Azaylide-based gemini amphiphiles displaying unique self-assembling behavior via an even–odd effect of alkyl linker chain length

Abstract

Herein, we report a straightforward synthesis of azaylide-based gemini amphiphiles using bis(diphenylphosphino)alkanes via the Staudinger reaction. The prepared gemini amphiphiles exhibited an even–odd effect in their self-assembly behavior depending on the length of the alkyl linkers. Furthermore, the assembled micelles had high host capability toward hydrophobic guests in water.

---

Gemini amphiphiles consist of two hydrophilic moieties and two hydrophobic moieties connected through a linker.^1–5 These amphiphiles are generally highly functional compared with their monomeric amphiphiles;^6–9 therefore, they have been utilized in various applications including drug and gene delivery,^10–14 antibacterial agents,^15–18 corrosion inhibitors,^19,20 phase transfer catalysts,²¹ and microemulsion stabilizers.^22–24 In terms of the even–odd effect,^25–28 which is the oscillation in properties depending on the even or odd number of connected repeating units (e.g., methylene group), some gemini amphiphiles comprising aliphatic moieties (e.g., n-alkane) as building blocks display a unique self-assembling behavior.^29–31 Although these examples are based on long alkyl linkers, e.g., (CH₂)_n with n = 10 and 12, the even–odd effect of short alkyl chains can be expected to dramatically change the aggregation behavior owing to their considerable impact on the molecular structure.^32,33 However, systematic studies on the self-assembly of gemini amphiphiles with short alkyl linkers are scarce.^31,34 We previously reported the facile transformation of hydrophobic triarylphosphines toward monomeric amphiphiles via the Staudinger reaction.^35,36 The obtained azaylide-based (i.e., N=P bond) amphiphiles self-assembled into ca. 2-nm-sized spherical micelles that could encapsulate various hydrophobic dyes in water. In addition to monophosphines, diphosphines such as bis(diphenylphosphino)alkanes with various alkyl chains exhibit different catalytic activities depending on the alkyl chain length and are commercially available (Fig. 1a).³⁷ In this study, we focused on azaylide-based gemini amphiphiles³⁸ with a short alkyl linkage (NdpPn; n = 1–6, Fig. 1b) based on bis(diphenylphosphino)alkanes (dpPn) as the building blocks and investigated the unique self-assembling behavior of NdpPn in water along with that of the corresponding monomeric amphiphile NPPh2Me (Fig. 1c and d) for comparison.

Fig. 1 Chemical structures of (a) diphosphines dpPn (n = 1–6) and (c) monophosphine PPh2Me. (b) Chemical structures of azaylide-based gemini amphiphiles NdpPn (n = 1–6) and (d) the corresponding monomeric amphiphile NPPh2Me.

A series of azaylide-based gemini amphiphiles were successfully synthesized by simply mixing hydrophilic azide 1 and bis(diphenylphosphino)alkanes in CH₃CN. To prevent residual free hydrophilic azide 1 and intermediates (e.g., monosubstituted product), a small excess of bis(diphenylphosphino)alkanes was added to 1 two times. First, 0.45 equiv. of bis(diphenylphosphino)propane (dpP3) was added to a CH₃CN solution of hydrophilic azide 1 (1.0 equiv.) and stirred at 60 °C until complete consumption of dpP3 (ca. 10 min). Subsequently, a CHCl₃ solution of dpP3 (0.075 equiv.) was slowly added to the resultant solution over 10 min to afford the desired gemini amphiphile NdpP3 in excellent yield (96%), as confirmed by NMR and MS analyses (Fig. 2a). After the formation of the azaylide (N=P bond), a ³¹P NMR signal attributable to NdpP3 appeared at 7.3 ppm, which shifted downfield by +25.1 ppm relative to that of free dpP3, and the ¹H NMR signals of the dpP3 moiety showed downfield shift (Δδ = +0.69 ppm, Fig. 2b and c). The mass spectrum of dicationic NdpP3 showed a prominent signal at m/z = 496.18 for [M−2Cl]²⁺ (Fig. 2d). The other gemini amphiphiles were synthesized in a similar manner in 87–98% yields (Fig. S1, ESI†). Interestingly, the synthesis of NdpP1 required a longer reaction time (3 h), higher concentration (113 mM), and higher temperature (80 °C) due to the increased steric repulsion between the diphenylphosphine moieties. The crowded aromatic rings and phosphine moieties exerted synergistic shielding effects, resulting in a remarkable downfield shift of the ¹H signal of the methylene linker and an upfield shift of the ³¹P signal compared with those of other NdpPn.

Fig. 2 (a) Synthesis of azaylide-based gemini amphiphile NdpP3via the Staudinger reaction. (b) ¹H NMR (500 MHz, DMSO-d₆, 298 K) and (c) ³¹P NMR (202 MHz, DMSO-d₆, 298 K) spectra of dpP3 (top) and NdpP3 (bottom). (d) ESI-TOF MS (CH₃CN/CH₃OH) spectrum of NdpP3. (e) X-ray crystal structure of NdpP3^I. Thermal ellipsoids are set at 50% probability. (f) Time-course of the decomposition of NdpP3 and nNdpP3 in D₂O.

A single-crystal X-ray diffraction (SXRD) analysis of NdpP3^I, in which the counter ions in NdpP3 were replaced with two iodide ions, unambiguously revealed the chemical structure of gemini amphiphile NdpP3 (Fig. 2e and Fig. S2, ESI†), which consisted of two hydrophilic parts connected to the dpP3 moiety in the same direction, i.e., in the syn-conformation. One phenyl group of dpP3 and a phenyl azaylide moiety formed a trans-stilbene-like structure.³⁵NdpP3 displayed an absorption band (λ_max = 318 nm) and a weak blue emission (λ_em = 450 nm) in CH₃CN (Fig. S4, ESI†), similar to those of a previously reported azaylide-based amphiphile with PPh₃.³⁵ The other azaylide compounds NdpP1 and NdpP5 showed excellent water solubility. Non-chlorinated nNdpP3 was extremely fragile in D₂O, exhibiting a lifetime (τ_1/2) of 14 min (Fig. 2f and Fig. S5 and S6, ESI†), whereas a triphenylphosphine-based azaylide³⁵ showed a relatively long τ_1/2 of 9.8 h. This result indicates that electron-rich yet less sterically hindered alkyl phosphines accelerate the hydrolysis of the azaylide moiety. Nevertheless, chlorinated NdpP3 exhibited high stability against hydrolysis even in water (Fig. 2f and Fig. S7, ESI†). Thus, our strategy can be applied to bis(diphenylphosphino)alkanes with reactive phosphine sites.

Next, the self-assembling behavior of the gemini amphiphiles was investigated in water. First, the amphiphiles with an odd number of methylene linkers, i.e., NdpP1, NdpP3, and NdpP5, were examined. When NdpP3 was dissolved in D₂O (2.0 mM), (NdpP3)_m micelle aggregates were quantitatively and spontaneously formed (Fig. 3a (top)). The ¹H and ³¹P NMR signals of NdpP3 in D₂O were substantially broadened owing to the dynamic behavior (Fig. S8 and S9, ESI†). A diffusion-ordered spectroscopy (DOSY) analysis revealed a single band at a small diffusion coefficient (D) of 2.28 × 10⁻¹⁰ m² s⁻¹ (Fig. S20, ESI†). According to the Stokes–Einstein equation, this D value corresponds to aggregates with a diameter of ca. 1.7 nm. A dynamic light scattering (DLS) analysis revealed the formation of aggregates with a diameter of ca. 2.1 nm in water with a narrow size distribution (Fig. 3b and Fig. S21, ESI†). Atomic force microscopy (AFM) measurements confirmed the existence of spherical aggregates (NdpP3)_m with a height of ca. 2–3 nm (Fig. 3c and Fig. S28, ESI†). These results indicate that the size of the (NdpP3)_m micelle aggregates corresponds to four molecules of NdpP3, as estimated by molecular mechanics (MM) calculations (Fig. S22, ESI†). The critical micelle concentration (CMC) of NdpP3 was measured using concentration-dependent ¹H NMR and DOSY analyses. The D values suddenly decreased when the concentration reached 1.0 mM, which was accompanied by a broadening of the ¹H NMR signals (Fig. S23 and S26, ESI†). Thus, the CMC of NdpP3 was determined to be 1.0 mM, which is considerably lower than that of the monomeric amphiphile NPPh2Me (10 mM, Fig. S25, ESI†). The other amphiphiles NdpP1 and NdpP5 also formed micelle aggregates in water with relatively small CMC values (1.0 and 0.20 mM, respectively, Fig. S10–S13, S20–S22, S24, and S26, ESI†). Owing to the increased hydrophobicity, the CMC values gradually decreased with increasing methylene linker length.

Fig. 3 (a) Schematic of differences in assembling behavior into micelles and layered structures in water. (b) DLS charts of a 2.0 mM aqueous suspension of NdpP3 and NdpP2 (298 K). AFM images of (c) micelles of NdpP3 and (d) the layered structure of NdpP2. (e) X-ray packing structure of NdpP2. (f) PXRD patterns of a white dispersion of NdpP2 and simulation based on the results of SXRD of NdpP2.

Different results were obtained for the amphiphiles with an even number of methylene linkers, i.e., NdpP2 and NdpP4. When NdpP2 (4.0 μmol) was added to H₂O (2.0 mL), a white suspension was obtained instead of a clear aqueous solution (Fig. 3a (bottom)). A DLS analysis of the suspension revealed the existence of two kinds of aggregates, one type with a diameter of ca. 2.6 nm and a narrow distribution and another type with a diameter of ca. 1.4 μm and a broadened distribution (Fig. 3b). ¹H NMR analyses showed sharp signals with a large D value of 5.25 × 10⁻¹⁰ m² s⁻¹ at 298 K in water (Fig. S14, S15 and S20, ESI†). The fact that NdpP2 has low water-solubility (ca. 0.41 mM) suggests that NdpP2 is in a monodisperse state or forming premicelles in water. Apart from these NdpP2, most of the NdpP2 molecules formed a white suspension consisting of ca. 1–2 μm aggregates in water. AFM and transmission electron microscopy (TEM) observations showed large aggregates (ca. 1–2 μm) and spherical small aggregates (ca. 3–10 nm), which are consistent with the DLS results (Fig. S28 and S29, ESI†). In particular, the AFM image revealed that the large aggregates exhibit a multilayered structure (Fig. 3d), implying the formation of microcrystals.^39,40

To gain more insight into the self-assembled structures, X-ray diffraction analyses were conducted. Slow evaporation of a CH₃CN/CHCl₃ solution of NdpP2 resulted in plate-shaped single crystals, whose SXRD analysis revealed that the two hydrophilic moieties possessed an anti-conformation, differing from that of NdpP3 (Fig. 2e). In the crystalline packing, elongated NdpP2 molecules were aligned to form a layered structure; specifically, the amphiphiles formed 2D-sheets through multiple CH–π interactions along the a- and c-axes, and these 2D-sheets stacked in the b-axis (Fig. 3e). The powder X-ray diffraction (PXRD) pattern of white dispersions of NdpP2 in water was consistent with the simulation pattern of NdpP2 based on the SXRD analysis (Fig. 3f). A similar PXRD pattern was found for NdpP4 (Fig. S27, ESI†). Thus, unlike the micelle formation of NdpP1, NdpP3, and NdpP5, both NdpP2 and NdpP4 exhibited high crystallinity with packing structures similar to that shown in Fig. 3e in water. These differences stem most likely from the dipole moments of NdpP3 (38.66 D) and NdpP2 (4.84 D), with the latter resulting in a strong packing force.

Contrary to our expectations, NdpP6 with an even number of methylene linkers was highly soluble in water, resulting in the formation of micelles with a diameter of ca. 2–3 nm instead of a layered structure (Fig. S18–S21, ESI†). Increasing the length of the alkyl linker chain should endow the gemini amphiphiles with conformational flexibility, leading to the formation of micelle aggregates in water. This observation indicates that the remarkable even–odd effect of the alkyl linker on the assembly behavior is unambiguous for the amphiphiles with short linkers up to n = 5.

Finally, the host capability of the gemini amphiphiles was investigated. NdpP3 (4.0 μmol) and perylene (Per; 1.0 equiv.) were ground in a mortar for 3 min, followed by the addition of water (2.0 mL), and the mixture was stirred vigorously for 1 h. After removal of excess guest powders via filtration, a yellow solution of (Per)_x·(NdpP3)_m was obtained (Fig. 4a). The UV-vis spectrum clearly displayed absorption bands around 401 nm ascribable to encapsulated hydrophobic Per (Fig. 4b). Encapsulated Per exhibited weak excimer emission (Fig. S30, ESI†). The host–guest complex (Per)_x·(NdpP3)_m was ca. 2.7 nm in diameter, as confirmed by DLS measurement (Fig. 4c). The ¹H NMR spectrum of a DMSO-d₆ solution of the solid revealed that the amount of encapsulated Per was 0.87 mM in 2.0 mM of NdpP3 aqueous solution (Fig. S36, ESI†). Considering these results, MM calculation provided (Per)₂·(NdpP3)₅ as an average host–guest complex model (Fig. 4a (right) and Fig. S39, ESI†). Similarly, NdpP3 successfully encapsulated other hydrophobic guests such as pyrromethene 546 (BP) and fullerene (C60) in water (Fig. 4a and Fig. S30–S39, ESI†). It should be noted that the guest uptake ability of gemini amphiphile NdpP3 toward Per was much higher than that of the monomeric amphiphile NPPh2Me (0.072 mM, Fig. 4d). Interestingly, the host capability toward Per of the other gemini amphiphiles, i.e., NdpP1, NdpP5, and NdpP6, was similar regardless of the length of the alkyl linker (Fig. 4d and Fig. S30, S33, S36, S39, ESI†). This tendency was observed in the case of BP (Fig. S37 and S38, ESI†). Considering that aromatic moieties can effectively encapsulate guest molecules in water,^9,35,36,41 the bound hydrophobic cores may play a key role in the effective molecular encapsulation.

Fig. 4 (a) Schematic of the encapsulation of hydrophobic guests (pyrromethene 546 (BP), perylene (Per), and fullerene (C60)) by NdpP3 in water and the optimized structure of (Per)₂·(NdpP3)₅. (b) UV-vis spectra and (c) DLS charts of (Per)_x·(NdpP3)_m, (BP)_x·(NdpP3)_m, and (C60)_x·(NdpP3)_m (H₂O, 298 K, 2.0 mM based on NdpP3). (d) Comparisons of the uptake quantity of Per by NPPh2Me and NdpPn (n = 1, 3, 5, 6). The numbers above the bars indicate relative uptake ratio to NPPh2Me (note: NPPh2Me was used 4.0 mM to unify the concentration of diphenylphosphino moiety).

In conclusion, we have successfully synthesized a series of azaylide-based gemini amphiphiles from bis(diphenylphosphino)alkanes via the Staudinger reaction. The gemini amphiphiles exhibited an even–odd effect of the alkyl chain length on their assembly behavior. Our findings provide useful guidance for designing new types of multilinked amphiphiles (e.g. gemini and bola⁴¹ amphiphiles) in the near future.

This study was supported by a JSPS Grant-in-Aid for Scientific Research (20H02721), Grants-in-Aid for Early-Career Scientists (22K14662), and the Cooperative Research Program of “NJRC Mater. & Dev. (MEXT).” The authors thank Tokyo Tech. Open Facility Center. The X-ray analysis was performed under the approval of the Photon Factory Program Advisory Committee (2021G046, PF-BL5a). We appreciate Prof. Masaki Kawano (Tokyo Tech.) for X-ray analysis and Prof. Kei Goto (Tokyo Tech.) for the FTIR measurements. We also appreciate valuable discussions with Dr Eiji Tsurumaki (Tokyo Tech.).

Data availability

The data underlying this study are available in the published article and its ESI.†

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. F. M. Menger and C. A. Littau, J. Am. Chem. Soc., 1991, 113, 1451–1452.
2. F. M. Menger and J. S. Keiper, Angew. Chem., Int. Ed., 2000, 39, 1906–1920.
3. R. Zana, Adv. Colloid Interface Sci., 2002, 97, 205–253.
4. D. Shukla and V. K. Tyagi, J. Oleo Sci., 2006, 55, 381–390.
5. R. Sharma, A. Kamal, M. Abdinejad, R. K. Mahajan and H.-B. Kraatz, Adv. Colloid Interface Sci., 2017, 248, 35–68.
6. E. Alami, G. Beinert, P. Marie and R. Zana, Langmuir, 1993, 9, 1465–1467.
7. Z. Wang, Y. Li, X.-H. Dong, X. Yu, K. Guo, H. Su, K. Yue, C. Wesdemiotis, S. Z. D. Cheng and W.-B. Zhang, Chem. Sci., 2013, 4, 1345–1352.
8. K. Hamaguchi, R. Ichikawa, S. Kajiyama, S. Torii, Y. Hayashi, J. Kumaki, H. Katayama and T. Kato, ACS Appl. Mater. Interfaces, 2021, 13, 20598–20605.
9. T. Nishioka, K. Kuroda, M. Akita and M. Yoshizawa, Angew. Chem., Int. Ed., 2019, 58, 6579–6583.
10. C. Bombelli, L. Giansanti, P. Luciani and G. Mancini, Curr. Med. Chem., 2009, 16, 171–183.
11. S. Mahajan and R. K. Mahajan, Adv. Colloid Interface Sci., 2013, 199, 1–14.
12. M. Rodrigues, A. C. Calpena, D. B. Amabilino, M. L. Garduño-Ramírez and L. Pérez-García, J. Mater. Chem. B, 2014, 2, 5419–5429.
13. A. K. Singh, K. Achazi, S. Ehrmann, C. Böttcher, R. Haag and S. K. Sharma, Polym. Chem., 2020, 11, 6772–6782.
14. A. Mittal, F. Zabihi, F. Rancan, K. Achazi, C. Nie, A. Voggt, R. Haag and S. K. Sharma, RSC Adv., 2022, 12, 23566–23577.
15. M. C. Grenier, R. W. Davis, K. L. Wilson-Henjum, J. E. LaDow, J. W. Black, K. L. Caran, K. Seifert and K. P. C. Minbiole, Bioorg. Med. Chem. Lett., 2012, 22, 4055–4058.
16. A. Pinazo, M. A. Manresa, A. M. Marques, M. Bustelo, M. J. Espuny and L. Pérez, Adv. Colloid Interface Sci., 2016, 228, 17–39.
17. K. Taleb, M. Mohamed-Benkada, N. Benhamed, S. Saidi-Besbes, Y. Grohens and A. Derdour, J. Mol. Liq., 2017, 241, 81–90.
18. R. Qi, N. Zhang, P. Zhang, H. Zhao, J. Liu, J. Cui, J. Xiang, Y. Han, S. Wang and Y. Wang, ACS Appl. Mater. Interfaces, 2020, 12, 17220–17229.
19. B. Brycki and A. Szulc, J. Mol. Liq., 2021, 344, 117686.
20. A. R. Ahmady, P. Hosseinzadeh, A. Solouk, S. Akbari, A. M. Szulc and B. E. Brycki, Adv. Colloid Interface Sci., 2022, 299, 102581.
21. C. Borde, V. Nardello, L. Wattebled, A. Laschewsky and J.-M. Aubry, J. Phys. Org. Chem., 2008, 21, 652–658.
22. M. Rodrigues, A. C. Calpena, D. B. Amabilino, D. Ramos-López, J. de Lapuente and L. Pérez-García, RSC Adv., 2014, 4, 9279–9287.
23. M. E. Alea-Reyes, A. González, A. C. Calpena, D. RamosLópez, J. de Lapuente and L. Pérez-García, J. Colloid Interface Sci., 2017, 502, 172–183.
24. S. Wang, X. Xin, H. Zhang, J. Shen, Y. Zheng, Z. Song and Y. Yang, RSC Adv., 2016, 6, 28156–28164.
25. F. Tao and S. L. Bernasek, Chem. Rev., 2007, 107, 1408–1453.
26. E. Yashima, N. Ousaka, D. Taura, K. Shimomura, T. Ikai and K. Maeda, Chem. Rev., 2016, 116, 13752–13990.
27. T. Shimizu, Polym. J., 2003, 35, 1–22.
28. A. D. Bond, New J. Chem., 2004, 28, 104–114.
29. Y. Li, P. Li, J. Wang, Y. Wang, H. Yan and R. K. Thomas, Langmuir, 2005, 21, 6703–6706.
30. D. V. Perroni, C. M. Baez-Cotto, G. P. Sorenson and M. K. Mahanthappa, J. Phys. Chem. Lett., 2015, 6, 993–998.
31. S. Mantha, J. G. McDaniel, D. V. Perroni, M. K. Mahanthappa and A. Yethiraj, J. Phys. Chem. B, 2017, 121, 565–576.
32. K. E. Jelfs, E. G. B. Eden, J. L. Culshaw, S. Shakespeare, E. O. Pyzer-Knapp, H. P. G. Thompson, J. Bacsa, G. M. Day, D. J. Adams and A. I. Cooper, J. Am. Chem. Soc., 2013, 135, 9307–9310.
33. Z. Xiong, J. Zhang, J. Z. Sun, H. Zhang and B. Z. Tang, J. Am. Chem. Soc., 2023, 145, 21104–22113.
34. R. Zana, J. Colloid Interface Sci., 2002, 248, 203–220.
35. M. Yamashina, H. Suzuki, N. Kishida, M. Yoshizawa and S. Toyota, Angew. Chem., Int. Ed., 2021, 60, 17915–17919.
36. H. Suzuki, Y. Akiyama, M. Yamashina, Y. Tanaka and S. Toyota, Chem. – Eur. J., 2023, 29, e202303017.
37. A. L. Clevenger, R. M. Stolley, J. Aderibigbe and J. Louie, Chem. Rev., 2020, 120, 6124–6196.
38. K. J. Sommers, M. E. Michaud, C. E. Hogue, A. M. Scharnow, L. E. Amoo, A. A. Petersen, R. G. Carden, K. P. C. Minbiole and W. M. Wuest, ACS Infect. Dis., 2022, 8, 387–397.
39. B. A. Pindzola, J. Jin and D. L. Gin, J. Am. Chem. Soc., 2003, 125, 2940–2949.
40. H. Oshiro, T. Kobayashi and T. Ichikawa, Mol. Syst. Des. Eng., 2022, 7, 1459–1466.
41. J.-H. Fuhrhop and T. Wang, Chem. Rev., 2004, 104, 2901–2937.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 2342817 and 2342818. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4sm00789a

---