Synthesis, electronic properties and self-assembly on Au{111} of thiolated (oligo)phenothiazines

Summary (Oligo)phenothiazinyl thioacetates, synthesized by a one-pot sequence, are electrochemically oxidizable and highly fluorescent. SAMs can be readily formed from thiols prepared by in situ deprotection of the thioacetates in the presence of a gold-coated silicon wafer. Monolayer formation is confirmed by ellipsometry and the results compared to those obtained by force field and DFT calculations.

consequence of their electronic properties. In particular, their reversible formation of stable radical cations [26][27][28][29][30][31], their tunable redox and fluorescence properties [32][33][34], and their tendency to self-assemble on surfaces by π-π interactions [35] make them eligible for use as redox-switchable molecular entities. In addition, the inherent folded conformation of phenothiazines [36], with a folding angle of 158.5°, represents an intriguing new aspect for the formation of self-assembled monolayers (SAMs) of this class of compounds. Furthermore, the transformation of phenothiazines into stable planar radical cations with excellent delocalization [37] qualifies them as excellent models for switchable conductive or semiconductive molecular wires. Encouraged by successful electrode modifications with conjugated thiolated anilines [38] and SAM formation of thiolated phenylethynyl phenothiazines [39], and in continuation of our investigations directed towards the synthesis and study of (oligo)phenothiazine-based functional π-systems [40][41][42][43][44][45][46], we have now focused our attention on thiolated phenothiazines and (oligo)phenothiazines as "alligator-clips". Here, we report the synthesis of phenothiazines and their oligomers bearing "alligator-clips" and their electronic properties as studied by cyclic voltammetry (CV), spectroscopic and spectrometric methods. Furthermore, their chemisorption and SAM formation on Au{111} were studied by ellipsometry.

Synthesis
The facile bromine-lithium exchange of bromo phenothiazines [47][48][49] and the subsequent electrophilic trapping reactions of the resulting lithio phenothiazines [50,51] with different electrophiles set the stage for a straightforward synthesis of thiolated (oligo)phenothiazines. Therefore, the synthesis of thiofunction-alized phenothiazines can be accomplished according to a standard protocol [18]. Thus, solutions of bromo phenothiazines 1 [32,52] were cooled to −78 °C and reacted with n-BuLi (1a and 1b) or t-BuLi (1c-e and 3), respectively, to give the corresponding lithio phenothiazines via bromine-lithium exchange. Subsequent addition of elemental sulfur, followed by stirring for 5 min at −78 °C, and the addition of freshly distilled acetyl chloride furnished the desired (oligo)phenothiazinyl thioacetates 2 and 4 in moderate to good yields (Scheme 1). However, in the case of dyad 1c thiolation was only accomplished by addition of acetylsulfur chloride [53] to the lithio species at low temperature, albeit the thiofunctionalized derivative 2c was obtained in only 15% yield. The structures of the (oligo)phenothiazinyl thioacetates 2 and 4 were unambiguously supported by 1 H and 13 C NMR spectroscopy, mass spectrometry and elemental analysis.

Electronic properties
The electronic properties of the (oligo)phenothiazinyl thioacetates 2 and 4 were investigated by absorption and emission spectra, and cyclic voltammetry (Table 1). Optical spectroscopy (UV-vis and fluorescence spectra) revealed that only the triad 2d and the tetrad 2e displayed considerable fluorescence with emission of greenish-blue light and large Stokes shifts ( Figure 1, Δ 6400-6600 cm −1 ). While the absence of fluorescence of monophenothiazines 2a, 2b, and 4 with heavy atom substitution and consequently, increased spin-orbit coupling is not too surprising, the presence of a diphenothiazine unit (2c) is not sufficient. Hence, at least two covalently bound phenothiazines without an additional sulfur substituent appears to be the prerequisite for intense fluorescence of oligophenothiazinyl thioacetates.  Electrochemical data for (oligo)phenothiazinyl thioacetates 2 and 4 were obtained by cyclic voltammetry in the anodic region (scan area up to 1.5 V). The reversible first oxidations to the radical cations of monophenothiazines 2a, 2b, and 4 were shifted anodically in comparison to unsubstituted monophenothiazines [54] as a consequence of the electron-withdrawing nature of the thioacetate. Due to unsymmetrical substitution, the dyad 2c showed two distinctly separated, reversible oxidations at E 0 0/+1 = 668 mV and E 0 +1/+2 = 853 mV. The cyclic voltammogram of the triad 2d displayed three distinctly separated, reversible oxidations at E 0 0/+1 = 608 mV, E 0 +1/+2 = 765 mV, and E 0 +2/+3 = 876 mV ( Figure 2). However, the electrochemistry of the tetrad 2e is more complicated. Only three distinctly separated, reversible oxidations were evident. The first oxidations at E 0 0/+1 = 597 mV and E 0 +1/+2 = 690 mV are in accordance with Nernstian behavior, while the third oxidation at E 0 = 842 mV reveals a large difference of ΔE = 132 mV for the current peaks of the oxidation and the reduction wave. Presumably, the expected third and fourth oxidations coincide and give rise to a combined quasi-reversible peak.

Self-assembly and ellipsometry
SAMs on a Au{111}-coated silicon wafer substrate were prepared from (oligo)phenothiazinyl thioacetates 2 or 4 by in situ saponification with degassed aqueous ammonia in THF at room temperature for 24 h (Scheme 2).
Based upon thorough surface analysis of the previously studied thiolated phenylethynyl phenothiazines chemisorbed on Au{111} by ellipsometry, contact angle measurements, X-ray photoelectron spectroscopy, and infrared reflection absorption spectroscopy (IRRAS) [39], we applied ellipsometry in combination with molecular modeling at the force field and DFT levels of theory for the characterization of SAMs of in situ liberated (oligo)phenothiazinyl thiols on Au{111}. The ability of the molecules to form SAMs was investigated by solution adsorption of different systems onto gold films of 100 nm thickness Scheme 2: Preparation of SAMs from (oligo)phenothiazinyl thioacetates 2 or 4 on a Au{111}-coated silicon wafer substrate. thermally evaporated onto Si wafers using 10 nm of Ti as adhesion promoter. This procedure is known to yield polycrystalline gold films with preferential {111} orientation [55].
The thickness of the layer was determined by ellipsometry as described above. As an estimate for the molecular dimensions of the monolayers, the structures of the (oligo)phenothiazines 2 and 4 were computed at the MM2 and DFT levels of theory (Table 2) [56].
To minimize computational time in the latter calculations, the hexyl substituents were truncated to methyl groups. From these calculations, the theoretical layer thickness was calculated according to d th = l mol cos φ + l Au-S , where l mol is the calculated length of the respective molecule, φ is the molecules' tilt angle with the surface normal, and l Au-S = 2.1 Å is the Au-S bond length [57]. For φ, we refer to a recent electron spectroscopic analysis on similar aromatic systems, which determined φ = 23° for anthracene-2-thiol [58]. Using this value, we made the reasonable assumption that the Au-S-C bond is mainly influenced by the adjacent phenyl system. Table 2 shows d th for the different molecules along with the experimental thickness d exp as determined by ellipsometry. The theoretical thicknesses are given for MM2 as well as DFT calculations. As a simple measure of monolayer formation of the different systems, the relative coverage θ obtained experimentally is calculated from θ = d exp /d th as given in Table 2. From these values it is clear that of 2, only 2a and 2d show good SAM formation, suggesting an odd-even effect on film growth, which might be related to steric hindrance during adsorption when an even number of phenothiazine units are present, e.g., because of a back bending of the thiol-bound molecule to the gold surface in these cases, supported by additional gold-π-interactions with the terminal phenothiazine, which thus would hamper the formation of a SAM with an almost parallel intermolecular orientation. In corroboration of such disorder effects, coverage seems to decrease with increasing molecule length for even-numbered molecules (cf. Table 2). The highest coverage was obtained with 4, which is not surprising, because the thiol bifunctionality allows chemisorption of the molecule at either side, which reduces the impact of steric effects on the adsorption kinetics and thus may lead to a more densely packed film. As a consequence, thiolated mono-and terphenothiazines 2 (n = 1, 3) and the dithiolated derivative 4 can be easily self-assembled to give stable monolayers on gold surfaces. This feature makes this class of redox-active molecular entities highly interesting for the fabrication of functionalized electroactive surfaces and nanostructured devices.

Conclusion
In summary we have shown a concise, general synthetic access to (oligo)phenothiazinyl thioacetates that are suitable precursors for the formation of thiol-bound (oligo)phenothiazines on gold surfaces. Whereas the first oligomers are non-fluorescent, the triad and the tetrad display intense greenish-blue fluorescence in addition to distinct multiple reversible oxidation. The in situ deprotection of the thioacetates to thiols in the presence of a gold-coated silicon wafer was used to prepare self-assembled monolayers, which were unambiguously characterized by ellipsometry and accompanying force field and DFT calculations. The chemical trigger of gradual thiol liberation enables better control of film formation and adsorption kinetics, which can be very useful, for example, for co-adsorption of the moieties with a second, nonconductive molecule, which serves as an insulating matrix. Further studies directed toward such morecomplex (oligo)phenothiazine SAMs on gold and functionalized redox manipulable surfaces, the nanoscopic characterization of the monolayers as well as their manipulation with external stimuli are currently underway.

Experimental General considerations
Reagents, catalysts, ligands, and solvents were purchased reagent grade materials and used without further purification. THF and acetyl chloride were dried and distilled according to standard procedures [59]. The bromo phenothiazines 1a [50,51], 1b [50,51], 1c and 1d [32,52] and 3 [50,51], and acetylsulfur chloride [53] were prepared according to literature procedures.  [50,51], and 2.45 g (17.7 mmol) potassium bicarbonate were dissolved in 100 mL of DME and 20 mL of water. The mixture was degassed by purging with argon gas for 20 min. After the addition of 136 mg of tetrakis(triphenylphosphan)palladium (118 μmol, 4 mol %), the reaction mixture was stirred for 12 h at 85 °C. After cooling to room temperature, 50 mg of Na 2 SO 3 was added and the reaction mixture stirred for 14 h at room temperature. Then, 3.17 g (6.49 mmol) of 3-bromo-10-hexyl-7-iodo-10H-phenothiazine [51] was added and the mixture stirred for 4 d at 85 °C. After the addition of 100 mL of water, the crude product was extracted several times with dichloromethane. The combined organic phases were dried with magnesium sulfate and the solvents removed in vacuo. The residue was chromatographed on silica gel (hexane/acetone 50:1) to give 1.53 g (36%) of 1e as a yellow resin. 1

SAM preparation and ellipsometry
The (oligo)phenothiazinyl thioacetates 2a, 2c-e, and 4 were dissolved under an argon atmosphere in dry THF to give a 10 −4 M solutions. Au-coated silicon wafers (surface area: 1 cm 2 ) were placed in these solutions. Upon the addition of a few drops of a concentrated solution of aqueous ammonia the thioacetates were saponified to liberate the thiol functionality necessary for chemisorption and SAM formation on gold. After 24 h the wafers were removed from the solution and rinsed several times with dry THF.
The thickness of the formed organic layers was determined by means of spectral ellipsometry (M-44, J.A. Woollam, USA) applying a 3-layer model consisting of gold substrate, organic layer, and ambient [62]. The organic layer was described by means of a Cauchy model, with the first two Cauchy parameters chosen such to yield a refractive index of 1.490 at 500 nm, which resulted from a study on biphenylthiolates on gold in excellent agreement with theory [63].