Comment on ‘‘Determination of the Exciton Binding Energy in Single-Walled Carbon Nanotubes’’ Recently, Wang et al. [1] used Raman spectroelectro- chemistry for the determination of the binding energy of the exciton, associated with the transition between the second pair of van Hove singularities in semiconducting single-walled carbon nanotubes (SWCNTs). The central idea was that the electronic energy gap, in which the Raman intensity of the radial breathing mode (RBM) remains unaffected, is a superposition of exciton binding energy and the energy of the incident or scattered photons. The attenuation of RBM intensity was followed by electro- chemical charging of SWCNTs deposited on an ITO (in- dium tin oxide) working electrode. A single Pt electrode served both as the reference electrode and the counter- electrode. Presumably, the electrolyte solution was aque- ous and contacted air. Although the found exciton binding energies for tubes 10; 3 and 7; 5, respectively, were comparable to those from independent studies, one can worry that the agreement was just accidental. The potential of the Pt ‘‘reference’’ electrode is poorly defined by a complex redox equilibrium between Pt, sur- face oxides on Pt, H  , OH  , H 2 O, dissolved O 2 , etc. This electrode potential is unstable, even if no net current flows through the Pt/electrolyte solution interface. The voltage applied between the Pt and ITO/SWCNT electrodes ranged from 0:73 to 2.8 V [1]. The standard redox potentials of H  =H 2 and O 2 =H 2 O redox couples are 0 and 1.229 V, respectively. Hence, Faradaic current flows through the electrochemical cell at certain applied voltages outside this region, even if we admit non-negligible overvoltages on Pt and ITO/SWCNT electrodes. The electrode reactions depend on pH, which was, probably, not buffered. The changes of local pH near the electrode surface can only be negligible, if the Faradaic reactions, such as oxygen and hydrogen evolution, are strongly retarded, but this cannot be assessed because the effective current is unknown. Moreover, the overvoltages of Faradaic reactions on the ITO surface and the SWCNT deposit may be different. Hence, the actual electrochemical potential applied at SWCNTs is out of any control. Wang et al. [1] further state that Pt forms an Ohmic contact to SWCNTs, and this, reportedly, allows an equat- ing of the applied potential on the working electrode with the movement in the Fermi level (E F ). This statement is unclear for two reasons. First, SWCNTs were, actually, deposited on ITO and not on Pt in Ref. [1]. Second, the Fermi level shift does not seem to be that easily derivable from the applied potential [2 – 4]. Independent studies re- ported that the proportionality constant between E F and the applied electrochemical voltage was between 0.3 to 0:7 eV=V [2,3]. The broad region of potentials, where the RBM intensity of semiconducting SWCNTs was unaf- fected (2.4 to 2.6 V) is exceptional in the Letter [1] com- pared to other reports. For instance, Ref. [5] shows that this potential region hardly exceeds 1 V for bundled SWCNTs, and is even narrower for single isolated SWCNTs [6]. Hence, the suggested [1] calculation of exciton binding energy would lead to meaningless values. Wang et al. [1] further reported that the frequency of the G-band and D-band phonons did not change with potential in the whole region of applied voltages. However, other authors report that the p doping (anodic charging) causes significant blueshifts of the G band [2,3,7–10]. This shift expectedly indicates the stiffening of graphene modes by holes in the  band [2,7,9]. The cathodic n doping caused redshifts or blueshifts [2,3,7,8] for yet unclear reasons. But the absence of any shift in a broad region of p=n doping [1] is a true curiosity. The second order Raman modes (G 0 and D modes) do shift after electrochemical charging too [9,10], which again contradicts the data by Wang et al. [1]. There is no doubt that spectroelectrochemistry is a powerful tool for study of electronic properties of carbon nanostructures. However, the experimental conditions used in Ref. [1] are unsuitable for control of charge-transfer doping of SWCNTs. Therefore, the reported determination of exciton binding energy is questionable. Ladislav Kavan, 1, * Martin Kalba ´c ˘, 1,2 Marke ´ta Zukalova ´, 1 and Lothar Dunsch 2 1 J. Heyrovsky ´ Institute of Physical Chemistry Academy of Sciences of the Czech Republic Dolejs ˇkova 3, CZ-182 23 Prague 8, Czech Republic 2 Leibniz Institute of Solid State and Materials Research Helmholtzstraße 20, D-01069 Dresden, Germany Received 24 February 2006; published 4 January 2007 DOI: 10.1103/PhysRevLett.98.019701 PACS numbers: 78.67.Ch, 71.20.Tx, 71.35.y, 78.30.Na *To whom correspondence should be addressed. Electronic address: Kavan@jh-inst.cas.cz [1] Z. Wang, H. Pedrosa, T. Krauss, and L. Rothberg, Phys. Rev. Lett. 96, 047403 (2006). [2] P.M. Rafailov, J. Maultzsch, C. Thomsen, and H. Kataura, Phys. Rev. B 72, 045411 (2005). [3] S. B. Cronin et al., Appl. Phys. Lett. 84, 2052 (2004). [4] I. Heller et al., J. Am. Chem. Soc. 128, 7353 (2006). [5] L. Kavan, M. Kalbac, M. Zukalova, and L. Dunsch, J. Phys. Chem. B 109, 19 613 (2005). [6] K. Okazaki, Y. Nakato, and K. Murakoshi, Phys. Rev. B 68, 035434 (2003). [7] L. Kavan et al., J. Phys. Chem. B 105, 10 764 (2001). [8] L. Kavan and L. Dunsch, Chem. Phys. Chem. 4, 944 (2003). [9] P. Corio et al., Chem. Phys. Lett. 370, 675 (2003). [10] P.M. Rafailov and C. Thomsen, J. Optoelect. Adv. Mat. 7, 461 (2005). PRL 98, 019701 (2007) PHYSICAL REVIEW LETTERS week ending 5 JANUARY 2007 0031-9007= 07=98(1)=019701(1) 019701-1  2007 The American Physical Society