© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1433 www.advmat.de www.MaterialsViews.com CORRESPONDENCE wileyonlinelibrary.com Adv. Mater. 2011, 23, 1433–1435 Torsten Schwich, Marie P. Cifuentes, Paul A. Gugger, Marek Samoc, and Mark G. Humphrey* Electronic, Molecular Weight, Molecular Volume, and Financial Cost-Scaling and Comparison of Two-Photon Absorption Efficiency in Disparate Molecules (Organometallic Complexes for Nonlinear Optics. 48.) – A Response to “Comment on ‘Organometallic Complexes for Nonlinear Optics. 45. Dispersion of the Third-Order Nonlinear Optical Properties of Triphenylamine-Cored Alkynylruthenium Dendrimers.’ Increasing the Nonlinear Response by Two Orders of Magnitude.” The Comment by Pérez-Moreno and Kuzyk [1] and our original paper [2] both focus inter alia on the crucially important problem of comparing nonlinear absorption efficiency across different types of material. Pérez-Moreno and Kuzyk stress the need for normalization of the nonlinear parameters in order to perform comparisons of nonlinear optical (NLO) merit. For reasons of convenience, many authors have scaled NLO data by molec- ular weight, [3] but such comparisons are biased against metal- containing molecules (a point that we made in the original manuscript), so a general approach to compare disparate molecules is needed to facilitate materials improvement. Pérez-Moreno and Kuzyk are correct in saying that it is more sensible from a theoretical perspective to compare the two-photon absorption (TPA) merit of chromophores by scaling by the square of the number of electrons rather than just the number of electrons. However, scaling the molecular TPA coefficient σ 2 by the number of electrons contributing to the response, as we did in the original paper, and as they have improved upon in the preceding Comment, is arguably somewhat arbitrary because it assumes that electrons in a molecule can be clearly separated into those that form a “con- tiguous” system and thereby contribute to the nonlinearity, and those that do not contribute to the nonlinearity. This may be defensible when one distinguishes between σ electrons and π electrons, but in cases where free electron pairs and metal valence electrons contribute to the delocalization/conjugation, such separation may be dubious. One other concern is that, in certain molecular assemblies, the contiguous π-system inter- actions may not be solely responsible for the NLO merit: there may be interactions “through space”, a possibility that certainly exists in dendrimers where the arms may “talk” to each other through exciton interactions as well as through the π-network. There are other options for comparing NLO results of dis- parate chemicals. One possibility is to scale NLO data by the molecular volume V m , a justification being that compounds or materials showing NLO effects may be used in devices or struc- tures for which decreasing the size is important (e.g. lumines- cent nanoparticles) and for which a (NLO merit)/ V m parameter will be crucial. A second possibility is to consider the cost of production, which would also be of interest in “downstream” applications. Neither approach has been explored previously, so we have taken the first steps to calculation of the volume-scaled and cost-scaled nonlinear absorption of the organic and organo- metallic dendrimers discussed in our earlier paper ( Figure 1), [2,3] and contrasted the outcomes with those obtained by the mole- cular weight scaling and “number of effective electrons” scaling procedures in the Comment and our earlier paper. The V m values in this Response were calculated using a variety of computational methods (Supporting Information). The cost of production of 1 mmol of each dendrimer has been evaluated from the published syntheses by considering all required starting mate- rials, the labour cost of synthesis, and the yields of reactions (Sup- porting Information). While there are approximations involved in both scaling procedures (perhaps comparable to the errors asso- ciated with the nonlinear absorption experiments that afforded the σ 2 data, viz ± 20%), they do provide additional insight into comparative NLO efficiency of disparate chemicals, and the broad trends are not strongly dependent on the volume or economic cost choices that were made when developing these comparisons. The Figure 2 left and right panels compare the zero- and first-generation organometallic and organic dendrimer pairs DOI: 10.1002/adma.201004348 T. Schwich, Dr. M. P. Cifuentes, P. A. Gugger, Prof. M. G. Humphrey Research School of Chemistry Australian National University Canberra, ACT 0200, Australia E-mail: Mark.Humphrey@anu.edu.au Prof. M. Samoc Institute of Physical and Theoretical Chemistry Wroclaw University of Technology 50–370 Wroclaw, Poland