Thomas Lee Elifritz

http://arxiv.org/abs/1207.6061

Self Energy and Fluctuation Spectra in Cuprates: Comparing Optical and Photoemission Results, R.S. Markiewicz, Tanmoy Das, A. Bansil

We compare efforts to extract self energies and fluctuation spectra of the cuprates using optical and photoemission techniques. The fluctuations have contributions from both the coherent and incoherent parts of the band, which are spread over the full bare bandwidth of >2eV. Many experimental studies concentrate on the coherent part of the band and hence miss higher energy fluctuations. Our study establishes the universal presence of high energy bosonic fluctuations across various spectroscopies as a key ingredient in the high temperature superconducting cuprates.

We find that an important contribution to the high-energy fluctuations has been ‘hiding in plain sight’. And we provide an example of an optical glue recovery which includes the high energy contribution, and which bears a striking similarity to the calculated results. Our study thus resolves a puzzling discrepancy between optical vs other experiments related to the nature of bosonic fluctuations, and clearly demonstrates that experimental attempts to extract the optical glue need to probe a higher energy range to weigh in on the issue of a possible U-scale glue involved in the mechanism of high-Tc superconductivity.

In conclusion, our study provides a number of insights into attempts to derive self energies and glue functions from experimental studies of the cuprates. We find a surprising variability in the magnitude of self energy reported in different studies. We have identified possible sources and recommend use of first principles dispersions in the analyses to minimize the problem. With respect to specifically optical studies, we find that the self energy is relatively momentum independent, so these studies should be useful for extracting momentum averaged fluctuation spectra. A possible weak link is relating the optical spectrum to an underlying self energy, as the usual Σ_op is found to deviate from the true Σ at energies above 1 eV.

Most importantly, we have shown that in the overdoped regime the optical spectra of the cuprates should be described in a single band model for energies below 2.5 eV. When this is done, the resulting fluctuation spectrum or glue function displays substantial spectral weight in the high energy region extending to 1.5 eV. Our study thus finds additional high energy bosonic fluctuations in the optical spectra and reconciles a puzzling discrepancy in this regard involving optical and other spectroscopies. In the conventional terminology of optical studies, these contributions should also be considered as part of the optical ‘glue’ function.

Of course, from optical studies there is no way of determining which of the observed fluctuations promote d-wavesuperconductivity, which play no role in superconductivity, and which are actually pairbreaking. Nevertheless, we must abandon the common perception that optical studies “prove” that only low frequency bosonic fluctuations are important for high-Tc superconductivity – particularly when those studies are restricted to energies below 1 eV.

Read it and weep.

lifeform@charter.net