Efficient infiltration of low molecular weight polymer in nanoporous TiO 2 Erik M.J. Johansson a,⇑ , Sulena Pradhan a , Ergang Wang b , Eva L. Unger a , Anders Hagfeldt a , Mats R. Andersson b a Department of Physical and Analytical Chemistry, Uppsala University, SE-751 05 Uppsala, Sweden b Department of Chemical and Biological Engineering, Polymer Technology, Chalmers University of Technology, SE-412 96 Göteborg, Sweden article info Article history: Received 19 October 2010 In final form 21 December 2010 Available online 23 December 2010 abstract The polymer APFO3 was prepared with different molecular weights to study how the infiltration into nanoporous TiO 2 films of different thickness depends on the size of the polymer. Also two different sizes of TiO 2 nanoparticles were investigated to understand the effect of different pore size. It was observed that the lowest molecular weight polymer dissolved in chlorobenzene could infiltrate the nanoporous TiO 2 network up to several micrometer thick films. It was concluded that efficient polymer infiltration into thick nanoporous layers was possible for the polymers with an estimated average chain length smal- ler than the diameter of the nanoparticles. Ó 2010 Elsevier B.V. All rights reserved. 1. Introduction Polymer/fullerene based organic solar cells [1,2] and dye-sensi- tized solar cells (DSC) [3] are promising alternatives to conven- tional solar cells. The DSC is based on an inorganic network of nanoparticles sensitized with dye-molecules and a liquid electro- lyte with a redox couple transporting charges to the dye from the backcontact. The liquid electrolyte may be replaced by a solid state hole-conducting material, which may be advantageous [4]. The dominating hole-conductor material in solid state dye-sensi- tized solar cells is 2,2 0 ,7,7 0 -tetrakis (N,N-di-p-methoxypheny- amine)-9,9 0 -spirobifluorene (spiro-MeOTAD) [4,5], meanwhile, other hole-conducting materials have also been used (see e.g. [6–10]) although resulting in less efficient devices. Hole-conduct- ing polymers that are usually used in the polymer/fullerene organ- ic solar cells have also shown to be rather effective in the dye-sensitized solar cells [8–16]. Also, the combination of poly- mers and TiO 2 or ZnO nanoparticles, without dye molecules, has been shown to give a photocurrent, where the polymer in addition to the hole conduction also is used to absorb the light and inject electrons to the nanoparticles, i.e. the polymer act both as a light absorber and hole-conductor [17–25]. In the solid state dye-sensitized solar cell, the nanoparticles are fused together in a sintering process and thereafter the hole-con- ductor material is infiltrated into the porous structure. An impor- tant task is to infiltrate the hole-conductor into the nanoporous structure to be able to transport all the photogenerated holes to the contact. The spiro-MeOTAD hole-conductor molecules, spin-coated on TiO 2 nanoporous electrodes, have been shown to infiltrate the nanoporous structure rather efficient [26–28] and other smaller hole-conductor molecules have been shown to be possible to infiltrate also by melting [29]. However, it has been dis- cussed whether the large polymer molecules can infiltrate the nanoporous structure or not [17,30,31]. In this report we investigate if a short polymer can be infiltrated into a TiO 2 nanoporous material of different thickness. We also study if the infiltration of the polymer depends on the molecular weights of the polymer and investigate the effect of two different sizes of the nanoparticles. Absorption spectroscopy and scanning electron microscopy (SEM) measurements were used to obtain information about the infiltration. It is very important to elucidate how efficient polymers can infiltrate such nanoporous materials, how thick nanoporous layers that may be infiltrated and the dependence on the polymer molecular weights, to be able to de- sign more efficient solar cell devices based on these hybrid combi- nations of nanoparticle networks and polymers. 2. Results and discussion The polymers poly(2,7-(9,9-dioctyl-flourene)-alt-5,5-(4 0 ,7 0 -di-2- thienyl-2 0 ,1 0 ,3 0 -benzothiadiazole)) [32,33], were prepared with dif- ferent molecular weights; polymer A: number average molecular weight (M n ) = 2000, and weight average molecular weight (M w )= 2700; B: M n = 3900, M w = 6000; C: M n = 6000, M w = 14 000; D: M n = 15 000, M w = 24 000 measured from Gel Permeation Chromatography (GPC) with polystyrene as a reference (see Table 1). In Figure 1 the absorption spectra of the polymers in a chlorobenzene solution are shown together with the molecular structure. We observe that the absorption spectra of the polymers 0009-2614/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2010.12.062 ⇑ Corresponding author. Fax: +46 (0)18 471 3633. E-mail address: Erik.Johansson@fki.uu.se (E.M.J. Johansson). Chemical Physics Letters 502 (2011) 225–230 Contents lists available at ScienceDirect Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett