@@ -1435,6 +1435,347 @@ \subsubsection{9.8 Symmetric limit for identical
14351435symmetric identical-system limit, that is precisely the condition for
14361436destructive interference between the \( bb\) and \( dd\) virtual pathways.
14371437
1438+ \subsubsection {9.9 Solving for the actual low-energy
1439+ roots }\label {solving-for-the-actual-low-energy-roots }
1440+
1441+ We now keep the symmetric identical-system assumptions of Section 9.8
1442+ and go one step further: instead of treating \( E\) as a common external
1443+ parameter, we solve the Bloch-Horowitz self-consistency equations for
1444+ the actual low-energy roots.
1445+
1446+ For this subsection, take
1447+
1448+ \[
1449+ E_0=0,
1450+ \qquad
1451+ E_\chi =E_0=0,
1452+ \label {eq:E0andEchiZero }
1453+ \]
1454+
1455+ and define
1456+
1457+ \[
1458+ E_\pi \equiv \omega _\pi .
1459+ \label {eq:EpiDef }
1460+ \]
1461+
1462+ Then Eq. \( \ref {eq:commonDeltaEdef }\) reduces to
1463+
1464+ \[
1465+ \Delta (E)=E_\pi -E.
1466+ \label {eq:DeltaSimple }
1467+ \]
1468+
1469+ It is also convenient to absorb the symmetric coupling into
1470+
1471+ \[
1472+ \Omega ^2 \equiv 2W^2.
1473+ \label {eq:OmegaDefLast }
1474+ \]
1475+
1476+ With this notation, the diagonal entries of Eq.
1477+ \( \ref {eq:HABdiagIdentical }\) become
1478+
1479+ \[
1480+ \lambda _{bb}(E)=-\frac {2\Omega ^2}{E_\pi -E},
1481+ \label {eq:lambdabbSimple }
1482+ \]
1483+
1484+ \[
1485+ \lambda _{bd}(E)=\lambda _{db}(E)=-\frac {\Omega ^2}{E_\pi -E},
1486+ \label {eq:lambdabdSimple }
1487+ \]
1488+
1489+ \[
1490+ \lambda _{dd}(E)=0.
1491+ \label {eq:lambdaddSimple }
1492+ \]
1493+
1494+ The physical dressed energies are found from the self-consistency
1495+ conditions
1496+
1497+ \[
1498+ E_{bb}=\lambda _{bb}(E_{bb}),
1499+ \qquad
1500+ E_{bd}=E_{db}=\lambda _{bd}(E_{bd}),
1501+ \qquad
1502+ E_{dd}=0.
1503+ \label {eq:selfConsLast }
1504+ \]
1505+
1506+ \paragraph {Exact quadratic equations }\label {exact-quadratic-equations }
1507+
1508+ For the singly-bright states \( bd\) and \( db\) ,
1509+
1510+ \[
1511+ E_{bd}=-\frac {\Omega ^2}{E_\pi -E_{bd}},
1512+ \label {eq:EbdImplicit }
1513+ \]
1514+
1515+ so
1516+
1517+ \[
1518+ E_{bd}(E_\pi -E_{bd})=-\Omega ^2,
1519+ \label {eq:EbdQuadraticStep }
1520+ \]
1521+
1522+ and therefore
1523+
1524+ \[
1525+ E_{bd}^2-E_\pi E_{bd}-\Omega ^2=0.
1526+ \label {eq:EbdQuadratic }
1527+ \]
1528+
1529+ Thus
1530+
1531+ \[
1532+ E_{bd}^{(\pm )}=E_{db}^{(\pm )}=\frac {E_\pi\pm\sqrt {E_\pi ^2+4\Omega ^2}}{2}.
1533+ \label {eq:EbdRoots }
1534+ \]
1535+
1536+ For the doubly-bright state \( bb\) ,
1537+
1538+ \[
1539+ E_{bb}=-\frac {2\Omega ^2}{E_\pi -E_{bb}},
1540+ \label {eq:EbbImplicit }
1541+ \]
1542+
1543+ so
1544+
1545+ \[
1546+ E_{bb}(E_\pi -E_{bb})=-2\Omega ^2,
1547+ \label {eq:EbbQuadraticStep }
1548+ \]
1549+
1550+ and hence
1551+
1552+ \[
1553+ E_{bb}^2-E_\pi E_{bb}-2\Omega ^2=0.
1554+ \label {eq:EbbQuadratic }
1555+ \]
1556+
1557+ Therefore
1558+
1559+ \[
1560+ E_{bb}^{(\pm )}=\frac {E_\pi\pm\sqrt {E_\pi ^2+8\Omega ^2}}{2}.
1561+ \label {eq:EbbRoots }
1562+ \]
1563+
1564+ Finally,
1565+
1566+ \[
1567+ E_{dd}=0.
1568+ \label {eq:EddExactLast }
1569+ \]
1570+
1571+ The plus roots in Eqs. \( \ref {eq:EbdRoots }\) and \( \ref {eq:EbbRoots }\)
1572+ sit near the pion scale and are not the low-energy branches we want. The
1573+ relevant low-energy solutions are therefore
1574+
1575+ \[
1576+ E_{bd}^{\rm low}=E_{db}^{\rm low}=\frac {E_\pi -\sqrt {E_\pi ^2+4\Omega ^2}}{2},
1577+ \label {eq:EbdLow }
1578+ \]
1579+
1580+ \[
1581+ E_{bb}^{\rm low}=\frac {E_\pi -\sqrt {E_\pi ^2+8\Omega ^2}}{2},
1582+ \label {eq:EbbLow }
1583+ \]
1584+
1585+ \[
1586+ E_{dd}=0.
1587+ \label {eq:EddLow }
1588+ \]
1589+
1590+ \paragraph {Small-parameter expansion }\label {small-parameter-expansion }
1591+
1592+ Define the small parameter
1593+
1594+ \[
1595+ x\equiv \frac {\Omega ^2}{E_\pi ^2}\ll 1.
1596+ \label {eq:xSmall }
1597+ \]
1598+
1599+ Using
1600+
1601+ \[
1602+ \sqrt {1+u}=1+\frac {u}{2}-\frac {u^2}{8}+O(u^3),
1603+ \label {eq:sqrtExpand }
1604+ \]
1605+
1606+ we expand Eq. \( \ref {eq:EbdLow }\) as
1607+
1608+ \[
1609+ E_{bd}^{\rm low}
1610+ =
1611+ \frac {E_\pi -E_\pi\sqrt {1+4x}}{2}
1612+ =
1613+ -E_\pi x+E_\pi x^2+O(E_\pi x^3),
1614+ \label {eq:EbdExpandStep }
1615+ \]
1616+
1617+ so
1618+
1619+ \[
1620+ E_{bd}^{\rm low}=E_{db}^{\rm low}
1621+ =
1622+ -\frac {\Omega ^2}{E_\pi }
1623+ +
1624+ \frac {\Omega ^4}{E_\pi ^3}
1625+ +
1626+ O\! \left (\frac {\Omega ^6}{E_\pi ^5}\right ).
1627+ \label {eq:EbdExpand }
1628+ \]
1629+
1630+ Similarly, Eq. \( \ref {eq:EbbLow }\) gives
1631+
1632+ \[
1633+ E_{bb}^{\rm low}
1634+ =
1635+ \frac {E_\pi -E_\pi\sqrt {1+8x}}{2}
1636+ =
1637+ -2E_\pi x+4E_\pi x^2+O(E_\pi x^3),
1638+ \label {eq:EbbExpandStep }
1639+ \]
1640+
1641+ so
1642+
1643+ \[
1644+ E_{bb}^{\rm low}
1645+ =
1646+ -\frac {2\Omega ^2}{E_\pi }
1647+ +
1648+ \frac {4\Omega ^4}{E_\pi ^3}
1649+ +
1650+ O\! \left (\frac {\Omega ^6}{E_\pi ^5}\right ).
1651+ \label {eq:EbbExpand }
1652+ \]
1653+
1654+ And of course
1655+
1656+ \[
1657+ E_{dd}=0.
1658+ \label {eq:EddExpand }
1659+ \]
1660+
1661+ \paragraph {Energy differences }\label {energy-differences }
1662+
1663+ Now compare the singly-bright state to its neighbours above and below.
1664+
1665+ The exact spacings are
1666+
1667+ \[
1668+ E_{dd}-E_{bd}^{\rm low}
1669+ =
1670+ \frac {\sqrt {E_\pi ^2+4\Omega ^2}-E_\pi }{2},
1671+ \label {eq:ddMinusBdExact }
1672+ \]
1673+
1674+ and
1675+
1676+ \[
1677+ E_{bd}^{\rm low}-E_{bb}^{\rm low}
1678+ =
1679+ \frac {\sqrt {E_\pi ^2+8\Omega ^2}-\sqrt {E_\pi ^2+4\Omega ^2}}{2}.
1680+ \label {eq:bdMinusBbExact }
1681+ \]
1682+
1683+ Expanding these to the same order gives
1684+
1685+ \[
1686+ E_{dd}-E_{bd}^{\rm low}
1687+ =
1688+ \frac {\Omega ^2}{E_\pi }
1689+ -
1690+ \frac {\Omega ^4}{E_\pi ^3}
1691+ +
1692+ O\! \left (\frac {\Omega ^6}{E_\pi ^5}\right ),
1693+ \label {eq:ddMinusBdExpand }
1694+ \]
1695+
1696+ \[
1697+ E_{bd}^{\rm low}-E_{bb}^{\rm low}
1698+ =
1699+ \frac {\Omega ^2}{E_\pi }
1700+ -
1701+ 3\frac {\Omega ^4}{E_\pi ^3}
1702+ +
1703+ O\! \left (\frac {\Omega ^6}{E_\pi ^5}\right ).
1704+ \label {eq:bdMinusBbExpand }
1705+ \]
1706+
1707+ So the two spacings are equal at leading order but differ at the next
1708+ order:
1709+
1710+ \[
1711+ \big (E_{dd}-E_{bd}^{\rm low}\big )-\big (E_{bd}^{\rm low}-E_{bb}^{\rm low}\big )
1712+ =
1713+ \frac {2\Omega ^4}{E_\pi ^3}
1714+ +
1715+ O\! \left (\frac {\Omega ^6}{E_\pi ^5}\right ).
1716+ \label {eq:asymmetryExpand }
1717+ \]
1718+
1719+ Equivalently, the midpoint shift is
1720+
1721+ \[
1722+ E_{bd}^{\rm low}-\frac {E_{bb}^{\rm low}+E_{dd}}{2}
1723+ =
1724+ -\frac {\Omega ^4}{E_\pi ^3}
1725+ +
1726+ O\! \left (\frac {\Omega ^6}{E_\pi ^5}\right ).
1727+ \label {eq:midpointShift }
1728+ \]
1729+
1730+ So the \( bd/db\) pair lies slightly \textbf {below } the exact midpoint
1731+ between \( bb\) and \( dd\) once the self-consistent \( E\) -dependence is
1732+ kept.
1733+
1734+ \paragraph {\texorpdfstring {Writing the asymmetry in terms of
1735+ \( E_{bd}\) }{Writing the asymmetry in terms of E\_\{ bd\} } }\label {writing-the-asymmetry-in-terms-of-e_bd }
1736+
1737+ Introduce the leading-order singly-bright energy
1738+
1739+ \[
1740+ E_{bd}^{(0)}\equiv -\frac {\Omega ^2}{E_\pi }.
1741+ \label {eq:Ebd0 }
1742+ \]
1743+
1744+ Then
1745+
1746+ \[
1747+ \frac {\Omega ^2}{E_\pi }\frac {E_{bd}^{(0)}}{E_\pi }
1748+ =
1749+ -\frac {\Omega ^4}{E_\pi ^3}.
1750+ \label {eq:Ebd0Identity }
1751+ \]
1752+
1753+ So Eq. \( \ref {eq:midpointShift }\) can be written as
1754+
1755+ \[
1756+ E_{bd}^{\rm low}-\frac {E_{bb}^{\rm low}+E_{dd}}{2}
1757+ =
1758+ \frac {\Omega ^2}{E_\pi }\frac {E_{bd}^{(0)}}{E_\pi }
1759+ +
1760+ O\! \left (\frac {\Omega ^6}{E_\pi ^5}\right )
1761+ \label {eq:midpointShiftEbd0 }
1762+ \]
1763+
1764+ Likewise Eq. \( \ref {eq:asymmetryExpand }\) becomes
1765+
1766+ \[
1767+ \big (E_{dd}-E_{bd}^{\rm low}\big )-\big (E_{bd}^{\rm low}-E_{bb}^{\rm low}\big )
1768+ =
1769+ -2\, \frac {\Omega ^2}{E_\pi }\frac {E_{bd}^{(0)}}{E_\pi }
1770+ +
1771+ O\! \left (\frac {\Omega ^6}{E_\pi ^5}\right ).
1772+ \label {eq:asymmetryEbd0 }
1773+ \]
1774+
1775+ To the order kept here, one may replace \( E_{bd}^{(0)}\) on the
1776+ right-hand sides by the full low-energy \( E_{bd}^{\rm low}\) , because
1777+ the difference between them only feeds in at \( O(\Omega ^6/E_\pi ^5)\) .
1778+
14381779\printbibliography
14391780
14401781
0 commit comments