1: \documentclass[genTeX]{nrc1}

2: 

3: \usepackage{cite} 

4: \usepackage{amsmath} 

5: \usepackage{amsfonts}

6: \usepackage{bm} 

7: \usepackage{bbm} 

8: \usepackage{epsf}

9: \usepackage{times}

10: %\usepackage{showkeys}

11: \usepackage{nicefrac}

12: \usepackage{graphicx}

13: 

14: \def\buildrel#1\under#2{\mathrel{\mathop{\kern0pt #2}\limits_{#1}}}

15: \def\Res#1{{\buildrel {\scriptstyle #1} \under {\textstyle \rm Res}}\,}

16: \def\bbox#1{\mbox{\boldmath$#1$}}

17: \def\bm#1{\mbox{\boldmath$#1$}}

18: \def\eq{\;\;\Longleftrightarrow\;\;}

19: \def\corresponds{\sim}

20: \def\succsim{\succ\kern-.9em_\sim\kern.3em}

21: \def\precsim{\prec\kern-1em_\sim\kern.3em}

22: \def\slantfrac#1#2{\kern1em^{#1}\kern-.3em/\kern-.1em_{#2}}

23: \def\lfrac#1#2{{}^{#1\!}\kern-.0em/_{#2}}

24: \def\trunc#1{[\mkern - 2.5 mu  [#1] \mkern - 2.5 mu ]}

25: 

26: \sloppy{}

27: 

28: \journal{Canadian Journal of Physics}

29: 

30: \volyear{1}[2004]{2004}

31: 

32: \bibliographystyle{myprsty}

33: 

34: \setcounter{page}{1}

35: 

36: \begin{document}

37: 

38: \title{Some Recent Advances in Bound--State Quantum Electrodynamics}

39: 

40: \author{U.~D.~Jentschura and J.~Evers}

41: 

42: \address{Max--Planck--Institut f\"ur Kernphysik,

43: Saupfercheckweg 1, 69117 Heidelberg, Germany}

44: 

45: \maketitle

46: 

47: \shortauthor{Jentschura and Evers}

48: 

49: \begin{abstract}

50: We discuss recent progress in various problems

51: related to bound-state quantum electrodynamics:

52: the bound-electron $g$ factor,

53: two-loop self-energy corrections 

54: and the laser-dressed Lamb shift.

55: The progress relies on various advances 

56: in the bound-state formalism, including 

57: ideas inspired by effective field theories

58: such as Nonrelativistic Quantum Electrodynamics.

59: Radiative corrections in dynamical processes

60: represent a promising field for further 

61: investigations.

62: \end{abstract}

63: 

64: \keywords{Calculations and mathematical techniques in atomic and

65: molecular physics, \\

66: quantum electrodynamics -- specific calculations;\\

67: {\it PACS numbers:} 31.15.-p, 12.20.Ds}

68: 

69: % 

70: % Introduction 

71: % 

72: \section{Introduction}

73: 

74: This is a brief summary of a number of recent advances

75: in our understanding of bound-state quantum electrodynamic

76: (QED) effects. The topics are (i) two-loop corrections

77: to the bound-electron $g$ factor, (ii) higher-order 

78: two-loop corrections to the self-energy of a bound electron,

79: and (iii) the laser-dressed Lamb shift. 

80: 

81: The first two of these rather diverse topics are 

82: related to two-loop effects. The investigation 

83: of these is simplified considerably by the use of 

84: effective field-theory techniques inspired

85: by Nonrelativistic QED (NRQED)~\cite{CaLe1986,Pa1993,PiSo1998}.

86: The Wilson coefficients multiplying the effective 

87: operators in the NRQED Lagrangian are matched against

88: those of the full relativistic theory, providing a simplified 

89: framework for the calculation of bound-state effects.

90: Scale-separation parameters such as the photon mass $\mu$

91: are cancelled at the end of the calculation.

92: The analysis of higher-order corrections to the $g$ factor 

93: of the bound electron is simplified further by 

94: a transformation to the length gauge, which results in 

95: a lesser number of terms to be considered than would be 

96: necessary in the velocity gauge. This fact has inspired 

97: the development of Long-wavelength QED (LWQED)~\cite{Pa2004},

98: a theory which is obtained after Power--Zienau and

99: Foldy--Wouthuysen transformations of the first-quantized 

100: Lagrangian; the second quantization is carried out by formulating

101: the path integral. Consequently, an improved understanding

102: and a tremendous simplification results for the 

103: calculation of a number of QED corrections for bound states,

104: such as the $g$ factor and higher-order corrections to the 

105: self-energy.

106: 

107: A further field of recent studies has been concerned with 

108: the interaction of a laser-dressed bound electron with the radiation 

109: field~\cite{JeEvHaKe2003,JeKe2004aop,EvJeKe2004,JeEvKe2004}. 

110: This process entails corrections which can only be understood

111: if the analysis is carried out right from the start

112: in the framework of the laser-dressed states, which are 

113: the eigenstates of the quantized atom-laser Hamiltonian

114: in the rotating-wave approximation~\cite{CT1975}.

115: 

116: %

117: % Bound-electron $g$ factor 

118: %

119: \section{Bound-electron $g$ factor}

120: 

121: In this section, we briefly summarize the 

122: results of a recent investigation~\cite{PaJeYe2004}

123: of the bound-electron $g$ factor, which is based on 

124: NRQED. The central result of this investigation is the 

125: following semianalytic expansion 

126: in powers of $Z\alpha$ and $\ln(Z\alpha)$

127: for the bound-electron $g$ factor ($n{\rm S}$ state)

128: in the non-recoil limit, which is 

129: the limit of an infinite nuclear mass:

130: %

131: \begin{align}

132: \label{gbound}

133: g(n{\rm S}) & = 

134: \underbrace{2 - \frac{2\, (Z\alpha)^2}{3\,n^2} + 

135: \frac{(Z\alpha)^4}{n^3} \, \left( \frac{1}{2 n} - \frac23 \right) + 

136: {\cal O}{(Z\alpha)^6}}_{\mbox{Breit (1928),

137: Dirac theory}}

138: \nonumber\\[4ex]

139: &  

140: + \underbrace{{\frac{\alpha}{\pi}}\, 

141: \left\{ {2 \times \frac{1}{2}}\, {\bigg(}

142: 1 + \frac{(Z\alpha)^2}{6 n^2} {\bigg)}

143: + \frac{(Z\alpha)^4}{n^3} {\bigg\{}

144: a_{41}\, {\ln}[(Z\alpha)^{-2}]  + 

145: a_{40} {\bigg\}}

146: + {\cal O}(Z\alpha)^5 \right\}}_{\mbox{one-loop correction}}

147: \nonumber\\[4ex]

148: & 

149: + \underbrace{{ \left(\frac{\alpha}{\pi}\right)^2}\,  

150: \left\{ {-0.656958} \,  {\bigg(} 1 + 

151: \frac{(Z\alpha)^2}{6 n^2} {\bigg)}  + 

152: \frac{(Z\alpha)^4}{n^3} {\bigg\{}

153: b_{41}\,

154: {\ln}[(Z\alpha)^{-2}] + 

155: b_{40} {\bigg\}}

156: + {\cal O}(Z\alpha)^5 \right\}}_{\mbox{two-loop correction}}

157: \nonumber\\[4ex]

158: & + {\cal O}(\alpha^3)\,.

159: \end{align}

160: %

161: This expansion is valid through the order

162: of two loops (terms of order $\alpha^3$ are 

163: neglected). The notation is in part inspired by the usual conventions

164: for Lamb-shift coefficients: the (lower case) $a$ terms denote the 

165: one-loop effects, with $a_{kj}$ denoting the 

166: coefficient of a term proportional to 

167: $\alpha\,(Z\alpha)^k\,{\ln}^j[(Z\alpha)^{-2}]$.

168: The $b$ terms denote the two-loop corrections,

169: with $b_{kj}$ multiplying 

170: a term proportional to 

171: $\alpha^2\,(Z\alpha)^k\,{\ln}^j[(Z\alpha)^{-2}]$.

172: In~\cite{PaJeYe2004}, complete results are derived for the 

173: coefficients $a_{41}$, $a_{40}$ and $b_{41}$.

174: 

175: In general, the expression corresponding to (\ref{gbound}) 

176: for a free electron is obtained by letting the 

177: parameter $Z\alpha \to 0$ in every term of the loop

178: expansion (expansion in powers of $\alpha$).

179: In this limit, the known free-electron two-loop result is 

180: recovered~\cite{KaKr1950,So1957,So1958,Pe1957helv,Re1972a,Re1972b,Ad1989}.

181: 

182: \begin{figure} 

183: \begin{center}

184: \begin{minipage}{10cm}

185: \begin{center}

186: \includegraphics[width=0.8\linewidth]{fig1.eps}

187: \end{center}

188: \caption{\label{fig1} Feynman diagrams 

189: for the two-loop self-energy corrections

190: to the bound-electron $g$-factor.}

191: \end{minipage}

192: \end{center}

193: \end{figure}

194: 

195: Up to the relative order $(Z\alpha)^2$, the free-electron 

196: contribution in one-, two-, and higher-loop

197: order is multiplied by a relative factor

198: %

199: \begin{equation}

200: \label{Zalpha2}

201: 1 + (Z\alpha)^2 \, a_{20} = 

202: 1 + (Z\alpha)^2 \, b_{20} = 1 + \frac{(Z\alpha)^2}{6 n^2}\,.

203: \end{equation}

204: %

205: This result consequently holds for the three-loop and the 

206: four-loop term not shown in Eq.~(\ref{gbound}). 

207: The applicability of the relative factor (\ref{Zalpha2})

208: to the two-loop term, valid through $(Z\alpha)^2$,

209: had been stressed previously in~\cite{Ka2001proc}.

210: The result in Eq.~(\ref{Zalpha2}) had been obtained originally 

211: in~\cite{Gr1970prl,Gr1970pra,GrHe1971,Fa1970plb,Fa1970nc}

212: (for the 1S state).

213: As is evident from Eq.~(\ref{gbound}), 

214: the correction of relative order $(Z\alpha)^2$

215: is different on the level of the tree-level diagrams

216: and reads $g(n{\rm S}) \sim 2 \times [1 - (Z\alpha)^2/(3 n^2)]$.

217: 

218: Explicit results for the coefficients in (\ref{gbound}),

219: restricted to the one-loop self-energy, read~\cite{PaJeYe2004}

220: %

221: \begin{subequations}

222: \label{a4}

223: \begin{align}

224: \label{a41}

225: a_{41}(n{\rm S}) &= \frac{32}{9}\,, \\[2ex]

226: \label{a40}

227: a_{40}(n{\rm S}) &= \frac{73}{54} 

228: - \frac{5}{24 n}

229: - \frac{8}{9} \, \ln k_0(n{\rm S}) 

230: - \frac{8}{3} \, \ln k_3(n{\rm S})\,.

231: \end{align}

232: \end{subequations}

233: %

234: Here, $\ln k_0(nS)$ is the Bethe logarithm for an $nS$ state,

235: and $\ln k_3(nS)$ is a generalization of the Bethe logarithm

236: to a perturbative potential of the form $1/r^3$.

237: Vacuum polarization adds a further $n$-independent

238: contribution of $(-16/15)$ to $a_{40}$~\cite{Ka2000}.

239: The Bethe logarithms for 1S and 2S~\cite{DrSw1990}

240: read

241: %

242: \begin{subequations}

243: \label{lnk0}

244: \begin{eqnarray}

245: \label{lnk01S}

246: \ln k_0(1{S}) &=& 2.984~128~555\,,\\[1ex]

247: \label{lnk02S}

248: \ln k_0(2{S}) &=& 2.811~769~893\,,

249: \end{eqnarray}

250: \end{subequations}

251: %

252: and the corresponding values for $\ln k_3$ read~\cite{PaJeYe2004}

253: %

254: \begin{subequations}

255: \label{lnk3}

256: \begin{eqnarray}

257: \label{lnk31S}

258: \ln k_3(1{S}) &=& 3.272~806~545\,,\\[1ex]

259: \label{lnk32S}

260: \ln k_3(2{S}) &=& 3.546~018~666\,.

261: \end{eqnarray}

262: \end{subequations}

263: %

264: The quantity $\ln k_3(n{\rm S})$ is defined as,

265: %

266: \begin{equation}

267: \int_0^\epsilon {\rm d} k\, k^2\,

268: \langle\phi|\bm{r}\,\frac{1}{E-H_S-k}\,\frac{1}{r^3}\,

269: \frac{1}{E-H_S-k}\,\bm{r}\,|\phi\rangle

270: = -4\,\frac{(Z\alpha)^3}{n^3} \,

271: \left[\ln \frac{\mu}{(Z\alpha)^2}+\frac56-\ln k_3\right]

272: \end{equation}

273: %

274: where the ultraviolet cutoff $\epsilon$ is to be understood 

275: in the sense of~\cite{Pa1993}, and the matching of the noncovariant 

276: cutoff $\epsilon$ to the covariant photon mass $\mu$ 

277: is given as (see~\cite{ItZu1980}, pp.~361--362)

278: %

279: \begin{equation}

280: \ln \left( \frac{2\epsilon}{m} \right) \to 

281: \ln \left(\frac{\mu}{m}\right) + \frac{5}{6}\,.

282: \end{equation}

283: %

284: However, this replacement is not unique and the constant

285: term depends on the actual form of the integrand. A different

286: replacement has to be used for some of the low-energy photon

287: corrections to the $g$ factor~\cite{PaJeYe2004}.

288: 

289: The results for the two-loop coefficients read

290: %

291: \begin{subequations}

292: \label{b4}

293: \begin{align}

294: \label{b41}

295: b_{41}(n{\rm S}) &= \frac{56}{9}\,, \\[2ex]

296: \label{b40}

297: b_{40}(1{\rm S}) &= -18.5(5.5)\,.

298: \end{align}

299: \end{subequations}

300: %

301: Here, the result for $b_{40}$ is an estimate based

302: on an explicit calculation of a large contribution due to 

303: low-energy virtual photons, and an estimate

304: of the remaining, unknown contribution due to high-energy virtual

305: photons. The dominant logarithmic two-loop

306: term $b_{41}$ is caused

307: exclusively by the two-loop self-energy diagrams in Fig.~\ref{fig1}

308: alone. The other two-loop diagrams, 

309: which include closed fermion loops,

310: can be found in Fig.~21 of~\cite{Be2000}.

311: The logarithmic term $b_{41}$ is, however,

312: exclusively related to the gauge-invariant 

313: subset displayed in Fig.~\ref{fig1}.

314: 

315: The newly calculated $a_{41}$, $a_{40}$ and $b_{41}$

316: are bound-state corrections to the electron $g$-factor

317: of order $(\alpha/\pi)\, (Z\alpha)^4$ and

318: $(\alpha/\pi)^2\, (Z\alpha)^4$, multiplied by 

319: logarithmic terms. These corrections

320: are (at $Z=1$) formally of order $\alpha^5$ and $\alpha^6$

321: and therefore of the same order of magnitude as

322: the tenth- and twelfth-order corrections to the

323: free-electron anomaly, which barely are of experimental

324: or theoretical significance at the current level

325: of accuracy. One may therefore ask why these binding

326: corrections are of any phenomenological significance.

327: The reason is that at somewhat higher $Z$,

328: the situation changes drastically, due to $Z^4$

329: scaling of the binding corrections. In addition,

330: due to numerically large coefficients and logarithmic

331: factors, the ``hierarchy'' of the corrections changes drastically.

332: Roughly, one may say that at $Z=1$, the bound-electron

333: anomalous magnetic moment is approximately independent

334: of binding corrections of order  $(\alpha/\pi)\, (Z\alpha)^4$

335: and higher, whereas for higher $Z$, the situation is reversed,

336: and the binding corrections to the one- and two-loop

337: contributions are numerically much more 

338: significant than the higher-loop 

339: free-electron corrections. This ``transition

340: from free to bound-state quantum electrodynamics'' as a

341: function of $Z$ is a somewhat

342: peculiar feature of the bound-electron $g$-factor.

343: 

344: For, example, we consider the ratio

345: %

346: \begin{equation}

347: \label{r1def}

348: r_1(Z) = \frac{\displaystyle 

349: \left(\frac{\alpha}{\pi}\right)\,(Z\alpha)^4\,\ln[(Z\alpha)^{-2}]}

350: {\displaystyle \left(\alpha/\pi\right)^4}\,,

351: \end{equation}

352: %

353: which gives an order-of-magnitude estimate for the the ratio of the

354: one-loop self-energy binding correction to the

355: eighth-order anomalous magnetic moment of the free electron.

356: We have

357: %

358: \begin{equation}

359: \label{r1num}

360: r_1(Z=1) \approx 2\,, \qquad 

361: r_1(Z=10) \approx 10^4 \gg 1\,. \qquad 

362: \end{equation}

363: %

364: For the two-loop logarithmic binding correction, we

365: have

366: %

367: \begin{equation}

368: \label{r2def}

369: r_2(Z) = \frac{\displaystyle 

370: \left(\frac{\alpha}{\pi}\right)^2\,(Z\alpha)^4\,\ln[(Z\alpha)^{-2}]}

371: {\displaystyle \left(\alpha/\pi\right)^4}

372: \end{equation}

373: %

374: and consequently

375: %

376: \begin{equation}

377: \label{r2num}

378: r_2(Z=1) \approx 0.2 < 1\,, \qquad 

379: r_2(Z=10) \approx 2 \times 10^3 \gg 1 \,. \qquad 

380: \end{equation}

381: %

382: As is evident from these considerations, the 

383: $(Z\alpha)^4$ one-loop and two-loop binding corrections are 

384: roughly of the same order of magnitude as the 

385: highly problematic four-loop 

386: corrections~\cite{KiLi1990,Ki1995ieee,HuKi1999} for the free 

387: electron. However, the situation changes drastically even at

388: very moderate nuclear charge numbers, and the binding 

389: corrections to the one-loop and two-loop contributions

390: become dominant over the higher-loop effects. 

391: 

392: The NRQED one-loop calculation~\cite{PaJeYe2004} is divided into 

393: three parts, the first of which entails fully relativistic 

394: form-factor corrections including lower-order terms, 

395: the second of which corresponds to a spin-dependent 

396: scattering amplitude, and the third of which is a 

397: low-energy Bethe-logarithm type correction that contains

398: $\ln k_0$ and $\ln k_3$. The one-loop

399: correction in Eq.~(\ref{gbound}) can therefore be 

400: written in a natural way as 

401: $ \delta g^{(1)} = g_1^{(1)} + g_2^{(1)} + g_3^{(1)}$,

402: where

403: 

404: \begin{subequations}

405: \label{g1}

406: \begin{align}

407: \label{g11}

408: g_1^{(1)} &= \frac{\alpha}{\pi}\biggl[ 1+\frac{(Z\alpha)^2}{6\,n^2} -

409: \frac{(Z\alpha)^4}{n^3}\,\left(\frac76

410:   +\frac5{24n} +\frac{16}{3}\, \ln \mu \right) \biggr] \,, \\[2ex]

411: %

412: \label{g21}

413: g_2^{(1)} &= \frac{\alpha}{\pi}\,\frac{(Z\alpha)^4}{n^3}\,

414: \Bigl(4+\frac{16}{9}\,\ln\mu\Bigr)\,, \\[2ex]

415: %

416: \label{g31}

417: g_3^{(1)} &= \frac{\alpha}{\pi}\,\frac{(Z\alpha)^4}{n^3}\,\frac{32}{9}\,

418: \biggl[\ln\frac{\mu}{(Z\alpha)^2} -\frac{5}{12} -\frac{\ln k_0}{4}

419: -\frac{3}{4}\,\ln k_3 \biggr]\,.

420: \end{align}

421: \end{subequations}

422: 

423: The new contribution of order $(Z\alpha)^4$ can be

424: compared with the numerical results for the self-energy 

425: correction~\cite{YeInSh2002,YeInSh2004} complete to all orders in $Z\alpha$. 

426: Assuming correctness of the logarithmic term in Eq.~(\ref{g1}), 

427: a fit to numerical data yields 

428: $a_{40}^{(1)}({\rm 1S}) = -10.2(1)$ and 

429: $a_{40}^{(1)}({\rm 2S}) = -10.6(1.2)$ for the constant term, in excellent

430: agreement with the analytic results which read

431: $a_{40}^{(1)}({\rm 1S}) = -10.236~524~318(1)$ 

432: and $a_{40}^{(1)}({\rm 2S}) = -10.707~715~607(1)$ . 

433: 

434: Having verified the consistency 

435: of the analytic~\cite{PaJeYe2004}

436: and numerical results~\cite{YeInSh2002,YeInSh2004}, an interpolation 

437: procedure~\cite{IvKa2001proc} may now be used to extract 

438: a more accurate theoretical prediction 

439: at low and intermediate nuclear charge numbers, 

440: if combined with numerical results at

441: higher $Z$~\cite{YeInSh2002}. Thus, the 

442: results in Eqs.~(\ref{a4}) and (\ref{b4}) 

443: may be used in order to infer improved theoretical 

444: predictions for the bound-electron $g$ factor, 

445: notably in the experimentally important 

446: special cases of hydrogenlike carbon~\cite{HaEtAl2000prl}

447: and oxygen~\cite{VeEtAl2004}. Alternatively, the 

448: improved status of the theory may be used in order to 

449: infer a more accurate value of the 

450: electron mass. Specifically, the value from 

451: the carbon measurement~\cite{HaEtAl2000prl},

452: using the new theory, reads

453: %

454: \begin{equation}

455: m(^{12}{\rm C}^{5+}) = 0.000\,548\,579\,909\,41\, (29)(3)\ {\rm u} \,. 

456: \end{equation}

457: %

458: The first error comes from the 

459: experiment~\cite{HaEtAl2000prl}, and the 

460: second error corresponds to the theoretical uncertainty. 

461: The conclusion is that a further 

462: improvement of the experiment could lead to a much 

463: better determination of the electron 

464: mass; the new theory provides room 

465: for at least an improved determination by one order of magnitude.

466: 

467: For the calculation of yet higher-order binding corrections

468: to the one-loop and two-loop contributions,

469: a detailed understanding of the two-loop form-factors,

470: including their slopes, is required. 

471: The most recent calculations of these effects,

472: in both dimensional and photon-mass regularizations,

473: can be found in~\cite{BoMaRe2003,MaRe2003,BoMaRe2004}.

474: 

475: \begin{figure}[htb]

476: \begin{center}

477: \begin{minipage}{14cm}

478: \begin{center}

479: \includegraphics[width=1.0\linewidth]{fig2.eps}

480: \end{center}

481: \caption{\label{fig2} The two-loop corrections to the 

482: Lamb shift in hydrogenlike systems fall naturally 

483: into three gauge-invariant subsets, which are the 

484: pure two-loop self-energy terms (2LSE), the 

485: vacuum-polarization correction to the virtual-photon  

486: line in the self-energy (SVPE), and the 

487: self-energy vacuum-polarization and pure 

488: two-loop vacuum-polarization corrections (SEVP).

489: The two-loop Bethe logarithm is a numerically 

490: large correction to the nonlogarithmic term of order

491: $\alpha^2 (Z\alpha)^6$ and it is exclusively related 

492: to the 2LSE subset; however, for a complete result

493: in this order, contributions from the other gauge-invariant subsets 

494: will have to be considered as well.}

495: \end{minipage}

496: \end{center}

497: \end{figure}

498: 

499: %

500: % Two-loop corrections

501: %

502: \section{Two-loop Bethe logarithms}

503: 

504: As is well known~\cite{ApBr1970,Ka1993log,Pa1994prl,%

505: EiSh1995,Ka1996,Ka1996b,Ka1997,Pa2001}, 

506: the two-loop Lamb shift $\Delta E^{(2)}$, 

507: in the limit of an infinite nuclear mass, may be written as

508: %

509: \begin{eqnarray}

510: \Delta E^{(2)} =

511: \left(\frac{\alpha}{\pi}\right)^2 \, 

512: \frac{(Z\alpha)^4 \, m_{\rm e} \,c^2}{n^3} \,

513: H(Z\alpha)\,.

514: \end{eqnarray}

515: %

516: For S states, the dimensionless function $H(Z\alpha)$,

517: has a semianalytic expansion of the form

518: %

519: \begin{align}

520: H(Z\alpha) = &  B_{40} + (Z\alpha) \, B_{50}

521: \nonumber\\[2ex]

522: & + (Z\alpha)^2 \, \left[ B_{63} \,

523: \ln^3[(Z\alpha)^{-2}] + B_{62} \, \ln^2[(Z\alpha)^{-2}]

524: + B_{61} \, \ln[(Z\alpha)^{-2}] + B_{60} \right] \,,

525: \end{align}

526: %

527: where we ignore higher-order terms, and upper case is used 

528: for the $B_{ij}$ coefficients that multiply terms 

529: of order $\alpha^2\,(Z\alpha)^i\,\ln^j[(Z\alpha)^{-2}] \, m_{\rm e} \,c^2$.

530: The coefficients, restricted to the 

531: two-photon self-energy diagrams (Fig.~\ref{fig2}),

532: read as follows

533: %

534: \begin{subequations}

535: \begin{eqnarray}

536: B^{\rm (2LSE)}_{63}(n{\rm S}) &=&

537: -\frac{8}{27} = -0.296296\,, \\[2ex]

538: B^{\rm (2LSE)}_{62}(1{\rm S}) &=& 

539: \frac{16}{27} - \frac{16}{9}\, \ln(2) 

540: = -0.639\,669\,, \\[2ex]

541: %

542: B^{\rm (2LSE)}_{62}(n{\rm S}) &=& 

543: B^{\rm (2LSE)}_{62}(1{\rm S}) + 

544: \frac{16}{9} \left( \frac34 + \frac{1}{4 n^2} -

545: \frac1n - \ln(n) \! + \! \Psi(n) + C\right)\,.

546: \end{eqnarray}

547: \end{subequations}

548: 

549: The $n$-dependence of $B_{61}$ has been 

550: clarified in~\cite{Pa2001,Je2003jpa}

551: %

552: \begin{eqnarray}

553: B^{\rm (2LSE)}_{61}(n{\rm S}) &=& B^{\rm (2LSE)}_{61}(1{\rm S})

554: + \frac43 \, \left[ N(n{\rm S}) - N(1{\rm S})\right] 

555: \nonumber\\[0.5ex]

556: & & + \! \left( \frac{80}{27} - \frac{32}{9} \, \ln (2)\right) 

557: \left(\frac34 + \frac{1}{4 n^2} - \frac1n - \ln(n) \! + \! \Psi(n) 

558: \! + \! C\right)\,, 

559: \end{eqnarray}

560: %

561: where $C=0.577216\dots$ is Euler's constant,

562: $\Psi(n)$ is the logarithmic derivative of the Gamma function, 

563: and $N(n{\rm S})$ is related to a correction to the

564: Bethe logarithm induced by a Dirac-delta potential.

565: Explicit values for $N(n{\rm S})$ can be found in~\cite{Je2003jpa}

566: ($n = 1,\dots,8$).

567: 

568: \begin{figure}[htb]

569: \begin{center}

570: \begin{minipage}{8cm}

571: \begin{center}

572: \includegraphics[width=0.9\linewidth]{fig3.eps}

573: %

574: \caption{\label{fig3}

575: The integration regions for the two virtual photons in the 

576: two-loop self-energy problem comprise a low-energy regime 

577: with two low-energy photons, which gives rise to $b_L$.

578: The middle-energy regions with one low-energy and a one 

579: high-energy photon give rise to $b_M$. The 

580: high-energy contribution $b_F + b_H$ is as yet unknown.}

581: \end{center}

582: \end{minipage}

583: \end{center}

584: \end{figure}

585: 

586: The $B_{60}$ coefficients are the sum of several 

587: contributions

588: %

589: \begin{subequations}

590: \begin{eqnarray}

591: \label{2lse}

592: B^{\rm (2LSE)}_{60}(n{\rm S}) &=&

593: b_L + b_M + \left\{ b_F + b_H + b_{\rm VP} \right\}\,, \\[2ex]

594: b_L &=& b_L(n{\rm S}) \; \corresponds \;

595: \mbox{two-loop Bethe logarithm, two soft photons}\,,

596: \\[2ex]

597: b_M &=& b_M(n{\rm S}) \; = \;

598: \frac{10}{9} \, N(n{\rm S})

599: \; \corresponds \; \mbox{one soft, one hard photon}\,,

600: \\[2ex]

601: b_F &\corresponds&

602: \mbox{soft electron momenta, two hard photons}\,,

603: \\[2ex]

604: b_H &\corresponds&

605: \mbox{hard electron momenta, two hard photons}\,,

606: \\[2ex]

607: b_{\rm VP} &\corresponds&

608: \mbox{vacuum-polarization corrections}\,.

609: \end{eqnarray}

610: \end{subequations}

611: %

612: Only the terms $b_L + b_M$ are currently 

613: known~\cite{PaJe2003,Je2004b60}~(see also Fig.~\ref{fig2}).

614: The contributions in curly brackets in 

615: Eq.~(\ref{2lse}) remain to be evaluated.

616: However, an estimate for the total value of $B_{60}$

617: may be obtained, 

618: %

619: \begin{eqnarray}

620: B_{60}(n{\rm S}) &=&

621: b_L + b_M + \left\{ b_F + b_H + b_{\rm VP} \right\}\,, \nonumber\\[2ex]

622: B_{60}(n{\rm S}) &\approx&

623: b_L + b_M \; \pm \; 15\,\%\,.

624: \end{eqnarray}

625: %

626: This estimate~\cite{PaJe2003,Je2004b60} is based on corresponding 

627: one-loop calculations,

628: where the low-energy virtual photons give the by far 

629: dominant contribution to the constant term~\cite{Pa1993}.

630: The results for the two-loop Bethe logarithms of S states 

631: read~\cite{PaJe2003,Je2004b60}

632: %

633: \begin{subequations}

634: \label{bL}

635: \begin{eqnarray}

636: b_L(1S) &=& -81.4(3)\,, \\[2ex]

637: b_L(2S) &=& -66.6(3)\,, \\[2ex]

638: b_L(3S) &=& -63.5(6)\,, \\[2ex]

639: b_L(4S) &=& -61.8(8)\,, \\[2ex]

640: b_L(5S) &=& -60.6(8)\,, \\[2ex]

641: b_L(6S) &=& -59.8(8)\,.

642: \end{eqnarray}

643: \end{subequations}

644: %

645: A few clarifying remark might be in order. The $B_{60}$ coefficient

646: multiplies a correction of order

647: $\alpha^2 (Z\alpha)^6$, which is effectively an order-$\alpha^8$ 

648: contribution to the energy levels of hydrogen ($Z=1$). 

649: In order to complete the calculation at this 

650: order of magnitude, it would also be necessary to 

651: consider the four-loop Dirac form-factor slope of the electron,

652: as well as the three-loop binding correction

653: of order $\alpha^3\,(Z\alpha)^5\,m_{\rm e}\,c^2$~\cite{EiSh2004}.

654: The three-loop slope has recently been evaluated 

655: in~\cite{MeRi2000}, completing the theory of 

656: energy levels in hydrogen up to the order of $\alpha^7$.

657: 

658: %

659: % Laser-dressed Lamb shift

660: %

661: \section{Laser-dressed Lamb shift}

662: 

663: In the recent past, seminal advances have been obtained both 

664: in the techniques

665: of high-precision spectroscopy (e.g.,~\cite{ReEtAl2000}),

666: and in the coherent preparation 

667: and manipulation of media by external electromagnetic 

668: fields~\cite{ScZu1997,FiSw2005}.

669: Thus it is desirable to study the bound electrons 

670: interacting simultaneously both with the quantized 

671: radiation field and with an external driving 

672: field. An accurate theory of such systems,

673: including all dynamic effects, might eventually

674: open a possibility for a whole new class of high-precision

675: experiments, provided that technical problems related

676: to the required highly accurate intensity stabilization of the laser 

677: (and others) can be solved. 

678: Traditionally, radiative and relativistic corrections 

679: are treated with methods of QED, whereas studies 

680: related to the dynamical nature of the interaction of matter 

681: with driving laser fields are the domain of Quantum Optics (QO).

682: Obviously, a treatment of bound electrons in the presence

683: of both the radiation field and external driving fields requires

684: a combination of ideas from both subject areas:

685: While, {\em a priori}, the essential-state approximation

686: of QO~\cite{ScZu1997} is not sufficient to obtain the accuracy

687: of QED, a perturbative treatment of the interaction of the bound

688: electron with a strong external (laser) field as in QED is hopeless

689: because of the large coupling parameter.

690: 

691: In~\cite{JeEvHaKe2003,JeKe2004aop,EvJeKe2004}, 

692: an atom with two relevant energy levels driven by a strong

693: near-resonant monochromatic laser field is studied as the easiest

694: model system for the above problem. The incoherent

695: part of the resonance fluorescence spectrum emitted by this system

696: in QO is known as the Mollow spectrum, where the coupling strength

697: is characterized by the the Rabi frequency

698: $\Omega$ defined as ($\hbar = \epsilon_0 = c = 1$)

699: %

700: \begin{equation}

701: \Omega = -q \left <e|{\bm x}\cdot {\bm \epsilon}_{\rm L}|g \right >  

702: {\cal E}_{\rm L}\,,

703: \end{equation}

704: %

705: for a driving laser field ${\bm E}_{\rm L}(t) = {\cal E}_{\rm L} {\bm

706: \epsilon}_{\rm L} \cos (\omega_{\rm L}t)$ with frequency $\omega_{\rm L}$,

707: macroscopic classical amplitude ${\cal E}_{\rm L}$ and 

708: polarization ${\bm \epsilon}_{\rm L}$. Here,

709: $q=-|q|$ is the elementary charge.

710: The corresponding coupling constant $g_{\rm L}$ for 

711: the interaction of a quantized driving laser field with the 

712: main atomic transition

713: is defined by

714: %

715: \begin{equation}

716: g_{\rm L} = 

717: - q \, \langle g | \bm{\epsilon}_{\rm L} \cdot \bm{x} | e \rangle \,

718: {\cal E}_{\rm L}^{\rm (\gamma)}\, ,

719: \end{equation}

720: %

721: where ${\cal E}_{\rm L}^{\rm (\gamma)} = 

722: \sqrt{\omega_{\rm L}/2V}$ is the electric laser field per photon

723: and $V$ is the quantization volume. The matching of the electric

724: field per photon with the corresponding classical macroscopic

725: electric field ${\cal E}_{\rm L}$ is then given by

726: %

727: \begin{equation}

728: \label{matching}

729: 2 \: \sqrt{n+1} \: {\cal E}_{\rm L}^{\rm (\gamma)} \: 

730: \longleftrightarrow \: {\cal E}_{\rm L} \,.

731: \end{equation}

732: %

733: 

734: If $\Omega$

735: is larger than the natural decay width $\Gamma$ of the 

736: transition, then the Mollow spectrum approximately consists of

737: one central and two sideband peaks of Lorentzian shape, which

738: are located symmetrically around the driving laser field frequency.

739: The sideband peaks are shifted from the driving field frequency

740: by the generalized Rabi frequency 

741: $\Omega_{\textrm R} = \sqrt{\Omega^2 + \Delta^2}$, 

742: where $\Delta=\omega_{\textrm L} - \omega_{\textrm R}$ 

743: is the detuning of the laser field frequency $\omega_{\textrm L}$ 

744: with regard to the atomic transition frequency $\omega_{\textrm R}$. 

745: The shape of the Mollow spectrum

746: may easily be explained in terms of the dressed states, which

747: are defined as the eigenstates of the quantum optical interaction

748: picture Hamiltonian describing the matter-light interaction.

749: In transferring to the dressed state picture, the interaction with

750: the driving laser field is accounted for to all orders.

751: 

752: Thus, when evaluating radiative and relativistic corrections

753: to the Mollow spectrum, it is natural to start the analysis

754: from the dressed-state basis as opposed to the

755: unperturbed atomic bare-state basis. 

756: In~\cite{JeEvHaKe2003,JeKe2004aop}, it was shown

757: that this distinction in fact has to be made. It is not 

758: sufficient to modify the energies (which enter in the

759: formula for the dressed states) according to the usual

760: bare-state Lamb shift in order to obtain the correct

761: result for the corrections to the Mollow spectrum. Instead,

762: at nonvanishing detuning and nonvanishing Rabi frequency, 

763: a treatment starting from the dressed-state basis leads

764: to an additional nontrivial correction term. This term

765: gives rise to a shift of the Mollow sidebands relative

766: to the central peak given by

767: %

768: \begin{equation}

769: \label{deltaomegaplusminus}

770: \delta \omega_\pm^{(C)} =

771: \mp {\cal C}\, \frac{\Omega^2}{\sqrt{\Omega^2 + \Delta^2}}\,,

772: \end{equation}

773: %

774: where

775: %

776: \begin{eqnarray}

777: {\cal C} &=& \frac{\alpha}{\pi} \, \ln[(Z\alpha)^{-2}] \,

778: \frac{\left< \bm{p}^2 \right>_g + \left< \bm{p}^2 \right>_e}{m^2}

779: \end{eqnarray}

780: %

781: is a dimensionless constant. Here, the notation $\langle . \rangle_g$

782: and $\langle . \rangle_e$ denotes the expectation value evaluated

783: with the ground or excited atomic state, respectively.

784: 

785: Inspired by the interpretation of the bare Lamb shift correction

786: in terms of a ``summed'' shift as 

787: in~\cite{JeEvHaKe2003,JeKe2004aop}, this additional correction 

788: can be interpreted as a modification

789: to the Rabi frequency:

790: %

791: \begin{equation}

792: \label{summedC}

793: \delta \overline{\omega}^{(C)}_\pm = \pm

794: \left( \sqrt{\Omega^2 \, (1 - {\cal C})^2 +

795: \Delta ^2} - \sqrt{\Omega^2 + \Delta^2}\right)\,,

796: \end{equation}

797: %

798: with $\delta \overline{\omega}^{(C)}_\pm \approx 

799: \delta\omega_\pm^{(C)}$ because of the 

800: smallness of the correction. It should be noted that this

801: interpretation in terms of a summation is not trivial 

802: and was shown to be valid up to first order in the correction.

803: 

804: In~\cite{JeEvHaKe2003,EvJeKe2004}, the leading relativistic

805: and radiative corrections up to relative orders $(Z\alpha)^2$

806: and $\alpha(Z\alpha)^2$, respectively, have been evaluated,

807: as well as all other relevant correction terms up to the

808: specified order of approximation. It turns out that all corrections

809: may be interpreted as either corrections to the Rabi frequency

810: $\Omega$ or as corrections to the detuning $\Delta$, such that one

811: can define the fully corrected generalized Rabi frequency

812: $\Omega^{(j)}_{\mathcal C}$ by 

813: %

814: \begin{equation}

815: \label{fullOmega}

816: \Omega^{(j)}_{\mathcal C} =

817: \sqrt{\Omega^2 \cdot \left(1+{\hat{\Omega}_{\rm rad}^{(j)}}\right)^2 +

818: \left(\Delta - {\Delta^{(j)}_{\rm rad}}\right)^2}\,.

819: \end{equation}

820: %

821: Here, $\hat{\Omega}_{\rm rad}^{(j)}$ contains

822: all corrections to the Rabi frequency, namely

823: the relativistic and radiative corrections to the transition dipole 

824: moment, field-configuration dependent corrections, higher-order

825: corrections to the self-energy, and corrections to the secular 

826: approximation.

827: $\Delta^{(j)}_{\rm rad}$

828: consists of all corrections to the detuning, i.e.

829: the bare Lamb shift, Bloch-Siegert shifts, and off-resonant

830: radiative corrections.

831: The superscript $(j)$ indicates the dependence

832: of the result on the total angular momentum quantum number.

833: 

834: Equation~(\ref{fullOmega}) summarizes the main result of 

835: this study: In the presence of driving laser fields, the usual

836: bare state Lamb shift of the atomic states is augmented

837: by additional correction terms. These in part depend on the

838: laser field parameters $\Omega$ and $\Delta$, which span

839: a two-dimensional parameter manifold determining the

840: actual value of the dynamical Lamb shift.

841: 

842: A promising candidate for the experiment are

843: the hydrogen $1S_{\nicefrac{1}{2}}\leftrightarrow 2P_{\nicefrac{1}{2}}$ and 

844: $1S_{\nicefrac{1}{2}}\leftrightarrow 2P_{\nicefrac{3}{2}}$ 

845: transitions. 

846: We consider here

847: as a specific example the 

848: $1S_{\nicefrac{1}{2}}\leftrightarrow 2P_{\nicefrac{1}{2}}$ transition

849: with $\Omega = 1000\cdot \Gamma_{\nicefrac{1}{2}}$ and

850: $\Delta = 50 \cdot \Gamma_{\nicefrac{1}{2}}$ as the

851: laser field parameters. The Rabi frequency is shifted

852: with respect to the leading-order 

853: expression $\Omega_{\rm R} = \sqrt{\Omega^2 + \Delta^2}$ by

854: relativistic and radiative corrections as follows,

855: %

856: \begin{equation}

857: \pm \left (  \Omega^{(\nicefrac{1}{2})}_{\mathcal C} - \Omega_{\rm R} \right ) 

858: = \pm 738.282(60)\cdot 10^6 \: \rm{Hz}\, .

859: \end{equation}

860: %

861: This driving laser field parameter set is expected to be within reach 

862: of improvements of the currently available Lyman-$\alpha$ laser 

863: sources~\cite{EiWaHa2001} in the next few years.

864: The corresponding result for the 

865: $1S_{\nicefrac{1}{2}}\leftrightarrow 2P_{\nicefrac{3}{2}}$ transition 

866: with $\Omega = 1000\cdot \Gamma_{\nicefrac{3}{2}}$, $\Delta = 50 \cdot \Gamma_{\nicefrac{3}{2}}$

867: is

868: \begin{equation}

869: \pm \left (  \Omega^{(\nicefrac{3}{2})}_{\mathcal C} - \Omega_{\rm R} \right ) 

870: = \pm 734.871(60)\cdot 10^6 \: \rm{Hz}\, .

871: \end{equation}

872: %

873: All given uncertainties are due to unknown higher-order

874: terms~\cite{EvJeKe2004}. 

875: 

876: By a comparison to experimental data,

877: one may verify the presence of dynamical leading-logarithmic correction

878: to the dressed-state radiative shift in Eq.~(\ref{deltaomegaplusminus}), 

879: which cannot

880: be explained in terms of the bare Lamb shift alone. This allows to 

881: address

882: questions related to the physical reality of the dressed states.

883: On the other hand, the comparison with experimental results could

884: also be used

885: to interpret the nature of the evaluated radiative corrections in the sense

886: of the summation formulas which lead to the interpretation of

887: the shifts as arising from relativistic and radiative corrections

888: to the detuning and the Rabi frequency.

889: 

890: %

891: % Laser-dressed Lamb shift

892: %

893: \section{Conclusions}

894: 

895: One of the obvious conclusions to 

896: be drawn from the recent advances in bound-state

897: quantum electrodynamics is as follows.

898: A widespread opinion has been invalidated 

899: which suggested that two-loop corrections 

900: to the bound-electron $g$ factor, and two-loop

901: corrections to the Lamb shift in higher order

902: might be a severe, if not insurmountable,

903: obstacle against further theoretical progress.

904: Quite to the contrary, the recent advances have shown 

905: that an understanding of these effects is feasible

906: to a good accuracy. While some important 

907: contributions remain to be evaluated,

908: the further program (e.g., the evaluation of 

909: the remaining high-energy parts)

910: is clearly defined, and  decisive first 

911: steps toward a much improved understanding 

912: have been accomplished.

913: 

914: Further advances are possible both on the 

915: experimental as well as on the theoretical side: 

916: concerning the $g$ factor, the recent 

917: theoretical progress allows for an order-of-magnitude improvement 

918: of the value for the electron mass based on a potential new measurement alone.

919: Regarding the Lamb shift in hydrogen and low-$Z$

920: hydrogenlike systems, one has recently gained 

921: an improved understanding of the binding corrections of order 

922: $\alpha^2 \,(Z\alpha)^6\, \ln^j[(Z\alpha)^{-2}]\,

923: m_{\rm e}\,c^2$ (where $j$ may assume the values $j = 0,\dots,3$).

924: Recent numerical investigations in the regime 

925: of intermediate nuclear charge numbers~\cite{YeInSh2004nim}

926: have also contributed toward an improvement of our understanding.

927: The extrapolation of the low-$Z$ results

928: by deferred Pad\'{e} approximants~\cite{Je2003plb} suggests

929: a rather good general consistency of both approaches,

930: while some issues regarding the consistency of the

931: analytic and numerical approaches remain to be 

932: addressed~\cite{YeInSh2004nim}.

933: 

934: Other progress concerns the modifications of radiative 

935: corrections in dynamical processed as opposed to 

936: $S$-matrix energy shifts. The laser-dressed 

937: Lamb shift~\cite{JeEvHaKe2003,JeKe2004aop,EvJeKe2004,JeEvKe2004}

938: is a dynamical correction to the dressed-state~\cite{CT1975} 

939: quasi-energies. The self-energy, in this case,

940: gives the by far dominant effect. While the bulk of the 

941: laser-dressed Lamb shift can be understood in terms

942: of a radiative correction to the detuning, which is 

943: taken into account in a natural way by evaluating the 

944: detuning in terms of the observed (low-intensity) 

945: resonance frequency of the transition, a few tiny 

946: shifts persist which can only be understood if the 

947: system is treated in the dressed-state picture right 

948: from the start. The dynamic Lamb shift is an 

949: effect which depends on two parameters that determine

950: the dynamics of the system: (i) the detuning and (ii) the

951: Rabi frequency. Therefore, 

952: the dynamical Lamb shift could be mapped out as a function 

953: of these parameters in a possible experiment.

954: A rather promising candidate for a possible measurement

955: would be based on the hydrogen 1S--2P$_{\nicefrac{1}{2}}$ 

956: transition~\cite{JeEvHaKe2003}. However, a very attractive

957: alternative would be provided by a forbidden M1 transition~\cite{DrEtAl2003}

958: in Ar XIV, $2s^2 \,2p\, 

959: {}^2{\rm P}_{\nicefrac{1}{2}}$--${}^2{\rm P}_{\nicefrac{3}{2}}$,

960: provided the many-body QED effects can be treated to sufficient

961: accuracy. The system in question is described very well by 

962: a two-level formalism, and the resonance line width 

963: is small.

964: 

965: \section*{Acknowledgments}

966: 

967: The authors acknowledge insightful discussions with 

968: C. H. Keitel, V. A. Yerokhin and K. Pachucki. 

969: Helpful conversations with S. G. Karshenboim are also 

970: gratefully acknowledged.

971: 

972: \begin{thebibliography}{10}

973: 

974: \bibitem{CaLe1986}

975: W.~E. Caswell and G.~P. Lepage, Phys. Lett. B {\bf 167},  437  (1986).

976: 

977: \bibitem{Pa1993}

978: K. Pachucki, Ann. Phys. (N.Y.) {\bf 226},  1  (1993).

979: 

980: \bibitem{PiSo1998}

981: A. Pineda and J. Soto, Phys. Rev. D {\bf 59},  016005  (1998).

982: 

983: \bibitem{Pa2004}

984: K. Pachucki, Phys. Rev. A {\bf 69},  052502  (2004).

985: 

986: \bibitem{JeEvHaKe2003}

987: U.~D. Jentschura, J. Evers, M. Haas, and C.~H. Keitel, Phys. Rev. Lett. {\bf

988:   91},  253601  (2003).

989: 

990: \bibitem{JeKe2004aop}

991: U.~D. Jentschura and C.~H. Keitel, Ann. Phys. (N.Y.) {\bf 310},  1  (2004).

992: 

993: \bibitem{EvJeKe2004}

994: J. Evers, U.~D. Jentschura, and C.~H. Keitel, Relativistic and Radiative

995:   Corrections to the Mollow Spectrum, e-print quant-ph/0403202, submitted to

996:   Phys. Rev. A.

997: 

998: \bibitem{JeEvKe2004}

999: U.~D. Jentschura, J. Evers, and C.~H. Keitel, Relativistic and Radiative

1000:   Corrections to Multi-Level Mollow--Type Spectra, Laser Physics, at press

1001:   (special issue on the occasion of the 70${}^{\rm th}$ birthday of Prof.

1002:   Herbert Walther).

1003: 

1004: \bibitem{CT1975}

1005: C. Cohen-Tannoudji,  in {\em Aux fronti\`{e}res de la spectroscopie

1006:   laser/Frontiers in Laser Spectroscopy}, edited by R. Balian, S. Haroche, and

1007:   S. Liberman (North-Holland, Amsterdam, 1975), pp.\ 4--104.

1008: 

1009: \bibitem{PaJeYe2004}

1010: K. Pachucki, U.~D. Jentschura, and V.~A. Yerokhin, Phys. Rev. Lett. {\bf 93},

1011:   150401  (2004).

1012: 

1013: \bibitem{KaKr1950}

1014: R. Karplus and N.~M. Kroll, Phys. Rev. {\bf 77},  536  (1950).

1015: 

1016: \bibitem{So1957}

1017: C.~M. Sommerfield, Phys. Rev. {\bf 107},  328  (1957).

1018: 

1019: \bibitem{So1958}

1020: C.~M. Sommerfield, Ann. Phys. (N.Y.) {\bf 5},  26  (1958).

1021: 

1022: \bibitem{Pe1957helv}

1023: A. Petermann, Helv. Phys. Acta {\bf 30},  407  (1957).

1024: 

1025: \bibitem{Re1972a}

1026: E. Remiddi, Nuovo Cim. A {\bf 11},  825  (1972).

1027: 

1028: \bibitem{Re1972b}

1029: E. Remiddi, Nuovo Cim. A {\bf 11},  865  (1972).

1030: 

1031: \bibitem{Ad1989}

1032: G.~S. Adkins, Phys. Rev. D {\bf 39},  3798  (1989).

1033: 

1034: \bibitem{Ka2001proc}

1035: S.~G. Karshenboim,  in {\em The Hydrogen Atom}, edited by S.~G. Karshenboim and

1036:   F.~S. Pavone (Springer, Berlin, 2001), pp.\ 651--663.

1037: 

1038: \bibitem{Gr1970prl}

1039: H. Grotch, Phys. Rev. Lett. {\bf 24},  39  (1970).

1040: 

1041: \bibitem{Gr1970pra}

1042: H. Grotch, Phys. Rev. A {\bf 2},  1605  (1970).

1043: 

1044: \bibitem{GrHe1971}

1045: H. Grotch and R.~A. Hegstrom, Phys. Rev. A {\bf 4},  59  (1971).

1046: 

1047: \bibitem{Fa1970plb}

1048: R. Faustov, Phys. Lett. B {\bf 33},  422  (1970).

1049: 

1050: \bibitem{Fa1970nc}

1051: R. Faustov, Nuovo Cim. {\bf 69},  37  (1970).

1052: 

1053: \bibitem{Ka2000}

1054: S.~G. Karshenboim, Phys. Lett. A {\bf 266},  380  (2000).

1055: 

1056: \bibitem{DrSw1990}

1057: G.~W.~F. Drake and R.~A. Swainson, Phys. Rev. A {\bf 41},  1243  (1990).

1058: 

1059: \bibitem{ItZu1980}

1060: C. Itzykson and J.~B. Zuber, {\em Quantum Field Theory} (McGraw-Hill, New York,

1061:   NY, 1980).

1062: 

1063: \bibitem{Be2000}

1064: T. Beier, Phys. Rep. {\bf 339},  79  (2000).

1065: 

1066: \bibitem{KiLi1990}

1067: T. Kinoshita and W.~B. Lindquist, Phys. Rev. D {\bf 42},  636  (1990).

1068: 

1069: \bibitem{Ki1995ieee}

1070: T. Kinoshita, IEEE Trans. Instrum. Meas. {\bf 44},  498  (1995).

1071: 

1072: \bibitem{HuKi1999}

1073: V.~W. Hughes and T. Kinoshita, Rev. Mod. Phys. {\bf 71},  S133  (1999).

1074: 

1075: \bibitem{YeInSh2002}

1076: V.~A. Yerokhin, P. Indelicato, and V.~M. Shabaev, Phys. Rev. Lett. {\bf 89},

1077:   143001  (2002).

1078: 

1079: \bibitem{YeInSh2004}

1080: V.~A. Yerokhin, P. Indelicato, and V.~M. Shabaev, Phys. Rev. A {\bf 69},

1081:   052503  (2004).

1082: 

1083: \bibitem{IvKa2001proc}

1084: V.~G. Ivanov and S.~G. Karshenboim,  in {\em The Hydrogen Atom}, edited by

1085:   S.~G. Karshenboim and F.~S. Pavone (Springer, Berlin, 2001), pp.\ 637--650.

1086: 

1087: \bibitem{HaEtAl2000prl}

1088: H. H\"{a}ffner, T. Beier, N. Hermanspahn, H.-J. Kluge, W. Quint, J. Verd\'{u},

1089:   and G. Werth, Phys. Rev. Lett. {\bf 85},  5308  (2000).

1090: 

1091: \bibitem{VeEtAl2004}

1092: J. Verd\'{u}, S. Djeki\'{c}, S. Stahl, T. Valenzuela, M. Vogel, G. Werth, T.

1093:   Beier, H.~J. Kluge, and W. Quint, Phys. Rev. Lett. {\bf 92},  093002  (2004).

1094: 

1095: \bibitem{BoMaRe2003}

1096: R. Bonciani, P. Mastrolia, and E. Remiddi, Nucl. Phys. B {\bf 661},  289

1097:   (2003).

1098: 

1099: \bibitem{MaRe2003}

1100: P. Mastrolia and E. Remiddi, Nucl. Phys. B {\bf 664},  341  (2003).

1101: 

1102: \bibitem{BoMaRe2004}

1103: R. Bonciani, P. Mastrolia, and E. Remiddi, Nucl. Phys. B {\bf 676},  399

1104:   (2004).

1105: 

1106: \bibitem{ApBr1970}

1107: T. Appelquist and S.~J. Brodsky, Phys. Rev. A {\bf 2},  2293  (1970).

1108: 

1109: \bibitem{Ka1993log}

1110: S.~G. Karshenboim, JETP {\bf 76},  541  (1993), [ZhETF {\bf 103}, 1105 (1993)].

1111: 

1112: \bibitem{Pa1994prl}

1113: K. Pachucki, Phys. Rev. Lett. {\bf 72},  3154  (1994).

1114: 

1115: \bibitem{EiSh1995}

1116: M.~I. Eides and V.~A. Shelyuto, Phys. Rev. A {\bf 52},  954  (1995).

1117: 

1118: \bibitem{Ka1996}

1119: S.~G. Karshenboim, J. Phys. B {\bf 29},  L29  (1996).

1120: 

1121: \bibitem{Ka1996b}

1122: S.~G. Karshenboim, JETP {\bf 82},  403  (1996), [ZhETF {\bf 109}, 752 (1996)].

1123: 

1124: \bibitem{Ka1997}

1125: S.~G. Karshenboim, Z. Phys. D {\bf 39},  109  (1997).

1126: 

1127: \bibitem{Pa2001}

1128: K. Pachucki, Phys. Rev. A {\bf 63},  042503  (2001).

1129: 

1130: \bibitem{Je2003jpa}

1131: U.~D. Jentschura, J. Phys. A {\bf 36},  L229  (2003).

1132: 

1133: \bibitem{PaJe2003}

1134: K. Pachucki and U.~D. Jentschura, Phys. Rev. Lett. {\bf 91},  113005  (2003).

1135: 

1136: \bibitem{Je2004b60}

1137: U.~D. Jentschura, Two-Loop Bethe Logarithms for Higher Excited $S$ Levels,

1138:   Phys. Rev. A (2004), at press.

1139: 

1140: \bibitem{EiSh2004}

1141: M.~I. Eides and V.~A. Shelyuto, Phys. Rev. A {\bf 70},  022506  (2004).

1142: 

1143: \bibitem{MeRi2000}

1144: K. Melnikov and T. v.~Ritbergen, Phys. Rev. Lett. {\bf 84},  1673  (2000).

1145: 

1146: \bibitem{ReEtAl2000}

1147: J. Reichert, M. Niering, R. Holzwarth, M. Weitz, T. Udem, and T.~W. H\"{a}nsch,

1148:   Phys. Rev. Lett. {\bf 84},  3232  (2000).

1149: 

1150: \bibitem{ScZu1997}

1151: M.~O. Scully and M.~S. Zubairy, {\em Quantum Optics} (Cambridge University

1152:   Press, Cambridge, 1997).

1153: 

1154: \bibitem{FiSw2005}

1155: Z. Ficek and S. Swain, {\em Quantum Interference and Coherence --- Springer

1156:   Series in Optical Sciences vol. 100} (Springer, Berlin, Heidelberg, New York,

1157:   to appear in 2005, ISBN: 0-387-22965-5).

1158: 

1159: \bibitem{EiWaHa2001}

1160: K.~S.~E. Eikema, J. Walz, and T.~W. H\"{a}nsch, Phys. Rev. Lett. {\bf 86},

1161:   5679  (2001).

1162: 

1163: \bibitem{YeInSh2004nim}

1164: V.~A. Yerokhin, P. Indelicato, and V.~M. Shabaev, Two-loop self-energy

1165:   correction to the ground-state Lamb shift in H-like ions, e-print

1166:   hep-ph/0409048.

1167: 

1168: \bibitem{Je2003plb}

1169: U.~D. Jentschura, Phys. Lett. B {\bf 564},  225  (2003).

1170: 

1171: \bibitem{DrEtAl2003}

1172: I. Dragani\'{c}, J.~R. Crespo L\'{o}pez-Urrutia, R. DuBois, S. Fritzsche, V.~M.

1173:   Shabaev, R.~S. Orts, I.~I. Tupitsyn, Y. Zou, and J. Ullrich, Phys. Rev. Lett.

1174:   {\bf 91},  183001  (2003).

1175: 

1176: \end{thebibliography}

1177: 

1178: \end{document}

1179: 

1180: