Flexible Phenanthracene Nanotubes for Explosive Detection

Abstract

Phenanthracene nanotubes with arylene-ethynylene-butadiynylene rims and phenanthracene walls are synthesized in a modular bottom-up approach. One of the rims carries hexadecyloxy side chains, mediating the affinity to highly oriented pyrolytic graphite. Molecular dynamics simulations show that the nanotubes are much more flexible than their structural formulas suggest: In 1₂, the phenanthracene units act as hinges that flip the two macrocycles relative to each other to one of two possible sites, as quantum mechanical models suggest and scanning tunneling microscopy investigations prove. Unexpectedly, both theory and experiment show for 1₃ that the three phenanthracene hinges are deflected from the upright position, accompanied by a deformation of both macrocycles from their idealized sturdy macroporous geometry. This flexibility together with their affinity to carbon-rich substrates allows for an efficient host-guest chemistry at the solid/gas interface opening the potential for applications in single-walled carbon nanotube-based sensing, and the applicability to build new sensors for the detection of 2,4,6-trinitrotoluene via nitroaromatic markers is shown.

Figure 1.

Chemical structures of the PNTs 1₂ and 1₃, implementing arylene-ethynylene-butadiynylene rims and phenanthracene walls. The t-Bu and OC₁₆H₃₃ substituents mediate compound solubility, whereas the latter also provide an affinity to HOPG for STM imaging.

Figure 2.

Top and side views of molecular models of (a) 1₂ and (b) 1₃; D₁ = 2.4 nm; D₂ = 0.8 nm; h₁ = 0.7 nm; w₁ = 0.9 nm; w₂ = 1.6 nm; D₃ = 2.4 nm²; h₂ = 0.7 nm; w₃ = 0.8 nm; w₄ = 1.6 nm.

Figure 3.

Top, perspective, and side views of (a) 1₂ and (b) 1₃ on a graphene cutout optimized at the GFN2-xTB level of theory (with hexadecyloxy side chains omitted and graphene cut to C₆₀₀H₆₀ for clarity, cf. SI). Arrows 1 and 2 indicate phenanthracene units aligned in parallel to the surface and tilted, respectively; arrows 3/6 and 4/5 indicate the lower and upper macrocycles, respectively; and arrows 7 and 8 indicate crossings of the upper and lower macrocycles.

Figure 4.

(a, d, e) Scanning tunneling microscopy images, (b) proposed supramolecular model, (c) schematic model of the bottom rim, and (f) molecular models (neglecting the side chains) of (a–c) a self-assembled monolayer of 1₂ and (d–f) 1₃ at the solution/solid interface of the respective compound in 1,2,4-trichlorobenzene and highly oriented pyrolytic graphite. Image and unit cell parameters: (a) 1₂: c = 5 × 10⁻⁶ M, 30 × 30 nm², V_S = −1.3 V, I_t = 23 pA; a = (4.3 ± 0.2) nm, b = (3.7 ± 0.2) nm, γ(a,b) = (87 ± 2)°, γ(b,d₁) = (1 ± 2)°, γ(d₁,d₂) = (90 ± 4)°; (d) 1₃: c = 3 × 10⁻⁵ M, 40 × 40 nm² (internal scanner calibration), V_S = −0.7 V, I_t = 18 pA; (e) 1₃: c = 3 × 10⁻⁵ M, 9.4 × 9.4 nm² (internal scanner calibration), V_S = −0.7 V, I_t = 20 pA; all samples thermally annealed for 20 s at 80 °C. Red lines in panels (a–c) indicate the unit cell; white and black solid (and dashed) lines in panels (a, b) indicate the HOPG main axis (and armchair) directions; gray boxes and lines in panel (c) indicate the bottom rims and interdigitation pattern of the hexadecyloxy side chains; blue arrows in panel (c) indicate the tilting directions of 6 out of 9 top rims in the marked surface region in panel (a); black and white dashed ovals in panels (e, f) highlight the phenanthracene units.

Figure 5.

(a) Quartz crystal microbalance measurements with dimer 1₂ for different analytes (10 ppm in dry air); depicted is the change in the third harmonic of the resonant frequency with time upon analyte exposure; (b) current–voltage (I–V) characteristics of graphene-based field-effect transistors (GFETs) with 1₂ before and after 5 min exposure of 10 ppm of nitrobenzene in dry air. (N ≥ 16).

Figure 6.

Illustration of a swelling-based sensing mechanism, showing the expansion of the SWCNT network (d₂ > d₁) upon the binding of nitrotoluene to the PNTs.

Figure 7.

Chemiresistive responses of 1₂-SWCNT (yellow), 1₃-SWCNT (pink), and p-SWCNT (orange) upon exposure to 10 ppm of various analytes in dry air for 5 min. (N ≥ 4).

Figure 8.

Chemiresistive response-traces and sensor response to analyte concentration relationship for 1₂-SWCNT chemiresistors upon exposure to 2-nitrotoluene in dry air in different concentrations (a, c) below 1 ppm (b, d). (N ≥ 4).

Scheme 1.. Synthesis of the Phenanthracene Nanotubes (Top: Cyclodimer 1₂; Bottom: Cyclotrimer 1₃; Schematically)

Scheme 2.. Synthetic Approach Towards H-Shaped Monomer 11 and the Desired Nanotubular Target Structures 1₂ and 1₃.^a

^aReagents and Conditions: (a) NaOH, toluene, reflux, 1 h, 96%; (b) PdCl₂(PPh₃)₂ (cat.), PPh₃, CuI (cat.), piperidine:THF (2:1), 40 °C, 20 h, 62%; (c) CHCl₃, AcOH, reflux, 24 h, 92%; (d) NaOH (dry), toluene, reflux, 30 min, 61%; (e) Pd(OAc)₂ (cat.), XPhos, CuI (cat.), piperidine:THF (2:1), 80 °C, 20 h, 67%; (f) K₂CO₃ (anhydr.), MeOH:THF (1:2), r.t., 24 h, >99%; (g) PdCl₂(PPh₃)₂ (cat.), CuI (cat.), I₂, THF:DIPA (1:1), 50 °C, injection over 24 h, 31% 12₂ and 9% 12₃; (h) TBAF (1 M in THF), THF, 35 °C, 3 h; (i) PdCl₂(PPh₃)₂ (cat.), CuI (cat.), I₂, THF:DIPA (1:1), 50 °C, injection over 48 h, 25% 1₂ and 7% 1₃, over two steps respectively.