Revisiting the Formation and Tunable Dissociation of a [2]Pseudorotaxane Formed by Slippage Approach

Abstract

A new [2]pseudorotaxane DB24C8⊃1-H·PF₆ with dibenzo[24]crown-8 (DB24C8) crown ether-dibenzylammonium (1-H·PF₆) binding which was formed by slippage approach at different solvents and temperature, had been isolated and characterized by NMR spectroscopy and mass spectrometry. The [2]pseudorotaxane DB24C8⊃1-H·PF₆ was stable at room temperature. The dissociation rate of [2]pseudorotaxane DB24C8⊃1-H·PF₆ could be tuned by using different stimuli such as triethylamine (TEA)/diisopropylethylamine (DIPEA) and dimethyl sulfoxide (DMSO). In particular, the dissociation of [2]pseudorotaxane DB24C8⊃1-H·PF₆ by an excess of TEA/DIPEA base mixture possessed a long and sustained, complete dissociation over 60 days. Other stimuli by DMSO possessed a relatively fast dissociation over 24 h.

1. Introduction

Rotaxanes contain a linear dumbbell-shaped component bearing bulky end-groups or stoppers around which one or more macrocycles are trapped. On the other hand, pseudorotaxanes are temporally encircled around an unstoppered thread through noncovalent interactions from which they are readily susceptible to dissociation without breaking a covalent bond [1,2,3,4]. Rotaxanes and pseudorotaxanes have been studied extensively for the ability of the interlocked ring to be switched on demand by external stimuli such as pH [5,6], electrochemical reagents [7], heat [8,9], moisture [10,11], salt [12], light [13], etc. Coupled with their ability to be customized and optimized for nanoscale functions, these interlocked molecules are excellent candidates as movable elements in the construction of nanovalves [14] based on a porous, solid-phase support, for controlled substrate release. Many examples have been demonstrated about their relatively fast substrate releases within 24 h by tuning the pH values in the solution [12]. Currently, no pseudorotaxane or rotaxane building blocks are available and suitable for the construction of vehicles for sustained substrate release over a week.

The construction of rotaxane-like assemblies has recently relied upon thermodynamic, templated reactions with enhanced efficiencies [15,16,17,18,19,20,21,22,23,24]. The term “slippage” has been coined [25,26,27,28] for pseudorotaxane synthesis that employs thermodynamic threading of macrocycle to a “dumbbell”. This strategy utilizes (1) the size complementarily between the macrocycle and the “dumbbell’s” stoppers; and (2) the stabilizing noncovalent bonding interactions between the macrocycle and the “dumbbell’s” rod. In this strategy, the macrocycle and the dumbbell have been separately synthesized, prior to heating them together in solution so that the free energy of activation for the thermodynamic threading (slippage) of macrocycle to dumbbell can be overcome. The presence of a template on the dumbbell’s rod renders the pseudorotaxane structure more stable so that the free energy of activation for its dissociation becomes insurmountable when the solution has been cooled to ambient temperature. There are several disadvantages of the slippage approach in terms of reaction time and stability. Generally, it requires a long reaction time with a large slippage rate (k_on) for pseudorotaxane formation from usually a few days up to 90 days [27]. However, the dissociation of pseudorotaxane will sometimes have an uncontrollable, small slippage rate (k_off). Therefore, there is a need to study a better reaction condition for pseudorotaxane formation by slippage and to tune the pseudorotaxane dissociation rate (k_off << k_on) from days to hours.

2. Results and Discussion

Herein, we employ a slippage approach in one pot (Figure 1) to yield a thermodynamically stable [2]pseudorotaxane [29,30] by mixing an ammonium thread (dumbbell) 1-H·PF₆ with dibenzo[24]crown-8 (DB24C8). Benzo-crown ether DB24C8 is capable to recognize with secondary ammonium ions by virtue of [N⁺–H···O] and [N⁺–C–H···O] hydrogen bonds, electrostatic interactions, and augmented with some aromatic π-π interactions [31,32,33,34,35,36]. The 3,5-dimethoxyaryl moiety at one end of the dumbbell 1-H·PF₆, is a relatively bulky stopper that can effectively block the passageway of DB24C8 at high reaction temperature. On the other hand, the cyclohexyl ring of the dumbbell-like thread, in contrast, allow the crown ether to thread through to the dumbbell’s rod at elevated reaction temperature with the molecular flipping of between their chair-boat-chair forms [31,32,33,34,35,36,37].

Figure 1. Molecular structures of the crown ether DB24C8 and the ammonium thread “dumbbell” 1-H·PF₆. The 3,5-dimethoxyaryl group of the thread acts as an effective stopper to prevent the extrusion of macrocycle while the cyclohexyl group of the thread allows the slipping of macrocycle upon heating in selected solvents. The thermodynamically stable [2]pseudorotaxane DB24C8⊃1-H·PF₆ was isolated and subjected to dissociation studies with excess amine base (Et₃N and DIPEA) mixture and dimethyl sulfoxide (DMSO).

Figure 1. Molecular structures of the crown ether DB24C8 and the ammonium thread “dumbbell” 1-H·PF₆. The 3,5-dimethoxyaryl group of the thread acts as an effective stopper to prevent the extrusion of macrocycle while the cyclohexyl group of the thread allows the slipping of macrocycle upon heating in selected solvents. The thermodynamically stable [2]pseudorotaxane DB24C8⊃1-H·PF₆ was isolated and subjected to dissociation studies with excess amine base (Et₃N and DIPEA) mixture and dimethyl sulfoxide (DMSO).

The reaction time of the pseudorotaxane synthesis was fixed in 4 days with reasonable yields (~50%). The percentage yields (Table 1) of individual [2]pseudorotaxane after a reaction time of 4 days, have been determined by ¹H NMR spectroscopy according to their “bound” and “free” signal intensities [36,37,38]. From the results, the reaction yields and rates in synthesizing the [2]pseudorotaxanes are sensitive to different solvents and temperatures. In particular, the percentage yield of the [2]pseudorotaxane DB24C8⊃1-H·PF₆ was found to be the highest (46%) after a reaction temperature at 70 °C with MeCN compared to using PhMe (70 °C) and CH₂Cl₂ (40 °C). This is partially because, in polar solvents, the [2]pseudorotaxane might undergo intra and/or intermolecular aggregation of the hydrophobic alkyl chains and this will overcome the extrusion effect. The [2]pseudorotaxane DB24C8⊃1-H·PF₆ could be isolated by flash chromatography on silica gel (CH₂Cl₂/THF = 7/1) and characterized by NMR spectroscopy and mass spectrometry.

Table 1. Percentage yields of the [2]pseudorotaxanes after slippage reaction for 4 days. Analyzed by ¹H NMR spectroscopy (400 MHz, 298 K).

Solvent (Reaction Temperature in Sealed Tube)	DB24C8⊃1-H·PF6
MeCN (70 °C)	46%
PhMe (70 °C)	30%
CH2Cl2 (40 °C)	10%

Table 1. Percentage yields of the [2]pseudorotaxanes after slippage reaction for 4 days. Analyzed by ¹H NMR spectroscopy (400 MHz, 298 K).

Solvent (Reaction Temperature in Sealed Tube)	DB24C8⊃1-H·PF6
MeCN (70 °C)	46%
PhMe (70 °C)	30%
CH2Cl2 (40 °C)	10%

¹H NMR spectroscopy was employed to evaluate the structural features of [2]pseudorotaxanes. By way of an example, the isolated and pure [2]pseudorotaxane DB24C8⊃1-H·PF₆ reveals (Figure 2) proton chemical shifts of the –CH₂NH₂⁺CH₂– moieties at δ = 3.14 and 3.32 ppm (bound), comparing to the –CH₂NH₂⁺CH₂– moieties of its thread 1-H·PF₆ at δ = 2.81 and 3.06 ppm (free).

Figure 2. Partial ¹H NMR spectra (400 MHz, CD₃CN, 298 K) of crown ether DB24C8, isolated pure [2]pseudorotaxane DB24C8⊃1-H·PF₆, and ammonium thread 1-H·PF₆.

Figure 2. Partial ¹H NMR spectra (400 MHz, CD₃CN, 298 K) of crown ether DB24C8, isolated pure [2]pseudorotaxane DB24C8⊃1-H·PF₆, and ammonium thread 1-H·PF₆.

High-resolution electrospray ionization mass spectrometry (ESI-MS) has been employed (Figure 3) to further characterize the [2]pseudorotaxane. The molecular ion peak at m/z 740 which is the most abundant peak in the spectrum, is corresponded to the [M–PF₆]⁺ ion of the [2]pseudorotaxane.

Figure 3. Electrospray ionization mass spectrometry (ESI-MS) of the [2]pseudorotaxane DB24C8⊃1-H·PF₆, showing the molecular ion signal [M–PF₆]⁺ at m/z 740.

Figure 3. Electrospray ionization mass spectrometry (ESI-MS) of the [2]pseudorotaxane DB24C8⊃1-H·PF₆, showing the molecular ion signal [M–PF₆]⁺ at m/z 740.

Furthermore, the dissociation and stabilities of [2]pseudorotaxane towards organic amine bases and a hydrogen bond disrupting solvent were evaluated. In particular, pure [2]pseudorotaxane DB24C8⊃1-H·PF₆ was dissolved in CD₃CN and was treated with an excess of triethylamine (TEA)/diisopropylethylamine (DIPEA) mixture [5,6,12]. The ammonium ion of the thread 1-H·PF₆ could be successfully deprotonated by the bases wherein the DB24C8 loss its binding affinity towards the deprotonated, amine thread 1. Since the template effect is lost, the [2]pseudorotaxane is no longer stable whereas extrusion of macrocycle occurs with the molecular flipping of the cyclohexyl ring. This extrusion behavior was monitored (Figure 4) by observing a significant decrease of characteristic signal at δ = 6.16 ppm (bound ArH of 1-H·PF₆) as well as an increase of signal at δ = 6.38 ppm (free ArH of 1) from their ¹H NMR spectra over time. Interestingly, the [2]pseudorotaxane requires almost 60 days to dissociate completely into two separate components at ambient temperature. The half-life of dissociation was determined to be approximately 2 days (Figure 6).

Figure 4. Partial ¹H NMR spectra (400 MHz, CD₃CN, 298 K) showing the extrusion of DB24C8 macrocycle from the isolated [2]pseudorotaxane DB24C8⊃1-H·PF₆ with excess triethylamine (TEA)/diisopropylethylamine (DIPEA) mixture in CD₃CN over time.

Figure 4. Partial ¹H NMR spectra (400 MHz, CD₃CN, 298 K) showing the extrusion of DB24C8 macrocycle from the isolated [2]pseudorotaxane DB24C8⊃1-H·PF₆ with excess triethylamine (TEA)/diisopropylethylamine (DIPEA) mixture in CD₃CN over time.

On the other hand, by dissolving the pure [2]pseudorotaxane DB24C8⊃1-H·PF₆ in a hydrogen bond disfavored solvent—dimethyl sulfoxide (DMSO) [1,27], extrusion of DB24C8 from the thread 1-H·PF₆ occurred. Since the template effect is lost, the [2]pseudorotaxane is no longer stable whereas extrusion of macrocycle occurs with the molecular flipping of the cyclohexyl ring. This extrusion behavior was monitored (Figure 5) by observing the decreases of characteristic signals at δ = 3.81, and 4.14 ppm (bound –OCH₂CH₂O– of DB24C8) as well as the increases of signals at δ = 3.74, and 4.07 ppm (free –OCH₂CH₂O– of DB24C8) from their ¹H NMR spectra (in DMSO-d₆) over time. The [2]pseudorotaxane required only 24 h for a complete dissociation. The half-life of dissociation was determined to be approximately 4 h (Figure 6).

Figure 5. Stacked ¹H NMR spectra (400 MHz, CD₃SOCD₃, 298 K) showing the extrusion of DB24C8 macrocycle from the isolated [2]pseudorotaxane DB24C8⊃1-H·PF₆ in DMSO-d₆ over time.

Figure 5. Stacked ¹H NMR spectra (400 MHz, CD₃SOCD₃, 298 K) showing the extrusion of DB24C8 macrocycle from the isolated [2]pseudorotaxane DB24C8⊃1-H·PF₆ in DMSO-d₆ over time.

Figure 6. Dissociation data of the [2]pseudorotaxane DB24C8⊃1-H·PF₆ in the presence of base and DMSO, characterized by ¹H NMR spectroscopy. A plot of the concentration (%) of the [2]pseudorotaxane DB24C8⊃1-H·PF₆ versus time (day).

Figure 6. Dissociation data of the [2]pseudorotaxane DB24C8⊃1-H·PF₆ in the presence of base and DMSO, characterized by ¹H NMR spectroscopy. A plot of the concentration (%) of the [2]pseudorotaxane DB24C8⊃1-H·PF₆ versus time (day).

3. Experimental Section

General Information. ¹H NMR (400 MHz) and ¹³C NMR (100 MHz) spectra were recorded at room temperature in CDCl₃ unless otherwise stated. Each solvent residual signal was used as the internal standard. Chemical shifts were reported as parts per million (ppm) in δ scale and coupling constants (J) were reported in hertz. Mass spectra were obtained on a double focusing sector mass spectrometer with electrospray ionization (ESI) technique. The reported molecular mass (m/z) values, unless otherwise specified, were mono-isotopic mass. All reactions were carried out under N₂. All reactions were monitored by thin layer chromatography (TLC) performed on pre-coated silica gel 60 F₂₅₄ plates, and compounds were visualized with a spray of 5% (w/v) dodecamolybdophosphoric acid in ethanol and subsequent heating. Flash chromatography was carried out on columns of silica gel (230–400 mesh). Tetrahydrofuran (THF) was freshly distilled prior to use from sodium/benzophenone ketyl under N₂. CH₂Cl₂ was freshly distilled from CaH₂. Cyclohexanemethylamine (3) and dibenzo[24]crown-8 (DB24C8) were commercially available from Sigma-Aldrich (St. Louis, MO, USA) while the starting compound (2) was synthesized according to the literature procedures [39]. The synthetic scheme of new compounds is shown in Scheme 1.

Amide 4. 1-[3-(Dimethylamino)propyl]-3-ethylcarbodiimide methiodide (EDCI, 2.53 g, 8.52 mmol) was added to a stirred solution which contained 3-(3,5-dimethoxyphenyl)-propionic acid (2, 1.43 g, 6.81 mmol), cyclohexanemethylamine (3, 1.00 mL, 7.69 mmol) and 1-hydroxybenzotriazole (HOBt) (1.14 g, 8.45 mmol) in CH₂Cl₂ (25 mL). The reaction mixture was stirred for 18 h at 25 °C and concentrated under reduced pressure. The residue was purified by flash chromatography on silica gel (hexane/EtOAc = 4/1) to afford the amide (4) (1.50 g, 72%) as a white solid. M.p.: 72.6–74.3 °C. R_f: 0.58 (hexane/EtOAc = 4/1). ¹H NMR: δ 0.64–0.81 (2 H, m, CH₂), 0.92–1.14 (3 H, m), 1.20–1.34 (1 H, m), 1.43–1.62 (5 H, m), 2.38 (2 H, t, J = 7.6, CH₂CH₂), 2.77 (2 H, t, J = 7.6, CH₂C=O), 2.92 (2 H, t, J = 6.4, CHCH₂), 3.60 (6 H, s, CH₃), 6.16 (1 H, s, ArH), 6.23 (2 H, s, ArH), 6.61 (1 H, t, J = 5.2, NH). ¹³C NMR: δ 25.5, 26.1, 30.5, 31.8, 37.57, 37.63, 45.4, 54.7, 97.6, 106.0, 143.1, 160.5, 172.1. ESI-MS: m/z 328 ([M + Na]⁺, 100%). HRESI-MS: calcd m/z for C₁₈H₂₇NO₃Na: 328.1883, found: 328.1883 ([M + Na]⁺, 100%).

Scheme 1. Synthesis of the ammonium thread 1-H·PF₆.

Scheme 1. Synthesis of the ammonium thread 1-H·PF₆.

Amine 1. Lithium aluminum hydride (LAH) (0.77 g, 20.3 mmol) was added slowly to a stirred solution of amide (4) (1.43 g, 4.68 mmol) in THF (80 mL) at 0 °C. The solution was then heated to reflux for 24 h. The solution was cooled down to room temperature and then poured into an ice-water mixture. The resultant mixture was then extracted with CH₂Cl₂ (50 mL × 3) and the combined extracts were washed with brine, dried (MgSO₄), filtered and evaporated in vacuo to give the crude product which was purified by flash chromatography on silica gel (hexane/EtOAc/Et₃N = 240/120/1) to afford the amine (1) (1.25 g, 92%) as a colorless liquid. R_f: 0.35 (hexane/EtOAc/Et₃N = 240/120/1). ¹H NMR: δ 0.73–0.88 (2 H, m, CH₂), 0.99–1.23 (3 H, m), 1.28–1.42 (1 H, m), 1.50–1.83 (8 H, m), 2.34 (2 H, d, J = 6.8, CHCH₂NH), 2.45–2.58 (4 H, m, ArCH₂CH₂ and ArCH₂CH₂), 3.65 (6 H, s, CH₃), 6.20 (1 H, d, J = 2.0, ArH), 6.26 (2 H, d, J = 2.0, ArH). ¹³C NMR: δ 25.8, 26.4, 31.1, 31.2, 33.7, 37.6, 49.3, 54.8, 56.4, 97.4, 106.1, 144.2, 160.5. ESI-MS: m/z 292 ([M + H]⁺, 100%). HRESI-MS: calcd m/z for C₁₈H₃₀NO₂: 292.2271, found: 292.2276 ([M + H]⁺, 100%).

Ammonium salt 1-H·PF₆. To a stirred solution of amine (1) (1.25 g, 4.29 mmol) in a solvent mixture of CH₃CN/CH₂Cl₂ (3:1) (20 mL), excess hexafluorophosphoric acid (HPF₆) (1.20 mL, 8.73 mmol) was added dropwise at 0 °C. The reaction mixture was stirred for 3 h and then water (12 mL) was added to the mixture. The mixture was concentrated under reduced pressure and then extracted with CH₂Cl₂ (50 mL × 3). The combined extracts were washed with brine, dried (MgSO₄), filtered and evaporated in vacuo to afford the ammonium salt 1-H·PF₆ (1.50 g, 80%) as a white solid. M.p.: 173.6–174.8 °C. ¹H NMR (CD₃CN): δ 0.87–1.07 (2 H, m, CH₂), 1.08–1.34 (3 H, m), 1.57–1.82 (6 H, m), 1.93–2.12 (2 H, m, CH₂), 2.61 (2 H, t, J = 7.6, ArCH₂), 2.75–2.90 (2 H, m, CHCH₂NH₂), 2.93–3.10 (2 H, m, CH₂CH₂NH₂), 3.76 (6 H, s, CH₃), 6.34 (1 H, s, ArH), 6.41 (2 H, s, ArH), 6.61 (2 H, br s, NH₂). ¹³C NMR (CD₃CN): δ 26.5, 26.8, 28.1, 30.9, 33.4, 35.8, 49.4, 55.2, 56.0, 99.1, 107.5, 144.1, 162.2. ESI-MS: m/z 292 ([M–PF₆]⁺, 100%). HRESI-MS: calcd m/z for C₁₈H₃₀NO₂: 292.2271, found: 292.2265 ([M–PF₆]⁺, 100%).

Typical Synthesis of [2]Pseudorotaxane. A solution of 1-H·PF₆ (17.5 mg, 0.04 mmol) and DB24C8 (35.9 mg, 0.08 mmol) in a typical solvent (1.5 mL) was heated at 40 or 70 °C in a sealed tube for 4 days and then concentrated in vacuo. For an alternative method, a solution of 1-H·PF₆ (120 mg, 0.27 mmol) and DB24C8 (40 mg, 0.089 mmol) in MeCN (3 mL) was heated at 70 °C in a sealed tube for 4 days and then concentrated in vacuo. The resulting residue was purified by flash chromatography on silica gel (CH₂Cl₂/THF = 7/1) to afford the [2]pseudorotaxane DB24C8⊃1-H·PF₆ (34–36 mg, 46%) as a white solid. M.p. was not determined due to thermal instability. R_f: 0.65 (CH₂Cl₂/THF = 7/1). ¹H NMR (CD₃CN): δ 0.56–0.72 (2 H, m, CH₂), 0.78–0.93 (3 H, m), 1.40–1.55 (6 H, m, CH₂), 1.79–1.91 (2 H, m, CH₂), 2.41 (2 H, t, J = 7.2, ArCH₂), 3.08–3.18 (2 H, m, CHCH₂NH₂), 3.27–3.38 (2 H, m, CH₂CH₂NH₂), 3.58–3.70 (8 H, m, CH₂CH₂OCH₂), 3.72 (6 H, s, CH₃), 3.78–3.89 (8 H, m, ArOCH₂CH₂), 4.08–4.23 (8 H, m, ArOCH₂CH₂), 6.16 (2 H, d, J = 2.0, ArH), 6.31 (2 H, d, J = 2.0, ArH), 6.55–6.84 (2 H, br s, NH₂), 6.90–7.02 (8 H, m, catechol ArH). ¹³C NMR (CD₃CN): δ 26.1, 26.5, 28.7, 30.9, 33.4, 36.3, 49.2, 55.4, 55.9, 69.0, 71.3, 71.9, 98.7, 107.4, 113.6, 122.4, 144.0, 148.5, 162.0. ESI-MS: m/z 740 ([M–PF₆]⁺, 100%). HRESI-MS: calcd m/z for C₄₂H₆₂NO₁₀: 740.4368, found: 740.4359 ([M–PF₆]⁺, 100%).

Typical [2]Pseudorotaxane Dissociation. For dissociation with bases, [2]pseudorotaxane DB24C8⊃1-H·PF₆ (5 mg) was dissolved in CD₃CN (0.7 mL) in a NMR tube followed by the addition of 3 drops of Et₃N (TEA) and 3 drops of ⁱPr₂EtN (DIPEA) to the solution. ¹H NMR spectra were recorded over time at 298 K. For dissociation with solvent, [2]pseudorotaxane DB24C8⊃1-H·PF₆ (5 mg) was dissolved in CD₃SOCD₃ (0.7 mL) in a NMR tube. ¹H NMR spectra were recorded over time at 298 K.

4. Conclusions

In summary, a [2]pseudorotaxane DB24C8⊃1-H·PF₆ which was synthesized and isolated from a slippage approach, was found to be stable at room temperature. The [2]pseudorotaxane DB24C8⊃1-H·PF₆ was characterized by NMR spectroscopy and mass spectrometry. The dissociation rate of [2]pseudorotaxane DB24C8⊃1-H·PF₆ could be tuned by amine bases and DMSO solvent at room temperature from which the supramolecular interactions between the crown ether DB24C8 and the ammonium can be tuned. When the crown ether DB24C8 was lost its noncovalent interactions with the ammonium/amine, it could be detached from the dumbbell 1-H·PF₆ via the molecular flipping of between the chair-boat-chair forms of the cyclohexyl ring. In particular, the dissociation of [2]pseudorotaxane DB24C8⊃1-H·PF₆ with amine bases possessed a long and sustained, complete dissociation over 60 days. The use of DMSO possessed relatively fast and complete dissociation (24 h). The selectively tunable dissociation rates by potentially varying the ratios of stimuli for macrocycle extrusion from the novel [2]pseudorotaxane provide avenues for constructing novel functional nanovalves that sustained release substrates for a much longer period of time.