Determination of Surface Recombination Velocities at Contacts in Organic Semiconductor Devices Using Injected Carrier Reservoirs Oskar J. Sandberg, * Simon Sandén, Anton Sundqvist, Jan-Henrik Smått, and Ronald Österbacka Center for Functional Materials and Faculty of Science and Technology, Åbo Akademi University, Porthaninkatu 3, 20500 Turku, Finland (Received 31 August 2016; revised manuscript received 10 December 2016; published 15 February 2017) A method to determine surface recombination velocities at collecting contacts in interface-limited organic semiconductor devices, based on the extraction of injected carrier reservoirs in a single-carrier sandwich-type structure, is presented. The analytical framework is derived and verified with drift-diffusion simulations. The method is demonstrated on solution-processed organic semiconductor devices with hole-blocking TiO 2 =organic and SiO 2 =organic interfaces, relevant for solar cell and transistor applications, respectively. DOI: 10.1103/PhysRevLett.118.076601 Contacts play a crucial part in thin-film semiconductor devices, such as those based on organic and perovskite semiconductors. Most electronic devices require at least one contact that is either charge collecting or blocking. For instance, in organic field effect transistors, the source and drain constitute collecting contacts, while the gate electrode, covered with an insulating dielectric, needs to be blocking [1]. Many applications also require selective contacts that are able to efficiently either inject or collect one type of charge carrier while simultaneously blocking the other type. This is of particular importance in organic and perovskite solar cells, where contacts that are able to efficiently collect majority carriers, while simultaneously blocking minority carriers, are desired [2–9]. However, a comprehensive understanding of the processes taking place at the contacts in organic thin-film semiconductor devices is still lacking. The current of carriers flowing out from the semi- conductor (to the electrode) at a collecting contact is generally described in terms of an effective surface recom- bination current [10–13]: J R ¼ qS R ½n c - n 0 ; ð1Þ where S R is the associated surface recombination velocity, n c is the carrier density at the surface and n 0 is the corresponding equilibrium density, and q is the elementary charge. The surface recombination velocity is a character- istic for the quality of the surface and can be expressed as S R ¼ σ R v R N s , where σ R is an effective capture cross section, v R is the carrier emission velocity at the contact, and N s is the surface density of recombination centers [3,10]. At an ideal semiconductor-metal contact, acting as an infinite recombination center (σ R N s → 1), the upper limit of S R is typically on the order of 10 6 cm=s at temperature T ¼ 300 K in accordance with the thermionic emission theory [4,10]. A schematic picture of surface recombination for holes at a contact is shown in Fig. 1(a). In general, if S R is larger than the effective transport velocity v D ∼ μjFj of carriers within the semiconductor layer (S R ≫ v D ), the carrier collection is limited by the bulk (diffusion limited) [9,14–16]. In this case, the contact virtually acts as a perfect collector (S R → ∞). Here, μ is the carrier mobility and F the electric field. If S R <v D , on the other hand, the charge collection is controlled by kinetics at the contact as the carrier transport becomes limited by the interface [9,10,16,17]. The surface recombination at a contact that is blocking from the viewpoint of carrier collection is by definition also interface limited (S R ≪ v D ). However, the condition S R ≪ v D alone does not necessarily fulfill the requirements of a blocking contact. Ideally, a blocking contact with S R ¼ 0 is achieved by inserting an interlayer that prevents all carriers from leaving the device at the contact (σ R N s → 0). In practice, however, FIG. 1. (a) Schematic picture of surface recombination at a semiconductor-electrode contact for holes being collected at a metal electrode. In (b) and (c), a schematic picture of the CELIV technique is shown. A linear voltage pulse uðtÞ is applied to extract charge carriers at the dc voltage V. From the correspond- ing extraction current transient jðtÞ (corrected for the steady-state current), the extracted charge is obtained from Q extr ¼ R t extr 0 ½jðtÞ - j 0 dt, where j 0 ¼ðεε 0 =dÞðdu=dtÞ¼ðεε 0 u max =dt pulse Þ. PRL 118, 076601 (2017) PHYSICAL REVIEW LETTERS week ending 17 FEBRUARY 2017 0031-9007=17=118(7)=076601(5) 076601-1 © 2017 American Physical Society