}}u_{K^\prime} v_{K} u_{L} u_{L^\prime},$ (11)


$\displaystyle B_{KK^\prime,LL^\prime}$ $\textstyle =$ $\displaystyle \bar{V}^{\rm ph}_{K^\prime L^\prime\bar{K} \bar{L}}u_{L^\prime}v_...
...\rm ph}_{K L^\prime \bar{K^\prime} \bar{L}}u_{L^\prime}v_{L} u_{K} v_{K^\prime}$  
    $\displaystyle -\bar{V}^{\rm ph}_{K^\prime L \bar{K} \bar{L}^\prime}u_{L}v_{L^\p...
...\rm ph}_{K L \bar{K}^\prime \bar{L}^\prime}u_{L}v_{L^\prime} u_{K} v_{K^\prime}$  
    $\displaystyle +\bar{V}^{\rm pp}_{K^\prime {K} \bar{L} \bar{L}^\prime}v_{L}v_{L^...
...m pp}_{L^\prime {L} \bar{K} \bar{K}^\prime}v_{K}v_{K^\prime} u_{L} u_{L^\prime}$  
    $\displaystyle +\bar{V}^{\rm 3p1h}_{K^\prime {K} {L^\prime} \bar{L}}u_{L^\prime}...
...\rm 3p1h}_{K^\prime {K} {L} \bar{L}^\prime}u_{L}v_{L^\prime} u_{K} u_{K^\prime}$  
    $\displaystyle +\bar{V}^{\rm 3p1h}_{L^\prime {L} {K^\prime} \bar{K}}u_{K^\prime}...
...\rm 3p1h}_{L^\prime {L} {K} \bar{K}^\prime}u_{K}v_{K^\prime} u_{L} u_{L^\prime}$  
    $\displaystyle +\bar{V}^{\rm 1p3h}_{K^\prime \bar{K} \bar{L} \bar{L}^\prime}v_{L...
...}_{K \bar{K}^\prime \bar{L} \bar{L}^\prime}v_{L}v_{L^\prime} u_{K} v_{K^\prime}$  
    $\displaystyle +\bar{V}^{\rm 1p3h}_{L^\prime \bar{L} \bar{K} \bar{K}^\prime}v_{K...
..._{L \bar{L}^\prime \bar{K} \bar{K}^\prime}v_{K}v_{K^\prime} u_{L} v_{L^\prime},$ (12)


\begin{displaymath}
\bar{V}^{\rm ph}_{KL K^\prime L^\prime}
= \frac{ \delta^2 E...
...ppa^\ast] }{ \delta\rho_{K^\prime
K}\delta\rho_{L^\prime L} },
\end{displaymath} (13)


\begin{displaymath}
\bar{V}^{\rm pp}_{K^\prime K L^\prime L}
= \frac{ \delta^2 ...
...{ \delta\kappa^\ast_{K^\prime K}
\delta\kappa_{L^\prime L} },
\end{displaymath} (14)


\begin{displaymath}
\bar{V}^{\rm 3p1h}_{K^\prime K L^\prime L}
= \frac{ \delta^...
...\prime} }
={\bar{V}^{\rm 1p3h\ \ast}}_{L L^\prime K^\prime K},
\end{displaymath} (15)

where $K$ and $L$ are single-particle indices for the canonical basis, and the states are assumed to be ordered. The symbol $\bar{K}$ refers to the conjugate partner of $K$, $u_K$ and $v_K$ come from the BCS transformation associated with the canonical basis, and the $E_{KL}$ are the one-quasiparticle matrix elements of the HFB Hamiltonian (cf. Eq. (4.14b) of Ref. [60]). $X^k_{LL^\prime}$ and $Y^k_{LL^\prime}$ are the forward and backward amplitudes of the QRPA solution $k$, and $E_k$ is the corresponding excitation energy. $E[\rho,\kappa,\kappa^\ast]$ is the energy functional (see App. B for an explicit definition) and $\rho$ and $\kappa$ are the density matrix and pairing tensor. After taking the functional derivatives, we replace $\rho$ and $\kappa$ by their HFB solutions, in complete analogy with an ordinary Taylor-series expansion.

To write the equations in coupled form, we introduce the notation

\begin{displaymath}
K \equiv (n_\mu l_\mu j_\mu m_\mu) \equiv (\mu m_\mu), \ \
L \equiv (\nu m_\nu), \ \
\end{displaymath} (16)

where $(nljm)$ denote spherical quantum numbers. Using (i) rotational, time-reversal, and parity symmetries of the HFB state, (ii) the conjugate single-particle state2


\begin{displaymath}
\vert\overline{K}\rangle =
\vert\overline{\mu m_\mu}\rangle = (-)^{j_\mu-m_\mu}\vert\mu\: -m_\mu\rangle ,
\end{displaymath} (17)

and (iii) the relations
    $\displaystyle X^k_{KK^\prime} = \langle j_\mu m_\mu j_{\mu^\prime} m_{\mu^\prim...
...array}{cc}\sqrt{2}, &\mu=\mu^\prime,\\  1,
&\mbox{otherwise},\end{array}\right.$ (18)
    $\displaystyle Y^k_{KK^\prime} = (-)^{j_\mu-m_\mu}(-)^{j_{\mu^\prime}-m_{\mu^\pr...
...array}{cc}\sqrt{2}, &\mu=\mu^\prime,\\  1,
&\mbox{otherwise},\end{array}\right.$ (19)

with $J_k$ the angular momentum of the state $k$ and the factor $\sqrt{2}$ for convenience [12], one can rewrite the QRPA equation as


\begin{displaymath}
\sum_{\nu\leq\nu^\prime}
\left(
\begin{array}{cc}
A_{[\mu\mu...
...mu\mu^\prime]J_k}
\end{array}\right),\ \ \mu \leq \mu^\prime ,
\end{displaymath} (20)


$\displaystyle A_{[\mu\mu^\prime]J_k,[\nu\nu^\prime]J_k}$ $\textstyle =$ $\displaystyle \frac{1}{\sqrt{1+\delta_{\mu\mu^\prime}}}
\frac{1}{\sqrt{1+\delta...
...E_{\mu^\prime\nu}\:\delta_{\mu\nu^\prime}(-)^{j_\mu+j_{\mu^\prime}-J_k}
\right.$  
    $\displaystyle -E_{\mu\nu^\prime}\:\delta_{\mu^\prime\nu}(-)^{j_\mu+j_{\mu^\prime}-J_k}
+E_{\mu^\prime\nu^\prime}\:\delta_{\mu\nu}$  
    $\displaystyle +G(\mu\mu^\prime\nu\nu^\prime;J_k)
(u_{\mu^\prime} u_\mu u_\nu u_{\nu^\prime}
+v_{\nu} v_{\nu^\prime} v_{\mu^\prime} v_{\mu})$  
    $\displaystyle +F(\mu\mu^\prime\nu\nu^\prime;J_k)
(u_{\mu} v_{\nu^\prime} u_\nu v_{\mu^\prime}
+u_{\mu^\prime} v_{\nu} u_{\nu^\prime} v_{\mu})$  
    $\displaystyle -(-)^{ j_{\nu^\prime}+j_\nu-J_k }F(\mu\mu^\prime\nu^\prime\nu;J_k...
...} u_{\nu^\prime} v_{\mu^\prime}
+u_{\mu^\prime} v_{\nu^\prime} u_{\nu} v_{\mu})$  
    $\displaystyle -H(\mu\mu'\nu\nu';J_k)( v_\nu v_{\nu'} u_\mu v_{\mu'} +u_{\mu'} v_\mu u_\nu u_{\nu'} )$  
    $\displaystyle +(-)^{j_\mu+j_{\mu'}-J_k} H(\mu'\mu\nu\nu';J_k)
(v_\nu v_{\nu'} u_{\mu'} v_\mu + u_\mu v_{\mu'} u_\nu u_{\nu'} )$  
    $\displaystyle -H^\ast(\nu\nu'\mu\mu';J_k)( u_\mu u_{\mu'} u_{\nu'} v_\nu
+ u_\nu v_{\nu'} v_{\mu'} v_\mu)$  
    $\displaystyle +(-)^{j_\nu+j_{\nu'}-J_k}H^\ast(\nu'\nu\mu\mu';J_k)
(u_\mu u_{\mu'} u_\nu v_{\nu'} + u_{\nu'} v_\nu v_{\mu'} v_\mu)
\left.\right\},$ (21)


$\displaystyle B_{[\mu\mu^\prime]J_k,[\bar{\nu}\bar{\nu}^\prime]J_k}$ $\textstyle =$ $\displaystyle \frac{1}{\sqrt{1+\delta_{\mu\mu^\prime}}}
\frac{1}{\sqrt{1+\delta...
...\mu v_\nu v_{\nu^\prime}
+u_{\nu} u_{\nu^\prime} v_{\mu^\prime} v_{\mu})\right.$  
    $\displaystyle -(-)^{ j_\nu+j_{\nu^\prime}-J_k }F(\mu\mu^\prime\nu^\prime\nu;J_k...
...} v_{\mu^\prime} v_{\nu^\prime}
+u_{\mu^\prime} u_{\nu^\prime} v_{\nu} v_{\mu})$  
    $\displaystyle +(-)^{ j_\nu+j_{\nu^\prime}+j_\mu+j_{\mu^\prime}
}F(\mu^\prime\mu...
... u_{\nu} v_{\nu^\prime} v_{\mu}
+u_{\mu} u_{\nu^\prime} v_{\mu^\prime} v_{\nu})$  
    $\displaystyle +H(\mu\mu'\nu\nu';J_k)(v_\nu v_{\nu'} u_{\mu'} v_\mu + u_\mu v_{\mu'} u_\nu u_{\nu'})$  
    $\displaystyle -(-)^{j_\mu+j_{\mu'}-J_k} H(\mu'\mu\nu\nu';J_k)
(v_\nu v_{\nu'} u_\mu v_{\mu'} + u_{\mu'} v_\mu u_\nu u_{\nu'})$  
    $\displaystyle -H^\ast(\nu\nu'\mu\mu';J_k)
(v_\mu v_{\mu'} u_{\nu'} v_\nu + u_\nu v_{\nu'} u_\mu u_{\mu'})$  
    $\displaystyle +(-)^{j_\nu + j_{\nu'}-J_k}H^\ast(\nu'\nu\mu\mu';J_k)
(v_\mu v_{\mu'} u_\nu v_{\nu'} +u_{\nu'} v_\nu u_\mu u_{\mu'})
\left.\right\},$ (22)


$\displaystyle G(\mu\mu^\prime\nu\nu^\prime;J_k)$ $\textstyle =$ $\displaystyle \sum_{ m_\mu m_{\mu^\prime}m_\nu
m_{\nu^\prime} }
\langle j_\mu m...
...e} m_{\nu^\prime} \vert J_k M_k\rangle
\bar{V}^{\rm pp}_{K K^\prime L L^\prime}$  
  $\textstyle \equiv$ $\displaystyle \langle [\mu\mu^\prime]J_k \vert \bar{V}^{\rm pp}
\vert[\nu\nu^\prime]J_k\rangle ,$ (23)
$\displaystyle F(\mu\mu^\prime\nu\nu^\prime;J_k)$ $\textstyle =$ $\displaystyle \sum_{ m_\mu m_{\mu^\prime}m_\nu
m_{\nu^\prime} }
\langle j_\mu m...
...^\prime} \vert J_k M_k\rangle
\bar{V}^{\rm ph}_{K\bar{L^\prime}\bar{K^\prime}L}$  
  $\textstyle =$ $\displaystyle \sum_{J^\prime} (-)^{j_{ \mu^\prime}+j_\nu +J^\prime }
\left\{
\b...
...e [\mu\nu^\prime]J_k \vert \bar{V}^{\rm ph} \vert [\mu^\prime
\nu]J_k \rangle ,$ (24)
$\displaystyle H(\mu\mu'\nu\nu';J_k)$ $\textstyle =$ $\displaystyle \sum_{m_\mu m_{\mu'} m_\nu m_{\nu'}}
\langle j_\mu m_\mu j_{\mu'}...
...\nu'} m_{\nu'} \vert JM \rangle
\bar{V}^{\rm 3p1h}_{\bar{L}\bar{L}' K \bar{K}'}$  
  $\textstyle =$ $\displaystyle \sum_{J^\prime} (-)^{j_{ \mu}+j_\nu + 1 + l_{\nu^\prime} -J_k - J...
... [\mu\nu]J_k \vert \bar{V}^{\rm 3p1h} \vert [\mu^\prime
\nu^\prime]J_k \rangle.$  

We have represented the second derivatives of the energy functional $E[\rho,\kappa,\kappa^\ast]$ as unsymmetrized matrix elements of effective interactions $\bar{V}^{\rm pp}$, $\bar{V}^{\rm
ph}$, and $\bar{V}^{\rm 3p1h}$. These effective interactions are given in App. B. Although the ``matrix elements'' are unsymmetrized, the underlying two-quasiparticle states are of course antisymmetric. As a consequence, $A_{[\mu\mu^\prime]J_k,[\nu\nu^\prime]J_k}=
B_{[\mu\mu^\prime]J_k,[ \bar{\nu} \bar{\nu}^\prime ]J_k}=0$ if $J_k$ is odd and either $\mu=\mu^\prime$ or $\nu=\nu^\prime$.

The nuclear energy functional $E[\rho,\kappa,\kappa^\ast]$ is usually separated into particle-hole (ph) and pairing pieces (again, see App. B for explicit expressions). If the pairing functional, which we will call $E_{\rm pair}[\rho,\kappa,\kappa^\ast]$, depends on $\rho$ then the derivatives of $E_{\rm pair}[\rho,\kappa,\kappa^\ast]$ with respect to $\rho_{KK'}$ are called pairing-rearrangement terms [13]. In the QRPA, two kinds of pairing-rearrangement terms can arise in general. One has particle-hole character and is included in $\bar{V}^{\rm ph}_{KLK'L'}$; the other affects 3-particle-1-hole (3p1h) and 1-particle-3-hole (1p3h) configurations and is represented by $\bar{V}^{\rm 3p1h}_{K'KL'L}$ and $\bar{V}^{\rm 1p3h}_{K'KL'L}$. If the $\rho$-dependence of $E_{\rm pair}[\rho,\kappa,\kappa^\ast]$ is linear, then the ph-type pairing-rearrangement term does not appear. Furthermore, the 3p1h and 1p3h pairing-rearrangement terms arise only for $J^{\pi}=0^+$ modes if the HFB state has $J=0$. Most existing work uses a pairing functional that is linear in $\rho$, and so needs no pairing-rearrangement terms in $J^{\pi} \neq 0^+$ channels.


next up previous
Next: Interaction matrix elements (second Up: qrpa16w Previous: Conclusion
Jacek Dobaczewski 2004-07-29