Island size evolution and molecular diffusion during growth of organic thin films followed by time-resolved specular and off-specular scattering

We report on a combined off-specular and specular x-ray scattering growth study of ultrathin films of the prototypical organic semiconductor diindenoperylene (DIP, ${\mathrm{C}}_{32}{\mathrm{H}}_{16})$. We investigate the evolution of the in-plane correlation length and the growth kinetics of the films including their dependence on the substrate temperature and the growth rate. We observe a temperature-dependent collective rearrangement of DIP molecules from a transient surface induced to the thin-film phase, which can be rationalized by incorporating a thickness-dependent out-of-plane lattice parameter. We further observe that the nucleation behavior of DIP changes from the first to the second monolayer, which we relate to a difference in the diffusion length of the molecules.

Figure 1

Scattering geometry, typically used in GISAXS experiments. Recording data with submonolayer time resolution allows one to follow the growth in real time. A gradual filling of successive layers leads to an oscillating intensity of the vertical streaks in the image series. The separation $q_{y,\max }$ of these features around the Yoneda wings can be related to the average island-to-island distance $L$ and for a known molecular coverage in the layer also to the average island diameter $l$.

Figure 2

(a) Scattering intensity at the critical angle (Yoneda wings) as a function of adsorbed material for DIP grown at $T_{\mathrm{sub}}=25$ ${}^{\circ }\mathrm{C}$ and $R_{\mathrm{growth}}=0.1$ nm/min. Detector images from which the horizontal GISAXS sections have been extracted, were integrated for 20 s per frame leading to a resolution of $\simeq 15$ frames/ML. (b) Horizontal line sections [extracted from the GISAXS images in (a)] at coverages of $\theta =0.5$, $1.5$, and $2.5$ ML. It is clearly observed that the separation between the two correlation peaks decreases for higher film coverages. As the growth progresses, more material is adsorbed on the surface and the correlation peaks become more intense.

Figure 3

(a)–(c) Specular oscillations at the anti-Bragg point (scatter plots) for DIP grown at $T_{\mathrm{sub}}=25$, 50, and 100 ${}^{\circ }\mathrm{C}$ and $R_{\mathrm{growth}}=0.4$ nm/min. The data were fitted using the prevailing model (PM) but show a considerable deviation during the filling of the second ML. However, the fit improved remarkably once the refined model (RM) has been applied. (d) Using the refined model the thickness dependence of the out-of-plane lattice parameter was determined for three different temperatures, showing an orientational transition from the transient surface induced to the thin-film phase. Importantly, the transition is delayed for higher substrate temperatures.

Figure 4

(a) Ex situ postgrowth XRR scans (open symbols) of DIP grown at a substrate temperature of $T_{\mathrm{sub}}=25$ ${}^{\circ }\mathrm{C}$ and a growth rate of $R_{\mathrm{growth}}=0.1$ nm/min. The film thickness was varied between $0.5\text{and}5.5$ ML. Solid lines show fits based on the Parratt formalism. (b) Corresponding electron densities plotted against the nominal film thickness. (c) Out-of-plane lattice parameters obtained from the XRR data shown in panel (a) (open circles) and the respective tanh-functional fit (solid line).

Figure 5

AFM images of DIP grown at $T_{\mathrm{sub}}=25$ ${}^{\circ }\mathrm{C}$ and $R_{\mathrm{growth}}=0.1$ nm/min taken with a JPK Nanowizard II. Respective height distributions are shown in the insets. Line profiles (indicated by red lines) are shown below the respective images. The images are shown for coverages of $\sim 0.5$, $\sim 1.5$, $\sim 3.5$, and $\sim 5.5$ ML [(a)–(d)] and cover an area of $2\times 2$ $\mu$m${}^2$.

Figure 6

(a) Average island distance (left axis), and specular and diffuse intensity (right axis) obtained by fitting the GISAXS data for DIP grown at $T_{\mathrm{sub}}=25$ ${}^{\circ }\mathrm{C}$. (b) Arrhenius plot for two different growth rates and two different thicknesses. The solid lines correspond to linear fits of $R_{\mathrm{growth}}=0.4$ nm/min and the dashed lines to $R_{\mathrm{growth}}=1.1$ nm/min, respectively. (c) Effective activation energy of nucleation $E_{\mathrm{act}}^{\mathrm{eff}}$ calculated for different coverages, which are almost constant throughout the layer filling. Note that $E_{\mathrm{act}}^{\mathrm{eff}}$ was derived using Eq. (3), which is valid only for small coverages (shown here as shaded areas).