Synthesis of cyclic amphiphilic homopolymers and their potential application as polymeric micelles

Abstract

Although amphiphilic copolymers have been widely studied due to their ability to phase segregate in bulk and form micelle-like nanostructures in solution, previous research has focused primarily on block copolymers. Amphiphilic homopolymers, in which each monomer along the backbone contains an amphiphilic unit, have seen only limited exploration, while non-linear amphiphilic homopolymers remain largely unexplored. Building from methods established in our laboratories for the synthesis of cyclic polymers, bifurcated amphiphiles were attached via a highly efficient “click” coupling to access analogous sets of linear and cyclic amphiphilic homopolymers (the first reported example of cyclic amphiphilic homopolymers). These amphiphilic homopolymers showed solubility in a wide range of solvents with varying polarities and also have demonstrated the ability to encapsulate guests in incompatible solvents. For the mass range examined, the cyclic polymers showed only a negligible difference in guest encapsulation when compared to linear analogs.

---

Introduction

Polymer-based micelles have been the subject of increased research due to their ability to solvate molecules in incompatible environments. When compared to small molecule amphiphiles, amphiphilic polymers can form micelles with increased stability and more control over their size. They are of particular interest in the field of drug delivery to assist the administration of hydrophobic drug molecules which alone are incompatible with the largely hydrophilic nature of the body. Hydrophobic guest molecules can easily be encapsulated within the non-polar core of a polymer micelle while the hydrophilic corona imparts water-solubility upon the aggregate.

Both small molecule amphiphiles and amphiphilic block copolymers have been employed extensively for guest encapsulation.^1–3 In polar solution, amphiphilic polymers form globular micelles as the result of the segregation of the incompatible blocks. This phenomenon is thermodynamically driven and is very dependent upon the nature of the contrasting blocks as well as their interactions with the external solvent environment. For example, the self-assembly of normal polymer micelles in water is a result of both the polar block's compatibility with the external aqueous environment and the non-polar block's strong preference for interactions with itself. Despite the obvious utility of block-copolymer self-assembly into higher ordered micelles, such aggregates are in constant equilibrium with their unimers when in solution and therefore their stability is concentration dependent. This dependence upon concentration is manifested by an inherent critical micelle concentration (cmc) below which the micelle assembly disaggregates resulting in the release of encapsulated guests. Additionally, since micelles are formed through a thermodynamic equilibrium, the size and shape of the resultant micelles are inherent to the sizes and block ratios of the pre-formed block copolymers.

In order to enhance the robustness of polymer micelles to reduced concentrations, as well as changes in pH, temperature, ionic strength, and solvent polarity, their stabilization has been achieved through the covalent cross-linking of pre-formed micelles in solution.⁴ This can be achieved by cross-linking the core,^5–8 or in the case of “Shell Cross-linked Knedels” (SCKs) cross-linking the corona to yield a relatively hollow cavity for guest encapsulation.^9–12 As long as a sufficient degree of cross-linking has been achieved, the stabilized micellar structure can no longer dissociate back to the monomeric block copolymers, thus eliminating its critical micelle concentration. The ability to stabilize the micelles in a dynamic environment has yielded a variety of new application. While tailoring the cross-linking density provides modular tuning of the rigidity, penetrability and “void volume” of the stabilized micelle-like nanoparticles, control of their size and shape is still governed by the self-assembly process.

Alternatively, micellar structures can be built up covalently, piece-by-piece by synthesizing dendrimer or star block copolymer “unimolecular micelles” which maintain their micellar function via a covalently stabilized structure, and therefore are not susceptible to disaggregation. In addition, their three-dimensional, globular architecture enables the functionality and polarity of both the core and the corona to be selectively modified.^13–16 One complication, relative to their loading capacities, is that the core is largely filled with polymeric scaffold. While these macromolecules require demanding organic synthesis and purification relative to the self-assembly of amphiphilic block copolymers, they exhibit wider control over specific size, shape, and end-group chemistries than other routes.

Facially amphiphilic polymers, in which the polar/non-polar interface runs along the backbone, constitute another, less explored class of amphiphilic materials. Unlike block copolymers, where the interface between polar and non-polar domains runs perpendicular to the backbone, facially amphiphilic polymers exhibit either an amphiphilic monomeric unit, or alternating hydrophilic and hydrophobic monomers. The conformational arrangement of facially amphiphilic polymers in solution has been investigated with rigid linear backbones, including poly(phenyleneethynylene),^17–20 poly(thiophenes),²¹ and poly(phenylene)²² for use in electronic, antimicrobial, and materials applications. Additionally, amphiphilic homopolymers with more flexible polymer backbones based on poly(norbornene),^23,24 poly(oxanorbornene),²⁵ poly(styrene),^26,27 and poly(acrylamide)²⁸ have also been studied, but predominantly with a simple linear topology. Cyclic oligomers of methacrylic acid (DP = 6, 7, and 14) have also been used to construct unimolecular micelles in which the invertible structure is both pH and solvent dependant, however, this study was limited to small oligomeric species of specified sizes. Since the oligomers are formed from a templation of monomeric units around β-cyclodextrin, control of the ring-size is limited.^29,30

The cyclic amphiphilic polymer reported herein represents a new class of amphiphilic polymers in which the amphiphilic arms emanate from a cyclic polymer “core.” Owing to the flexibility of the monomer, the conformations of these polymer micelles in solution should be extremely dynamic and dependant on solvent (Fig. 1). For example, in a non-polar solvent, the hydrophilic chains should orient themselves inwards in order to minimize interaction with the outside environment while the lipophilic chains orient outwards to the solvent, providing a “reverse micelle-like” configuration. If the polarity of the solvent increases substantially, the backbone of the polymer should twist in order to invert the polar and non-polar side chains and provide a “normal micelle-like” conformation. It is envisioned that this change in surrounding polarity may allow for both the selective encapsulation and the stimulus-responsive release of guest molecules.

Fig. 1 Schematic representation of a cyclic polymer reverse micelle undergoing inversion due to changes in solvent polarity.

Although most of the previously reported amphiphilic units on amphiphilic homopolymers were rendered water-soluble through the use of charged groups (either carboxylate or ammonium functionalities), amphiphilicity which is independent of pH and charge is particularly important for many biological applications. Towards this end, our labs have reported a modular synthetic route towards the preparation of bifurcated amphiphilic units which are both neutral and independent of pH.³¹ The synthesis involves the successive attachment of a long alkane chain (in this case dodecane) and an oligoethylene glycol chain (in this case tetra(ethylene glycol)) to a tris-alcohol core to provide amphiphilic unit 1. This novel bifurcated amphiphile can be easily synthesized in five steps with an overall 76% yield.

In an attempt to ascertain the effect of polymer architecture (cyclic vs. linear) on the nature of amphiphilic homopolymers, both linear and cyclic amphiphilic homopolymers were prepared by attaching amphiphile 1 to analogous linear and cyclic polymer backbones. To our knowledge, this is the first report of an amphiphilic homopolymer which is cyclic in nature and it is expected that the cyclic amphiphilic homopolymer will exhibit some unusual properties as a result of its unique topology.

Results and discussion

The high functional group tolerance of Copper(I)-catalyzed Azide- Alkyne Cycloaddition (CuAAC) “click” chemistry³² as well as the ease of incorporating the requisite alkyne and azide functionalities onto complimentary moieties offers an attractive and versatile method for the attachment of amphiphilic units onto the cyclic polymer scaffolds. The CuAAC coupling reaction is particularly useful for macromolecular conjugations because the high efficiency of the reaction enables it to overcome the steric inhibition that typically limits the functionalization of polymers. This concept has been elegantly demonstrated with a range of macromolecular conjugation reactions including the attachment of dendrons onto linear polymer backbones to produce linear dendronized polymers,³³ the grafting of polymer chains onto linear polymer backbones to afford polymer brushes,³⁴ the grafting of linear polymers onto dendrimeric alkyne cores to yield well-defined stars,^35,36 as well as the coupling of both linear polymers³⁷ and dendrons³⁸ onto cyclic polymer backbones.

The combination of “click” cyclization to yield a cyclic polymer substrate followed by “click” functionalization (e.g. with dendrons to yield dendronized cyclic polymers), was reported for the first time, recently, and is particularly useful for examining structure/property relationships because exactly analogous functionalized polymers can be made from comparable linear and cyclic substrates.³⁸ This approach utilizes the previous established methods for “click” cyclization of linear α-alkynyl-ω-azido poly(acetoxystyrene) (PAS) polymers (3). ATRP polymerization of 4-acetoxystyrene from propargyl bromo-isobutyrate yielded the bromide terminated polymer (2) which could be converted by azidation of the terminal bromide, to produce linear-PAS precursor (3) for cyclization (Scheme 1). The cyclization was carried out under “pseudo high dilution” by adding the linear precursor (3) dropwise into a solution of the copper catalyst. The combination of the “click” cyclization technique with a range of living polymerization methods has been shown to be a versatile approach for the synthesis of cyclic polymers with a diversity of polymer backbones.^39–48 In agreement with previous polymer cyclizations, the c-PAS (4) showed a longer retention time than its linear counterpart (3) because the cyclic polymer is “locked” into a smaller conformation in solution (Fig. 2).

Scheme 1 Synthesis of linear (8) and cyclic-PHS-yne (6) scaffolds for click “grafting to” functionalization of polymer backbones.

Fig. 2 SEC traces of linear (blue) and cyclic (red) PAS highlighting a shift to a longer retention time post “click” cyclization.

The cyclic polymer, 4, represents an ideal substrate for preparing functionalized cyclic polymers and hybrid cyclic architectures since the acetoxy groups can be easily be hydrolyzed exposing a pendant hydroxyl group that can be readily functionalized. To this end, c-PAS (4) was quantitatively deprotected using mildly acidic conditions as confirmed by the loss of the acetate protons (2.1 ppm) via¹H-NMR of the resultant cyclic polyhydroxystyrene (c-PHS), 5 substrate. The linear analog, l-PHS (7) was prepared for comparison via LiAlH₄ reduction of the pendant acetoxy protecting groups of the polymer bromide (2) to ensure reduction of the ester moiety on the initiator of the polymer as well as of the esters on the repeat units. Both linear (7) and cyclic-PHS (5) were functionalized with alkynyl moieties through an activated ester functionalization reaction with 4-pentynoic acid and N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC). ¹H-NMR of the alkynylated PHS polymers showed clear incorporation of the alkynyl group and integration values comparing the aromatic protons of the styrenic backbone to those of the terminal alkynyl protons (approximately 4 to 1) verified the near quantitative functionalization of the repeat units. The SEC traces of alkynyl functionalized polymers 6 and 8 confirmed the retention of the cyclic architecture during the deprotection and alkynyl coupling steps, as c-PHS-yne 6 exhibited a distinctly more compact structure than its linear analog l-PHS-yne 8 (Fig. 3).

Fig. 3 SEC traces of linear (8) (blue) and cyclic (6) (red) alkynylated polymer scaffolds showing a longer retention time for the cyclic analogue.

The amphiphilic alcohol 1 was prepared from 1,1,1-tris(hydroxymethyl)ethane 9 because it contains 3 functionalities which could be selectively functionalized through protection/deprotection chemistry (Scheme 2). Two of the hydroxyls of 1,1,1-tris(hydroxymethyl)ethane 9 were first protected with a benzylidene functionality via the reaction of benzaldehyde dimethyl acetal in the presence of a catalytic amount of para-toluene sulfonic acid (p-TSA). The single free alcohol of compound 10 (isolated as a mixture of isomers, predominantly cis) was deprotonated with NaH and displaced the bromide of 1-bromododecane to introduce the non-polar dodecyl chain of acetal 11 in 97% yield. The benzylidene functionality of 11 was then partially deprotected by reaction with DiIsoButyl ALuminum Hydride (DIBAL-H) to unmask a second free alcohol functionality for selective functionalization, while protecting the remaining hydroxyl as a benzyl ether 12. The polar chain of amphiphile 1 was then introduced by reaction of the alkoxide anion of 12 with tetraethylene glycol monomethyl ether mesylate 13 to give benzyl protected amphiphile, 14, in 90% yield. The benzyl ether was then removed via hydrogenation with Pd/C catalyst at 10–15 psi of H₂ (g) to yield the racemic amphiphile, 1.

Scheme 2 Synthesis of azido-functionalized bifurcated amphiphilic unit for attachment to linear and cyclic polymer scaffolds.

To prepare the requisite amphiphilic azide, Amph-OH 1, was first converted to its mesylate derivative, 15, followed by nucleophilic displacement of the mesylate with NaN₃ to produce Amph-N₃, 16, in 86% yield. A large excess of NaN₃ (15 equiv) was used at elevated temperatures to ensure sufficient conversion to the azide despite the steric hindrance on the amphiphilic substrate.⁴⁹ Successful incorporation of the azide functionality was confirmed by ¹H-NMR monitoring the loss of the mesylate group, while FT-IR verified the appearance of the characteristic strong azide stretching frequency at 2100 cm⁻¹.

The chromatographically purified product, Amph-N₃16 was then coupled to linear and cyclic PHS-yne functionalized polymers, 8 and 6, respectively using typical “click” coupling conditions (Scheme 3). The successful coupling to yield cyclic (17) and linear (18) amphiphilic homopolymers was verified by ¹H-NMR, SEC, and MALDI-TOF MS. Using NMR, the structure of the product was confirmed by the disappearance of the terminal alkynyl proton (2.0–2.15 ppm), the appearance of the aromatic triazole proton (7.45–7.60 ppm), and the shift of the methylene units adjacent to the alkynyl moiety (∼2.70 ppm → ∼3.00 ppm) downfield upon conversion from the alkyne to the aromatic triazole functionality (Fig. 4).

Scheme 3 Synthesis of cyclic (17) and linear (18) amphiphilic homopolymers by “click” coupling procedure.

Fig. 4 ¹H-NMR of Amph-N₃16 (top), c-PHS-yne 6 (middle), and cyclic “click-functionalized” amphiphilic homopolymer 17 (bottom) confirming near quantitative coupling through loss of the terminal alkyne proton at 2.0 ppm.

Coupling efficiencies of the CuAAC conjugation reaction were calculated by ¹H-NMR integration values by comparing the aromatic protons of the styrenic polymer backbone (6.20–7.00 ppm) to the aryl proton of the formed triazole functionality (7.45–7.60) in addition to the loss of the terminal alkyne (2.00 ppm). In an attempt to increase the coupling efficiency, a range of reaction conditions were investigated including elevation of the reaction temperature as well as the use of a microwave reactor. It was found that increasing the temperature noticeably improved the coupling efficiency (∼74% at room temperature after 12 h; ∼90% at 100 °C after 30 min) as well as rate of the “click” coupling process but that microwave conditions (100 W, 100 °C) did not provide a quantifiable advantage beyond simply raising the temperature during the cycloaddition reaction.

SEC traces of cyclic amphiphilic homopolymer 17 exhibited a reduced retention time when compared to the alkynyl functionalized polymer scaffold c-PHS-yne 6 (Fig. 5). Additionally, the “click” coupled products remained monomodal in nature and retained their low polydispersities, further confirming the efficiency of the “click” coupling procedure. Comparison of the linear and cyclic “click”-functionalized amphiphilic polymers verified that the polymer architectures exhibit sufficient difference in retention time to be distinguished by SEC.

Fig. 5 SEC traces showing retention of monomodal peak and a shift to lower retention time upon “click” coupling (left) and exhibiting resolution between linear and cyclic amphiphilic homopolymer (right). The molecular weight values reported are based on calibration against linear polystyrene standards in THF.

To confirm the reproducibility of the “click” coupling process and its versatility in rapidly building up a library of amphiphilic homopolymers, two different sets of linear and cyclic cores (DP = 32 and 40) were functionalized using this route (Table 1).

Table 1 Summary of SEC and coupling efficiency data (calculated by ¹H-NMR integration values) for a series of linear and cyclic amphiphilic homopolymers of varying DP's. *

DP = 32	Linear				Cyclic
	Cpd. #	Mn^a	PDI	Coup. Eff.	Cpd. #	Mn^a	PDI	Coup. Eff.
PAS	3a	4 750	1.13	—	4a	4 440	1.13	—
PHS-yne	8a	6 860	1.11	99%	6a	6 750	1.11	99%
PHS-triaz-Amph	18a	14 240	1.08	96%	17a	13 000	1.10	92%

---

DP = 40	Linear				Cyclic
	Cpd. #	Mn^a	PDI	Coup. Eff.	Cpd. #	Mn^a	PDI	Coup. Eff.
a Mn values are based on calibration vs. linear polystyrene standards in THF.
PAS	3b	6 050	1.24	—	4b	5 010	1.29	—
PHS-yne	8b	8 890	1.16	99%	6b	8 500	1.24	99%
PHS-triaz-Amph	18b	15 660	1.13	96%	17b	15 670	1.15	93%

---

Both sets of amphiphilic homopolymers were also characterized by MALDI-TOF MS showing that linear and cyclic amphiphilic homopolymers exhibited very similar molecular weights (Fig. 6). The observed M_n matched closely with theoretical predictions confirming a >90% coupling of the amphiphiles to the core (Table 2). In addition, the spacing between major peaks crudely corresponded to that of the “click”-coupled repeat unit (∼703 Da) confirming the near-quantitative attachment of the amphiphile onto each repeat unit.

Fig. 6 MALDI-TOF Mass spectra of both linear (18) and cyclic (17) amphiphilic homopolymers exhibiting nearly identical molecular weights.

Table 2 Summary of molecular weight data obtained by MALDI-TOF MS highlighting linear and cyclic amphiphilic homopolymers of varying DP

MALDI-TOF MS
		Linear			Cyclic
DP	Mn,theor	Cpd #	Mn	PDI	Cpd #	Mn	PDI
32	22 700	18a	24 200	1.04	17a	23 300	1.05
40	28 320	18b	30 500	1.11	17b	30 800	1.10

---

After confirming the cyclic topology of amphiphilic homopolymer 17, and verifying the high coupling yields of the amphiphile, homopolymers 17 and 18 were tested for solubility in solvents with a wide-range of polarities. The resulting polymers exhibited solubility in non-polar solvents such as toluene and hexanes, as well as solubility in polar solvents such as methanol, although they did not exhibit solubility in water. Additionally, amphiphilic homopolymers 17 and 18 exhibited solubility in all of the “intermediate polarity” solvents investigated, such as tetrahydrofuran, methylene chloride, chloroform, dimethylformamide, etc. The broad range of solubility of these polymers in both hydrophilic and hydrophobic solvents allowed for further investigations of their host–guest chemistry.

In order to determine these material's abilities to encapsulate hydrophilic molecules in a contrasting non-polar environment, toluene was used as a medium for encapsulation experiments. Rose Bengal, a water-soluble dye, was used as a model compound for encapsulation because its concentration can be easily quantified by UV-Vis spectroscopy. Encapsulation experiments were performed by dissolving amphiphilic homopolymers 17 and 18 in toluene at a range of concentrations followed by the addition of an excess of Rose Bengal as a fine powder. The mixtures were then agitated for 12 h at room temperature via an orbital shaker and then filtered through a 200 μm in-line syringe filter to remove any unsolubilized dye from the heterogeneous solution. UV-Vis absorption spectra of the samples were obtained to quantify the solubility of the Rose Bengal into solutions of the polymer, whereas a control experiment performed in toluene but without polymer exhibited no UV signal for the Rose Bengal (Fig. 7). Rose Bengal solvated by amphiphilic homopolymer in non-polar toluene showed a bathochromic shift to λ_max = 562 nm (relative to Rose Bengal in ethanol, λ_max = 559 nm) suggesting that the environment around the Rose Bengal was less polar than that of bulk ethanol. This experiment confirms that the amphiphilic homopolymer can encapsulate Rose Bengal in hydrophobic environments and impart solubility in non-polar solvents. The saturation concentration of Rose Bengal exhibited a linear dependence with respect to the concentration of the amphiphilic polymer for both linear and cyclic polymer samples of DP = 40. Even at low concentrations (∼10 μM) the amphiphilic homopolymers imparted hydrophobic solubility on Rose Bengal, suggesting that both the linear and cyclic amphiphilic homopolymers can act as unimolecular reverse micelles (or have a very low CMC). Consistently, between concentrations of 10–300μM, the polymers could encapsulate approximately 3 equivalents of Rose Bengal per macromolecule. Interestingly, when comparing the linear and cyclic homopolymers, there did not appear to be a significant difference in either their solubilities or their Rose Bengal encapsulation efficiency (Fig. 8). A similarity in properties was also noted for the set of amphiphilic homopolymers with DP = 32. Although it seems theoretically likely that the entropically confined nature of the cyclic amphiphilic homopolymers may lead to a difference in solubility and encapsulation, such differences are either too minute to be easily measured or not substantial in the size range probed (DP= 32–40). Ultimately, the versatility of this synthetic approach allows the DP to be tuned during preparation of linear precursors enabling rapid access to linear and cyclic amphiphilic homopolymers in a wider range of sizes for subsequent studies.

Fig. 7 Visual confirmation of the encapsulation capabilities of c-PHS-Amph (18) for Rose Bengal dye. Saturated solution of Rose Bengal were prepared in a 10 mg mL⁻¹ solution of polymer in toluene (right) and a control solution of toluene without polymer (left).

Fig. 8 Graph of calculated Rose Bengal concentration vs. concentration of the polymer used to quantify encapsulation ability of amphiphilic homopolymer for hydrophilic Rose Bengal dye in toluene.

In order to test the stability of the reverse micellar assemblies in toluene, an extraction of the dye back into water was attempted. A small amount of water was added to the toluene/dye/polymer mixture and agitated for about 1 min and the mixture was allowed to phase separate. Despite the high solubility of RB in water, and it's insolubility in toluene (Fig. 7), the majority of the RB remained in the polymer/toluene solution (top layer, Fig. 9) even after repeated agitation. This suggests that the amphiphilic polymers exhibit a robust micellar conformation, and that the resulting polymer/dye complex is locked into a thermodynamic minimum which prevents sufficient rearrangement to enable interaction with, and rapid migration across the toluene/water interface.

Fig. 9 Attempted extraction of Rose Bengal from toluene solution containing amphiphilic homopolymer 17 by the addition of water and allowing phase separation (bottom layer).

Similar results have previously been reported by Thayumanavan and co-workers for Rose Bengal encapsulation with linear amphiphilic homopolymers. These authors attribute this observation to a highly stable micellar assembly within the “bad” solvent (for the guest) that prevents rapid dye release from the polymer and transfer into the “better” solvent.²⁷ Currently, these amphiphilic homopolymer systems are being studied in greater detail through the preparation of a range of polymer architectures bearing similar neutral amphiphilic units.

Conclusions

In conclusion, we have developed a novel bifurcated amphiphilic unit and optimized its attachment to both linear and cyclic polymer backbones through “click” coupling to an alkynyl functionalized polymeric scaffold. The “click” coupling strategy routinely yielded amphiphilic homopolymers with coupling efficiencies above 90%, as measured by ¹H-NMR integration. Two sets of both linear and cyclic amphiphilic homopolymers were prepared and their structure confirmed by ¹H-NMR, SEC, and MALDI-TOF MS.

These materials exhibited solubility in a diversity of solvents ranging from limited solubility in hexanes to good solubility in toluene and more moderately polar solvents as well as polar protic solvents such as methanol. Although these macromolecules were not water soluble, the versatility of the “click” synthetic approach should allow further explorations towards preparing uncharged water-soluble amphiphilic homopolymers. These amphiphilic homopolymers were shown to assist the dissolution of Rose Bengal, a water soluble dye, in non-polar toluene solvent. Within the size range investigated linear and cyclic amphiphilic homopolymers' encapsulation efficiency was investigated by UV/Vis measurements of dye encapsulation and shown to have nearly identical encapsulation abilities. Interestingly, efforts to extract Rose Bengal from the reverse micellar assembly in a poor solvent (toluene) into a good solvent (water) were inefficient confirming the stability of the amphiphilic homopolymer assemblies.

Experimental

Materials

All reagents were purchased from Aldrich or Alfa Aesar and were used without further purification unless noted otherwise. THF was distilled from CaH₂ and stored under dry N₂ gas before use. Tetraethylene glycol monomethyl ether was distilled prior to use. All other solvents were used directly from the bottle without additional purification. Deuterated solvents were purchased from Cambridge Isotope Laboratories, Inc. NMR (¹H, ¹³C) spectra were recorded on Varian-400MHz spectrometer at room temperature. Chemical shifts are reported in parts per million (ppm) relative to CDCl₃ (7.26 ppm for ¹H and 77.16 ppm for ¹³C), CD₂Cl₂ (5.32 ppm for ¹H and 54.00 ppm for ¹³C), and CD₃OD (3.31 ppm for ¹H and 49.00 ppm for ¹³C) as internal standard for calibration. Size Exclusion Chromatography (SEC) data were acquired on a Waters model 1515 isocratic pump (Milford, MA) with THF as the mobile phase with a 1 mL min⁻¹ flow rate with columns at a constant 30 °C. This system was operated with a three-series column set from Polymer Laboratories, Inc. consisting of PLgel 5μm MIXED-D (300mm × 7.5mm), PLgel 5μm 500 Å (300mm × 7.5mm), and PLgel 5μm 50 Å (300mm × 7.5mm) columns. The system was also equipped with a Model 2487 differential refractometer detector. The resulting molecular weights were calculated based on linear poly(styrene) standards using Waters Breeze software. MALDI spectra were collected using a Bruker-Daltonics MALDI-TOF Autoflex III mass spectrometer using trans-3-indole acrylic acid (20 mg mL⁻¹) as matrix, amphiphilic homopolymer (2 mg mL⁻¹), and sodium trifluoroacetate (2mg mL⁻¹) as counterion in a 20/10/3 ratio of matrix/polymer/counterion ratio. M_n and PDI of the resultant spectra were calculated using PolyTools software. UV-Vis data were performed with a Hewlett Packard 8452A UV diode array spectrometer at room temperature (25 °C). The synthesis of compounds 2–8³⁸ has previously been reported by our labs and SEC data describing these compounds are presented in Table 1 and NMR data agree with previous reported. The only deviation from the previously reported synthesis of 2 was the use of N,N,N′,N′,N′′-pentamethyl diethylene triamine (PMDETA) as a ligand during ATRP with anisole (25 wt%) as a co-solvent during ATRP. Bifurcated amphiphile precursors 10–12 has also been previously described by our labs.³¹

General procedure for encapsulation studies

Amphiphilic homopolymer was dissolved into a non-polar solvent (toluene) at a range of concentrations with a final volume of 2 mL. Excess dye/guest (∼0.010 g) was added and the vial sealed with parafilm. The collection of vials was shaken on an orbital shaker for approximately 12 h and then allowed to rest for an additional 4 h before decanting the liquid from the vial via a syringe and filtering through a 200 μm in-line syringe filter. The filtrate was then used to take UV/Vis spectra with a wavelength range from 450–650 nm at 25 °C. Dilution factors were utilized in order to keep the intensity of the UV signal within 0.1–0.6 absorbance units range. A molar extinction coefficient of 90,400 M⁻¹ cm⁻¹ (EtOH) was used to calculate the concentration of Rose Bengal in solution using Beer's Law of the corrected absorbances after taking into account the dilution factors used for each sample.

1-((2-((2-(2-(2-(2-Methoxyethoxy)ethoxy)ethoxy)ethoxy)methyl)-3-(dodecyloxy)-2-methylpropoxy)methyl)benzene (14)

To a 125 mL oven dried 2-neck round bottom flask containing a magnetic stir bar and fitted with a reflux condenser was added sodium hydride (60% wt. in mineral oil) (0.794 g w/min. oil, 0.476 g, 19.835 mmol) under N₂ atmosphere. The sodium hydride was washed with hexanes and decanted three times to remove mineral oil followed by placing under vacuum to remove excess solvent. The solid sodium hydride was then made into a slurry by addition of 20 mL of freshly distilled dry THF. The solution was then placed into an ice bath at 0 °C before the addition of a solution of alcohol 12 (1.669 g, 4.409 mmol) in 5 mL of THF via syringe dropwise. This solution was allowed to proceed for 30 min. before the addition of a solution of tetraethylene glycol monomethyl ether mesylate 13 (1.578 g, 5.511 mmol) in 5 mL of THF via syringe. The reaction was allowed to warm to room temperature before placing into a preheated oil bath at 70 °C and allowed to proceed for 24 h. It was then cooled to room temperature and placed into an ice bath whereupon water was added slowly to quench any excess sodium hydride. The reaction was then diluted with CH₂Cl₂ and washed with aqueous NaHSO₄ and brine solutions before being dried over MgSO₄, filtered, concentrated in vacuo, and chromatographed over silica gel with a 85/15 hexanes/EtOAc eluent to produce ether 14 as a clear liquid (90% yield). ¹H-NMR (CDCl₃, δ, ppm): 0.82 (t, 3H, J = 7.2 Hz, –CH₂CH₃), 0.90 (s, 3H, –CCH₃), 1.20 (br s, 18H, –OCH₂CH₂(CH₂)₉CH₃), 1.46 (p, 2H, J = 6.8 Hz, –OCH₂CH₂), 3.23 (s, 2H, –OCH₂C), 3.27–3.34 (m, 9H, –OCH₃, –C(CH₂O)₂, –OCH₂CH₂), 3.47–3.52 (m, 4H, –OCH₂CH₂), 3.52–3.63 (m, 12H, –O(CH₂CH₂O)₃–), 4.43 (s, 2H, –CH₂Ph), 7.18–7.31 (m, 5H, –PhH). ¹³C-NMR (CDCl₃, δ, ppm): 14.26, 17.64, 22.81, 26.31, 29.48, 29.64, 29.75, 29.78, 29.81, 32.04, 41.08, 59.15, 70.50, 70.61, 70.66, 70.70, 70.75, 71.16, 71.70, 72.01, 73.26, 73.31, 73.47, 74.17, 127.36, 127.38, 128.33, 139.13. ESI Mass calculated for [C₃₃H₆₀O₇ + H]⁺: 569.4412. Observed: 569.4490.

2-((2-(2-(2-(2-Methoxyethoxy)ethoxy)ethoxy)ethoxy)methyl)-3-(dodecyloxy)-2-methylpropan-1-ol (1)

The benzylic ether 14 (2.17 g, 3.815 mmol) was weighed directly into an oven dried 50 mL round-bottomed flask equipped with magnetic stir bar and stop-flow adapter. The ether starting material was then dissolved in 25 mL of methanol and 20% wt. palladium metal on charcoal (0.250 g) was added while stirring. The reaction flask was evacuated and back-filled twice with hydrogen gas supplied via a soccer ball pressurized with hydrogen. The reaction was allowed to proceed for 24 h with constant stirring. During the course of the reaction, the hydrogen supply was refreshed once. The reaction was then diluted with 50 mL of CH₂Cl₂ and the palladium was removed by vacuum filtration through a Celite cake. The cake was washed with a solution of MeOH and CH₂Cl₂ (1 : 4). The filtrate was collected and reduced under vacuum to give alcohol 1 as a clear oil (96% yield). ¹H-NMR (CDCl₃, δ, ppm): 0.85 (s, 3H,–CCH₃), 0.86 (t, 3H, J = 6.4 Hz, –CH₂CH₃), 1.24 (br s, 18H, –OCH₂CH₂(CH₂)₉CH₃), 1.53 (p, 2H, J = 7.2 Hz, –OCH₂CH₂), 2.95 (br s, 1H, –OH), 3.32–3.48 (m, 9H, –OCH₃, –C(CH₂O)₂, and –CCH₂CH₂), 3.53–3.57 (m, 4H, –OCH₂CH₂), 3.57–3.67 (m, 14H, –O(CH₂CH₂O)₃ and –CCH₂OH). ¹³C-NMR (CDCl₃, δ, ppm): 14.25, 17.65, 22.81, 26.26, 29.47, 29.59, 29.68, 29.73, 29.74, 29.79, 32.03, 40.66, 59.16, 69.23, 70.42, 70.60, 70.67, 70.69, 70.98, 71.94, 72.01, 75.57, 75.63. ESI Mass calculated for [C₂₆H₅₄O₇ + H]⁺: 479.3948. Observed: 479.4026. Analysis calculated for C₂₆O₇H₅₄: C, 65.23, H, 11.37, O, 23.40. Observed: C, 65.00, H, 11.22.

2-((2-(2-(2-(2-Methoxyethoxy)ethoxy)ethoxy)ethoxy)methyl)-3-(dodecyloxy)-2-methylpropyl methanesulfonate (15)

An oven dried 50 mL 2-neck round bottom flask equipped with magnetic stir bar and a 25 mL addition funnel under N₂ atmosphere was charged with amphiphilic alcohol 1 (0.715 g, 1.494 mmol), 15 mL of freshly distilled THF, and triethyl amine (0.302 g, 2.987 mmol). A solution of 10 mL of THF and methanesulfonyl chloride (0.342 g, 2.987 mmol) was added dropwise to the reaction flask in an ice bath at 0 °C and the reaction was allowed to slowly warm to room temperature and proceed for 16 h before diluting with CH₂Cl₂ and washed once each with aqueous solutions of NaHSO₄, NaHCO₃, and NaCl. The solution was then dried over MgSO₄, filtered, and concentrated in vacuo before being chromatographed over silica gel with a 60/40 hexanes/EtOAc eluent to produce a clear oil (96% yield). ¹H-NMR (CDCl₃, δ, ppm): 0.86 (t, 3H, J = 7.2 Hz, –CH₂CH₃), 0.99 (s, 3H, –CCH₃), 1.24 (br s, 18H, –OCH₂CH₂(CH₂)₉CH₃), 1.50 (p, 2H, J = 6.8 Hz, –OCH₂CH₂), 2.99 (s, 3H, –O₃SCH₃), 3.20–3.37 (m, 9H, –OCH₃, –C(CH₂O)₂, and –CCH₂CH₂), 3.51–3.67 (m, 16H, –O(CH₂CH₂O)₄), 4.12 (s, 2H, –CH₂O₃SCH₃). ¹³C-NMR (CDCl₃, δ, ppm): 14.25, 17.14, 22.80, 26.26, 29.47, 29.58, 29.68, 29.74, 29.75, 29.79, 32.02, 36.68, 40.51, 59.15, 70.49, 70.62, 70.66, 70.70, 70.71, 71.09, 71.78, 72.00, 72.61, 72.65, 73.03. ESI Mass calculated for [C₂₇H₅₆O₉S + H]⁺: 557.3723. Observed: 557.3802. Analysis calculated for C₂₇H₅₆O₉S: C, 58.24, H, 10.14, O, 25.86, S, 5.76. Observed: C, 58.04, H, 10.32.

1-(3-(2-(2-(2-(2-Methoxyethoxy)ethoxy)ethoxy)ethoxy)-2-azidomethyl)-2-methylpropoxy)dodecane (16)

Amphiphile mesylate 15 (1.100 g, 1.976 mmol) was placed into a oven dried 25 mL round bottomed flask fitted with magnetic stir bar and a reflux condenser under N₂ atmosphere. To this flask was added 11 mL of dry DMSO and NaN₃ (1.927 g, 29.634 mmol) and then placed into a pre-heated oil bath at 95 °C. The reaction was allowed to proceed at this temperature for 36 h before cooling to room temperature, diluting with EtOAc, and washed with a dilute brine solution. It was then dried over MgSO₄, filtered, and concentrated in vacuo. The crude reaction product was then chromatographed over silica gel using a 60/40 hexanes/EtOAc mobile phase to produce the azido-functionalized amphiphile 16 as a clear oil (86% yield). ¹H-NMR (CDCl₃, δ, ppm): 0.87 (t, 3H, J = 6.8 Hz, –CH₂CH₃), 0.93 (s, 3H, –CCH₃), 1.25 (br s, 18H, –OCH₂CH₂(CH₂)₉CH₃), 1.52 (p, 2H, J = 6.8 Hz, –OCH₂CH₂), 3.20 (dd, 2H, J = 9.2 Hz, 11.2Hz, –C(CHH)₂), 3.24–3.31 (m, 4H, –C(CHH)₂ and –CCH₂N₃), 3.36 (t, 2H, J = 6.8 Hz, –OCH₂CH₂), 3.38 (s, 3H, –OCH₃), 3.53–3.58 (m, 4H, –OCH₂CH₂), 3.60–3.67 (m, 12H, –O(CH₂CH₂O)₃). ¹³C-NMR (CDCl₃, δ, ppm): 14.28, 18.19, 22.83, 26.31, 29.50, 29.62, 29.71, 29.77, 29.82, 32.05, 41.27, 55.67, 59.19, 70.52, 70.64, 70.71, 70.75, 70.77, 71.09, 71.70, 72.04, 73.49, 74.19. IR (KBr): 1100 cm⁻¹ (C–O), 2100 cm⁻¹ (–N₃), 2800–2950 cm⁻¹ (C–H). ESI Mass calculated for [C₂₆H₅₃N₃O₆ + H]⁺: 504.4013. Observed: 504.4085, 476.4032 (–N₂). Analysis calculated for C₂₆O₆H₅₃N₃: C, 61.99, H, 10.61, N, 8.34. Observed: C, 62.04, H, 10.62, N, 8.18.

c-PHS-Amph (17a)

c-PHS-yne 6 (0.007 g, 0.035 mmol) was weighed out into a 5 mL RBF and dissolved into 2 mL of DMF. To this solution was added Amph-N₃ (0.019 g, 0.039 mmol) and PMDETA (0.006 g, 0.035 mmol) before degassing by passing the solution through 3 freeze-pump-thaw cycles. While frozen, Cu(I)Br (0.005 g, 0.035 mmol) and Cu(0) powder (0.001 g, 0.018 mmol) was added and the reaction was evacuated and backfilled with N₂ gas twice. Upon thawing, the sealed reaction vessel was placed at 100 °C for 30 min. before diluting with CH₂Cl₂ and washing the organic portion twice with aqueous NH₄Cl solutions, once with a brine solution, drying over MgSO₄, filtering, and concentrating in vacuo. The crude reaction mixture was then re-dissolved into ∼5 mL of CH₂Cl₂ and passed through a small plug of silica gel with CH₂Cl₂ as the mobile phase. The solution was then concentrated in vacuo and placed under high vacuum to give a clear glass. (75% yield). ¹H-NMR showed a 92% coupling efficiency. ¹H-NMR (CD₂Cl₂, δ, ppm): 0.76–0.96 (br s, 6H, –CCH₃ and –CH₂CH₃), 1.17–1.39 (br s, 18H, –OCH₂CH₂(CH₂)₉CH₃), 1.47–1.61 (br s, 2H, –OCH₂CH₂), 2.84–3.00 (br s, 2H, O=CCH₂CH₂C), 3.00–3.10 (br s, 2H, O=CCH₂CH₂C), 3.10–3.24 (br d, 4H, –C(CH₂O)₂), 3.24–3.32 (br s, 3H, –OCH₃), 3.32–3.42 (br m, 2H, –OCH₂CH₂), 3.42–3.50 (br s, 2H, –OCH₂CH₂O), 3.50–3.66 (br m, 14H, –OCH₂CH₂O(CH₂CH₂O)₃), 4.24–4.36 (s, 2H, –CCH₂–triazole), 6.20–7.10 (br m, 4H, -Ph–H), 7.42–7.64 (s, 1H, –CH=C (triazole)). ¹³C-NMR (CDCl₃, δ, ppm): 14.25, 18.27, 21.08, 22.78, 26.31, 29.45, 29.61, 29.75, 32.00, 33.97, 39.60–40.70, 41.05, 53.30, 59.10, 70.41, 70.56, 70.61, 70.64, 70.70, 70.89, 71.63, 71.95, 73.42, 73.65, 121.12, 123.35, 128.40, 140–144, 145.42, 148.65, 171.09. SEC: M_n = 13000, PDI = 1.10. MALDI-TOF MS (m/z): M_n = 23300, PDI = 1.05.

l-PHS-Amph (18a)

Reaction was performed using an identical procedure as for c-PHS-Amph 17, except that l-PHS-yne 8 was used as the alkyne portion for click coupling. (78% yield) ¹H-NMR showed a 96% coupling efficiency. ¹H-NMR (CD₂Cl₂, δ, ppm): 0.78–0.95 (br, 6H, –CCH₃ and –CH₂CH₃), 1.18–1.40 (br s, 18H, –OCH₂CH₂(CH₂)₉CH₃), 1.48–1.61 (br s, 2H, –OCH₂CH₂), 2.86–3.01 (br s, 2H, O=CCH₂CH₂C), 3.01–3.12 (br m, 2H, O=CCH₂CH₂C), 3.12–3.26 (br d, 4H, –C(CH₂O)₂), 3.26–3.34 (br s, 3H, –OCH₃), 3.35–3.43 (br s, 2H, –OCH₂CH₂), 3.43–3.50 (br s, 2H, –OCH₂CH₂O), 3.50–3.67 (br, 14H, –OCH₂CH₂O(CH₂CH₂O)₃), 4.23–4.38 (s, 2H, –CCH₂-triazole), 6.13–7.04 (br m, 4H, –Ph–H), 7.38–7.57 (br s, 1H, –CH=C (triazole)). ¹³C-NMR (CD₂Cl₂, δ, ppm): 14.48, 18.52, 21.54, 23.24, 26.78, 29.92, 30.08, 30.22, 32.47, 34.36, 40–41.06, 41.45, 53.65, 59.15, 70.82, 70.89, 71.01, 71.05, 71.09, 71.43, 72.04, 72.41, 73.86, 74.11, 121.74, 123.75, 128.92, 141.70–144.70, 145.84, 149.35, 171.66. SEC: M_n = 14240, PDI = 1.08. MALDI-TOF MS (m/z): M_n = 24200, PDI = 1.04.

Acknowledgements

The authors acknowledge Tulane University, the NSF (CAREER DMR#0844662) for financial support, the Louisiana Board of Regents for a Graduate Fellowship (B.A.L), the NSF (MRI #0619770) for enabling MALDI-TOF and ESI mass spectral characterization.

References

1. Y. Ma, T. Cao and S. E. Webber, Macromolecules, 1998, 31, 1773–1778.
2. V. P. Torchilin, J. Controlled Release, 2001, 73, 137–172.
3. G. Riess, Prog. Polym. Sci., 2003, 28, 1107–1170.
4. R. K. O'Reilly, C. J. Hawker and K. L. Wooley, Chem. Soc. Rev., 2006, 35, 1068.
5. K. Procházka, M. K. Baloch and Z. Tuzar, Makromol. Chem., 1979, 180, 2521–2523.
6. A. Guo, G. Liu and J. Tao, Macromolecules, 1996, 29, 2487–2493.
7. M. Iijima, Y. Nagasaki, T. Okada, M. Kato and K. Kataoka, Macromolecules, 1999, 32, 1140–1146.
8. B. S. Sumerlin, A. B. Lowe, D. B. Thomas, A. J. Convertine, M. S. Donovan and C. L. McCormick, J. Polym. Sci., Part A: Polym. Chem., 2004, 42, 1724–1734.
9. K. B. Thurmond, T. Kowalewski and K. L. Wooley, J. Am. Chem. Soc., 1996, 118, 7239–7240.
10. K. B. Thurmond, T. Kowalewski and K. L. Wooley, J. Am. Chem. Soc., 1997, 119, 6656–6665.
11. H. Huang, T. Kowalewski, E. E. Remsen, R. Gertzmann and K. L. Wooley, J. Am. Chem. Soc., 1997, 119, 11653–11659.
12. Q. Zhang, E. E. Remsen and K. L. Wooley, J. Am. Chem. Soc., 2000, 122, 3642–3651.
13. C. J. Hawker, K. L. Wooley and J. M. J. Frechet, J. Chem. Soc., Perkin Trans. 1, 1993, 1287.
14. A. Heise, J. L. Hedrick, C. W. Frank and R. D. Miller, J. Am. Chem. Soc., 1999, 121, 8647–8648.
15. D. R. Vutukuri, S. Basu and S. Thayumanavan, J. Am. Chem. Soc., 2004, 126, 15636–15637.
16. M. Liu, K. Kono and J. M. J. Fréchet, J. Controlled Release, 2000, 65, 121–131.
17. L. Arnt and G. N. Tew, J. Am. Chem. Soc., 2002, 124, 7664–7665.
18. L. Arnt and G. N. Tew, Macromolecules, 2004, 37, 1283–1288.
19. D. Shen, L. Wang, Z. Pan, S. Cheng, X. Zhu and L.-J. Fan, Macromolecules, 2011, 44, 1009–1015.
20. J. Kim and T. M. Swager, Nature, 2001, 411, 1030–1034.
21. N. Reitzel, D. R. Greve, K. Kjaer, P. B. Howes, M. Jayaraman, S. Savoy, R. D. McCullough, J. T. McDevitt and T. Bjørnholm, J. Am. Chem. Soc., 2000, 122, 5788–5800.
22. Z. Bo, J. P. Rabe and A. D. Schlüter, Angew. Chem., Int. Ed., 1999, 38, 2370–2372.
23. M. F. Ilker, K. Nüsslein, G. N. Tew and E. B. Coughlin, J. Am. Chem. Soc., 2004, 126, 15870–15875.
24. M. F. Ilker, H. Schule and E. B. Coughlin, Macromolecules, 2004, 37, 694–700.
25. K. Lienkamp, A. E. Madkour, A. Musante, C. F. Nelson, K. Nüsslein and G. N. Tew, J. Am. Chem. Soc., 2008, 130, 9836–9843.
26. S. Basu, D. R. Vutukuri, S. Shyamroy, B. S. Sandanaraj and S. Thayumanavan, J. Am. Chem. Soc., 2004, 126, 9890–9891.
27. S. Basu, D. R. Vutukuri and S. Thayumanavan, J. Am. Chem. Soc., 2005, 127, 16794–16795.
28. E. N. Savariar, S. V. Aathimanikandan and S. Thayumanavan, J. Am. Chem. Soc., 2006, 128, 16224–16230.
29. R. Saito and K. Yamaguchi, J. Polym. Sci., Part A: Polym. Chem., 2005, 43, 6262–6271.
30. R. Saito, K. Yamaguchi, T. Hara and C. Saegusa, Macromolecules, 2007, 40, 4621–4625.
31. K. A. Willham, B. A. Laurent and S. M. Grayson, Tetrahedron Lett., 2008, 49, 2091–2094.
32. H. C. Kolb, M. G. Finn and K. B. Sharpless, Angew. Chem., Int. Ed., 2001, 40, 2004–2021.
33. B. Helms, J. L. Mynar, C. J. Hawker and J. M. J. Fréchet, J. Am. Chem. Soc., 2004, 126, 15020–15021.
34. H. Gao and K. Matyjaszewski, J. Am. Chem. Soc., 2007, 129, 6633–6639.
35. H. Gao and K. Matyjaszewski, Macromolecules, 2006, 39, 4960–4965.
36. O. Altintas, B. Yankul, G. Hizal and U. Tunca, J. Polym. Sci., Part A: Polym. Chem., 2006, 44, 6458–6465.
37. K. Zhang, M. A. Lackey, Y. Wu and G. N. Tew, J. Am. Chem. Soc., 2011, 133, 6906–6909.
38. B. A. Laurent and S. M. Grayson, J. Am. Chem. Soc., 2011, 133, 13421–13429.
39. B. A. Laurent and S. M. Grayson, J. Am. Chem. Soc., 2006, 128, 4238–4239.
40. D. E. Lonsdale, C. A. Bell and M. J. Monteiro, Macromolecules, 2010, 43, 3331–3339.
41. D. M. Eugene and S. M. Grayson, Macromolecules, 2008, 41, 5082–5084.
42. A. S. Goldmann, D. Quemener, P.-E. Millard, T. P. Davis, M. H. Stenzel, C. Barner-Kowollik and A. H. E. Mueller, Polymer, 2008, 49, 2274–2281.
43. X.-P. Qiu, F. Tanaka and F. M. Winnik, Macromolecules, 2007, 40, 7069–7071.
44. J. Xu, J. Ye and S. Liu, Macromolecules, 2007, 40, 9103–9110.
45. Z. Ge, Y. Zhou, J. Xu, H. Liu, D. Chen and S. Liu, J. Am. Chem. Soc., 2009, 131, 1628–1629.
46. S. Binauld, C. J. Hawker, E. Fleury and E. Drockenmuller, Angew. Chem., Int. Ed., 2009, 48, 6654–6658.
47. J. N. Hoskins and S. M. Grayson, Macromolecules, 2009, 42, 6406–6413.
48. E. D. Pressly, R. J. Amir and C. J. Hawker, J. Polym. Sci., Part A: Polym. Chem., 2011, 49, 814–819.
49. Proper safety precautions such as performing the reactions on small scales, the use of a blast shield, and in the absence of halogenated solvents were used to be sure to prevent potential problems associated with azides.

---