In situ study of pentacene interaction with archetypal hybrid contacts: Fluorinated versus alkane thiols on gold

One approach developed to improve the performance of bottom contact source/drain electrodes is to treat the contacts with thiols before deposition of the semiconductor. There is evidence indicating that improvement is due to both morphological effects and improved work function matching. Especially suggestive evidence shows that thiols that increase the effective work function of the contacts (e.g., fluorinated thiols) yield better device performance than work function decreasing thiols (e.g., alkane thiols). Here, we compare two technologically relevant thiol treatments, an alkane thiol (1-hexadecanethiol), and a fluorinated thiol (pentafluorobenzenethiol), in pentacene organic field effect transistors. Using in situ semiconductor deposition, x-ray photoemission, and x-ray absorption spectroscopy, we are able to directly observe the interaction between the semiconductor and the thiol-treated gold layers. Our spectroscopic analysis suggests that there is not a site-specific chemical reaction between the pentacene and the thiol molecules. A homogeneous standing-up pentacene orientation was observed in both treated substrates, consistent with the morphological improvement expected from thiol treatment in both samples. Our study shows that both the highest occupied molecular orbital–Fermi level offset and $\text{C}1s$ binding energy are shifted in the two thiol systems, which can be explained by varied dipole direction within the two thiols, causing a change in surface potential. The additional improvement of the electrical performance in the pentafluorobenzenethiol case is originated by a reduced hole injection barrier that is also associated with an increase in the density of states in the lowest unoccupied molecular orbital.

Figure 1

The valence band edge measured on the (a) hexadecanethiol (Hex-thiol) and (b) pentafluorobenzenethiol (PFB) treated samples. The HOMO edge shifts from 0.5 eV in pentacene/Hex-thiol/Au sample to 0.1 eV in pentacene/PFB/Au sample, showing decreased hole injection barrier from Au contact to pentacene in the latter case which leads to an improved contact between the semiconductor and the source/drain.

Figure 2

XPS taken exciting the $1s$ carbon orbital on the (a) hexadecanethiol and (b) pentafluorobenzenethiol samples. The shape of peaks in both samples resembles the shape of peaks in gas-phase and thin-film-phase pentacene formed on nondipole interface. The shift of the $\text{C}1s$ peaks in pentacene/PFB/Au to the lower BE is caused by the dipole in the pentacene/PFB/Au.

Figure 3

NEXAFS taken on the (a) pentafluorobenzenethiol with a monolayer of pentacene, (b) hexadecanethiol with a monolayer of pentacene, (c) thick pentacene $\left(>10\text{nm}\right)$ on parylene, (d) pentafluorobenzenethiol only on Au, and (e) hexadecanethiol only on Au.

Figure 4

(Color online) Detailed comparison of the Pn/PFB and PN/HEX LUMO states (open markers in the top and bottom panels, respectively) with the pentacene bulk states from the thick film (filled markers in the middle panel). Multiple component fitting lines are superimposed on the experimental data points. We first fitted the thick film to Voigt functions (full lines) with free-fitting parameters. Then, the NEXAFS spectrum of the Pn/PFB SL was fitted by constraining the LUMO and $\text{LUMO}+1$ states to have the same energy position and shape parameters of the corresponding peaks in the thick film (i.e., only the peak intensity was left as fitting parameter). In the Pn/PFB spectrum, an additional peak (shadowed) accounts for the increase in the density of states at the NEXAFS edge. The fitting constraints were relaxed for the Pn/HEX spectrum in the 285–288 eV range because of the substrate contribution to the NEXAFS spectrum.

Figure 5

A schematic of dipole configuration within pentafluorobenzenethiol [(a) dipoles are facing downward to the substrate and hexadecanethiol and (c) dipoles are facing upward from the substrate]. (b) and (d) are the charge, electrical field and potential within the pentafluorobenzenethiol and hexadecanethiol, respectively. The difference of potential on the top of the two thiols $\left(\left|{\varphi }_1-{\varphi }_2\right|\right)$ with respect to Au is estimated to be 0.4–1.1 eV.

Figure 6

The energy level model proposed for Au/thiol/pentacene system. In (a), various samples are independently aligned with the vacuum level. Pentafluorobenzenethiol increases the work function of the Au surface while hexadecanethiol slightly decreases the work function of the Au surface. (b) and (c) are bands alignments estimated based on the discussion for pentacene/PFB/Au and pentacene/Hex-thiol/Au samples, respectively, in the text. PFB/Au has improved hole injection barrier as indicated.