Synthesis of well-defined core–shell nanoparticles based on bifunctional poly(2-oxazoline) macromonomer surfactants and a microemulsion polymerization process

Abstract

Particles in the sub-100 nm range have attracted widespread attention in the past few years due to their application in drug delivery and diagnostics. Here we describe the synthesis of two bifunctional, amphiphilic poly(2-oxazoline) macromonomers with multiple acrylate groups in their hydrophobic block and azide or primary amino end groups. The amphiphilic macromonomers were applied in a microemulsion polymerization to form well-defined core-crosslinked nanoparticles with surface functional azide or amine groups. Therefore, an amphiphilic poly(2-oxazoline) was prepared by cationic ring-opening polymerization of 2-methyl-2-oxazoline to form the hydrophilic block and a mixture of 2-heptyl-2-oxazoline and 2-(5-pentyl-[(1,2,3-triazol)-4-yl-methacrylat)]-oxazoline to form the hydrophobic block and was terminated with an azide moiety as end group. The introduction of multiple methacrylate groups into the poly(2-oxazoline) macromonomers serve as a stabilizer in the microemulsion process to covalently link the polymer to the particle core. Variable particle sizes of 20–75 nm have been prepared by encapsulating different amounts of 1,6-hexanedioldimethacrylate (HDDMA) to swell the micellar core before subsequent crosslinking takes place. Finally, particle surface functionalization was achieved by converting the terminal azide group via Staudinger-reaction to a primary amine group. Nanoparticles with surface primary amine groups were functionalized with folic acid (FA), a GRGDS-peptide derivative and fluorescein isothiocyanate (FITC) by simple amidation reaction (FA, RGD-peptide) or thiourea formation (FITC).

---

Introduction

The fabrication of biocompatible, non-toxic and blood stable nanoparticles has found increasing interest in the past few years due to their potential application in drug delivery^1–3 and diagnostics.^4,5 Therefore many different technologies have been developed to prepare such nanoparticles⁶ either from dispersions of preformed polymers,^7,8 cross-linking of polymer micelles in the core⁹ or shell,^10,11 and by heterogeneous polymerization techniques. The latter category includes emulsion,¹² miniemulsion¹³ and microemulsion^14,15 polymerization of monomers. The miniemulsion process has become very popular in the past decade due to its versatility to prepare a wide range of useful materials by using a high shear device and a hydrophobe for nanodroplet stabilization.^13,16–18 Moreover much progress has been made in the past years in the miniemulsion process to replace the low molecular weight surfactants by polymer based amphiphiles, employing PEG-based surfactants,¹⁹ macroinitiators,^20,21 and macromonomers^22,23 or MacroRAFT agents.²⁴

A second approach that attracted less attention in the past years to prepare well-defined nanoparticles with tunable size and functionalities is the microemulsion process. The main difference to the miniemulsion is the higher concentration of surfactant required for a complete particle coverage. The kinetics of the polymerization process has been described by the Candau–Leong–Fitch model (CLF-model)²⁵ where preformed monomer swollen micelles or microdroplets are initiated by radicals formed in the aqueous phase, monomers are then transported from unpolymerized micelles through the aqueous phase to grow the particles containing a polymer chain. The process leads in general to latex particles in a size range of 10–100 nm and empty micelles at the end of polymerization. Much work has been devoted in the past to elucidate the dependence of the particle size on initiator concentration,²⁶ surfactant concentration,²⁷ and type of surfactant and co-surfactant.^28,29 Moreover, apart from the free radical polymerization also controlled radical polymerization techniques such as NMP,³⁰ ATRP³¹ and RAFT^32,33 have been applied to produce polymer particles by microemulsion.³⁴ Although the particle size in the sub 100 nm range is considered to be ideal for drug delivery and diagnostics, the large amount of surfactants presents a major challenge for medical application of such particles especially when cationic surfactants are used.³⁵

A strategy to circumvent that issue is the use of polymerizable surfactants,³⁶ or polymer based macroinitiators and macromonomers that are covalently attached to the particle at the end of the polymerization process. Although various block copolymers, such as poly(styrene)-b-poly(ethyleneoxide) (PS-b-PEO),³⁷ poly(propylene oxide)-poly(ethylene oxide) (PPO-b-PEO)³⁸ block copolymers or poly(MMA-b-MAA)³⁹ were used as dispersant in microemulsions only two examples of using macromonomers in a microemulsion approach as a co-surfactant have been reported. David and coworker deployed poly(N-acetylethyleneimine) with cinnamoyl polymerizable end group⁴⁰ and maleic acid terminated poly(N-acetylethyleneimine),⁴¹ that were included in the microemulsion recipe as a co-surfactant with SDS to produce nanoparticles of methyl methacrylate and butyl methacrylate copolymers with a diameter of 40–100 nm.

For the synthesis of nanoparticles with the possibility to introduce also targeting and/or imaging moieties on the particle surface, bifunctional polymer surfactants are required where one functional group acts as macromonomer to covalently attach the polymer surfactant to the latex particle and a second orthogonal functional groups is required for further particle surface modification. A class of polymers that has recently attracted increased attention in the biomedical field are poly(2-oxazolines).⁴² They can be prepared by cationic ring-opening polymerization, allow the synthesis of different polymer architectures and many different monomers with various functional groups have been reported for post-analogous functionalization.^43–45 Moreover, depending on the chain length of the alkyl group, polymer polarity can be tuned from hydrophilic (alkyl = methyl or ethyl) to hydrophobic^46,47 (alkyl > propyl) which makes them also a promising alternative to polyethylene oxide (PEO).^48–51 In the early 90^th, different research groups have already prepared poly(2-oxazoline) macromonomers mainly for micron sized polymer particles.^52–55

Herein, we report the first synthesis and characterization of polymeric nanoparticles in a microemulsion polymerization based on bifunctional poly(2-oxazoline) macromonomers. The concept involves the use of an bifunctional amphiphilic polymer surfactant that carries multiple acrylate groups in the hydrophobic block to serve as a macromonomer during the formation of the nanoparticles. Particle sizes can be varied between 20 and 75 nm depending on the amount of core cross-linker used. Furthermore, to demonstrate the versatility of this approach an amino end-group of the poly(2-oxazoline) macromonomer was used for post-analogous modification of the particles with folic acid, a RGD-peptide derivative and fluorescein.

Results and discussion

Synthesis of the poly(2-oxazoline) monomer with methacrylate side chain

The synthesis of the acrylate functionalized 2-oxazoline monomer (5) is shown in Scheme 1. The azide precursor 2-(5-azidopentyl)-2-oxazoline (4) was synthesized according to Lav et al.⁵⁶ starting with 6-bromohexane acid and was obtained in an overall yield of 38%.

Scheme 1 Synthetic procedure for methacrylate-functionalized 2-oxazoline 5.

Monomer 5 was obtained as a yellow oil in 76% yield via a copper(I)-catalyzed-alkyne–azide-cycloaddition (CuAAC)⁵⁴ by reacting propargyl methacrylate with 4. The structure of monomer 5 was analyzed by ¹H NMR, ¹³C NMR spectroscopy and ESI/MS. The ¹H NMR spectrum of 5 is depicted in Fig. 1 showing two characteristic signals of the methylene groups of the oxazoline ring e and f at 3.70 and 4.18 ppm, respectively. Moreover, the two signals at 5.55/6.10 ppm can be assigned to the protons i and j of the vinyl group.

Fig. 1 ¹H NMR spectrum (CDCl₃) of 5.

Conversion of the azide moiety was also monitored by FTIR spectroscopy indicating the complete disappearance of the characteristic azide peak at ν = 2090 cm⁻¹ (Fig. 2). Moreover, the C=O stretching bond was observed at 1715 cm⁻¹ in agreement with the presence of the acrylate group.

Fig. 2 FTIR-spectra of compound 4 and 5.

Synthesis of the bifunctional poly(2-oxazoline) macromonomer

We decided to incorporate monomer 5 within the hydrophobic block together with 2-heptyl-2-oxazolines followed by the hydrophilic block based on 2-methyl-2-oxazolines to give P1 (Scheme 2). P1 was terminated with sodium azide to form an azide end group⁵⁷ that was subsequently converted to a primary amine end group using the Staudinger reaction⁵⁸ during the aqueous work up to give P2. FTIR spectroscopy and the disappearance of the signal at 2133 cm⁻¹ confirmed the successful conversion of the azide moiety to the amine group (see Fig. S4†). Formation of the amino end group was verified by performing a Ninhydrin test of P2 that turned blue due to the formation of the Ruthman's complex at λ = 577 nm (see Fig. S5A†). Moreover, quantification of the amino end group was carried out by reaction P2 with 2-bromomethyl naphthalene. Characterization of the functionalized polymer P2 revealed 72% amino end groups (see Fig. S5B†).

Scheme 2 Synthesis of the polymers P1 and P2.

Introduction of the bifunctional monomer 5 in P1 was successful and no cross reaction with the cationic polymerization mechanism was observed. SEC analysis displayed a monomodal SEC curve with a narrow polydispersity index of 1.13 for P1. Moreover, ¹H NMR and FTIR-spectroscopy (see Fig. 3 and S4†) confirmed that the acrylate function was still intact in P1 after the cationic polymerization process and P2 after the azido end group modification to primary amines. In water, the polymers formed micelles with a diameter of 19.0 ± 4.8 nm in agreement with literature results of other poly(2-oxazoline) block copolymers of similar molar mass and ratio of hydrophilic : hydrophobic block length (Table 1).^46–48

Fig. 3 ¹H NMR of P2 (CDCl₃) and SEC curves of P1 and P2.

Table 1 Analytical data of the polymers P1 and P2

Polymer	x^a	y^a	z^a	Mn^a/g mol^−1	Mn^b/g mol^−1	Đ^b	dh^c/nm
a Molar mass was determined by ^1H NMR spectroscopy and end group analysis, values in bracket quoted the theoretically used eq.b Molar mass and polydispersity indices (PDI) were obtained by SEC with PMMA standards in DMF/5 mg mL^−1 LiBr.c Hydrodynamic diameter (dh) were determined by DLS measurements from 1 mM polymer solutions in water at RT, PDI are given in brackets.
P1	5(4)	3(4)	30(30)	4380	8480	1.13	19.0 ± 4.8 (0.21 ± 0.01)
P2	5(5)	3(3)	32(30)	4520	7980	1.25	17.2 ± 3.7 (0.24 ± 0.01)

---

Nanoparticle synthesis via microemulsion polymerization

By using the microemulsion polymerization and different amount of core cross linker during the polymerization reaction, nanoparticles of different sizes were synthesized.

In our microemulsion approach (Scheme 3) the block copolymers P1 or P2 served as a surfactant with multiple acrylate groups in the hydrophobic side chain. This should allow the formation of stable polymeric nanoparticles by a free radical polymerization initiated with AIBN (see Table 2, NP1 to NP6). Furthermore, 1,6-hexanedioldimethacrylate (HDDMA) was used as a cross-linker in varies amounts to swell the micellar core formed by P1 or P2 and thus control particle size. The encapsulated HDDMA amount was varied from 0 to 100 wt%. Table 2 summarizes the results of particle sizes in methanol and water.

Scheme 3 Schematic representation of the microemulsion polymerization approach with the bifunctional, amphiphilic macromonomers P1 and P2 (A = homogenization step, 30 min deoxygenation with argon and 5 min sonication; B = polymerization step, 65 °C, over night).

Table 2 Analytical data of the nanoparticles NP1–NP6

Sample	Polymer	HDDMA^a/wt%	dh/nm (H2O)^b	dh/nm (MeOH)^b
a Microemulsion conditions: 0–100 wt% 1,6-hexanediol dimethacrylate, 5 wt% heptadecane, 0.01–0.10 wt% AIBN, 30 min degassed, 5 min sonication, 65 °C overnight.b Hydrodynamic diameter (dh) were determined by DLS measurements of 1 mg mL^−1 NP solutions at RT (Zetasizer SZ from Malvern), PDI of DLS measurements are given in brackets.
NP1	P1	0	25.63 ± 1.20 (0.46 ± 0.01)	20.80 ± 6.88 (0.37 ± 0.01)
NP2	P1	50	47.04 ± 6.84 (0.24 ± 0.02)	41.48 ± 2.34 (0.16 ± 0.01)
NP3	P1	100	72.64 ± 6.05 (0.56 ± 0.12)	67.99 ± 6.57 (0.17 ± 0.02)
NP4	P2	0	26.15 ± 3.07 (0.22 ± 0.01)	18.66 ± 3.57 (0.22 ± 0.01)
NP5	P2	50	42.16 ± 5.43 (0.56 ± 0.07)	37.80 ± 1.91 (0.19 ± 0.01)
NP6	P2	100	70.25 ± 16.60 (0.41 ± 0.01)	69.38 ± 2.66 (0.13 ± 0.01)

---

Several observation could be made. Firstly, the results showed a clear correlation between the amount of encapsulated HDDMA and the resulting particle size. Whereas the crosslinked micelle without any additional HDDMA displayed a size of 25.15 nm for P1 and 26.15 nm for P2, the use of additional HDDMA increased the particle size 2.8 times to 72.64 nm with P1 (see Table 2, NP1 to NP3) and 2.7 fold for P2 with the terminal amino groups (see Table 2, NP4 to NP6). Secondly, there is no major difference between particle size in methanol and water suggesting a densely crosslinked particle core that cannot swell anymore in methanol as non-selective solvent. On the other, rather small changes in particle diameter have still huge effect on particle volume. Particle volume for NP1 decreased by 47% in methanol compared to water whereas particle diameter decreased only from 25.6 to 20.8 nm by 19%. Particle size was always smaller in methanol and particle volume decreased from 3.7% to 64% in methanol compared to the values in water (see Table S1†). The effect on particle volume shrinkage is more pronounced for the smaller nanoparticles NP1 with 47% and NP4 and 64% and decreases with increasing particle size to reach 19% and 3.7% for NP3 and NP6, respectively (see Table S1†).

The results of the DLS measurements were confirmed by TEM measurements (Fig. 4). Moreover, this technique verified the spherical morphology of the particles and disclosed more information about the nature of the particles. A significant difference was observed between NP4 (Fig. 4a) and NP5/NP6 (Fig. 4b and c). NP4 showed a nanocapsule-like structure suggesting that the polymer micelle core is not densely packed most likely due to the sterically demanding side chains of the hydrophobic block.⁵⁵ However, when HDDMA was encapsulated in the micellar core and subsequently cross-linked to give NP5 and NP6, a much higher mass contrast of the particles was observed by TEM measurements suggesting a successful core polymerization of the HDDMA monomer.⁵⁹

Fig. 4 TEM images (c = 0.01 mg mL⁻¹ in MeOH) and DLS curves (c = 1 mg mL⁻¹ in MeOH) of NP4 (a), NP5 (b) and NP6 (c).

Surface functionalization of the amino functionalized nanoparticles (NP4–NP6)

To demonstrate the versatility of surface functionalization of our nanoparticles we decided to use two different targeting motifs, folic acid (FA) and a RGD-containing peptide (GRGDS(Ahx)₂F), and a fluorescein dye (FITC) as an imaging molecule. The choice of these moieties was based on several considerations. First, the up-regulation of folate receptors on the surface of different cancer cells such as ovary, lung, kidney and breast cancer cells is well-known.⁶⁰ In the past years it has been shown that various types of folic acid modified nanoparticles such as polymer micelles,^61–63 nanogels,⁶⁴ and nanoparticles based on chitosan⁶⁵ or albumin⁶⁶ can bind to folate receptors to facilitate receptor mediated endocytosis thus making folate receptors attractive targets. Secondly, active targeting based on integrin-receptor targeting motifs such as the RGD-peptide^67–69 has attracted also much attention in the years and has been mainly investigated for polymer micelles^70–72 and micelles with cross-linked core or shell^73,74 and represents an alternative approach to the folate receptor. Thirdly, FITC-labeling of polymers,⁷⁵ dendrimers⁷⁶ and nanoparticle⁷⁷ is a very common procedure to obtain fluorescently modified (macro)molecules for tracking in cell experiments. Folic acid was attached to the nanoparticle surface by using N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide (EDC) mediated amidation reaction. However, the bulky peptide GRGDS(Ahx)₂F could only be attached to the particle surface by using the more reactive tetramethylfluoroformamidinium hexafluorophosphate (TFFH) as coupling reagent via the formation of acid fluorides. Side products were separated by dialysis against water followed by lyophilization and precipitation in diethyl ether. Three nanoparticles NP4–NP6 were functionalized with FA to give NP4-NP6-FA and GRGDS(Ahx)₂F to give NP4-NP6-P (Scheme 4). The particles were afterwards characterized by DLS and by UV/Vis spectroscopy to determine their size and the degree of surface modification.⁷⁸ Finally, the fluorescent probe was attached to the particle surface via addition to the primary amine at 45 °C in the dark and formation of a thiourea bond. The work-up of the product was carried out by repeated precipitating in cold diethyl ether. The characteristic C=S stretch vibrations at ν = 1187 cm⁻¹ was observed in FTIR and confirmed the successful covalently bonding of the dye to the particle surface (Fig. 5).

Scheme 4 General procedure of surface functionalization of NP4–NP6 with (I) FA, (II) GRGDS(Ahx)₂F, and (III) FITC.

Fig. 5 NP5-FITC with primary amino groups before and after reaction with fluorescein isothiocyanate (FITC).

As can be seen from Table 3, modification of the nanoparticles displayed only a moderate increase of the particle size after modification with FA and FITC whereas particle size increased much more after modification with the sterically demanding GRGDS(Ahx)₂F (Table 3, NP4-NP6-RGD). Moreover, the size of the functional group had a large effect on the degree of surface modification with values of 1.74–2.0 × 10⁻⁴ mol L⁻¹ for folic acid modification (NP4-NP6–FA) compared to 0.11–0.18 × 10⁻⁴ mol L⁻¹ for peptide modification. Rossin et al. showed a degree of FA modification of 1% of available COOH groups on a 20 nm shell crosslinked NP. Here, the authors functionalized their NPs with further bioactive molecules, hence, a higher degree of modification was not required.⁷⁹ Furthermore, Pan and coworkers determined a shell functionalization of FA of 66% by calculating the aggregation number within the shell crosslinked NPs. However, the modification of these 40 nm sized NPs took not place with an end group but rather the side chains of the hydrophilic part were labeled.⁸⁰ In comparison, the RGD-functionalized NP of the workgroup of Yao exhibited a particle surface modification of 95% by using c(RGDyC) and a maleiimide-modified PEG-b-PLA-nanoparticles in a Michael reaction.^39,81 This high amount of RGD-functionalization depends on the dispensation with coupling reagents, e.g. TFFH. Mekuria showed a peptide functionalization of liposomes within a molar ratio of 16 : 1.⁸²

Table 3 Analytical data of the surface modified nanoparticles NP4–NP6

No.	HDDMA^a wt%	Surface functionality	dh^b/nm (MeOH)	c^c/10^−4 mol L^−1	Degree of modification^d/wt%
a Microemulsion conditions: 0–100 wt% 1,6-hexanediol dimethacrylate, 5 wt% heptadecane, 0.05 wt% AIBN, 30 min degassed, 5 min sonication, 65 °C overnight.b Hydrodynamic diameter (dh) were determined by DLS measurements of 1 mg mL^−1 NP solutions at RT (Zetasizer SZ from Malvern).c Determined via UV/Vis spectroscopy of 1 mg mL^−1 NP.d Modification degrees were calculated based on the concentration values determined by UV/Vis.
NP4-FA	0	Folic acid	21.38 ± 7.62 (0.47 ± 0.01)	1.74 ± 0.85	7.70 ± 3.77
NP5-FA	50	Folic acid	39.56 ± 5.35 (0.48 ± 0.09)	1.97 ± 1.05	8.70 ± 4.62
NP6-FA	100	Folic acid	60.64 ± 7.85 (0.48 ± 0.17)	2.00 ± 0.64	8.82 ± 2.81
NP4-RGD	0	RGD-peptide	28.86 ± 6.32 (0.42 ± 0.02)	0.11 ± 0.02	20.02 ± 9.36
NP5-RGD	50	RGD-peptide	41.29 ± 2.23 (0.45 ± 0.01)	0.18 ± 0.04	18.20 ± 2.55
NP6-RGD	100	RGD-peptide	68.20 ± 7.66 (0.24 ± 0.01)	0.13 ± 0.05	23.00 ± 0.25
NP4-FITC	0	FITC	24.67 ± 7.62 (0.41 ± 0.03)	1.25 ± 0.03	4.85 ± 0.10
NP5-FITC	50	FITC	38.68 ± 2.48 (0.25 ± 0.01)	1.16 ± 0.12	4.52 ± 0.48
NP6-FITC	100	FITC	59.42 ± 3.32 (0.34 ± 0.02)	1.04 ± 0.22	4.06 ± 0.84

---

In the case of FITC the degree of modification was roughly 50% compared to FA as determined by UV/Vis measurements. Successful nanoparticle formation could also be seen by FTIR and formation of the C=S stretching bond at 1187 cm⁻¹ (see Fig. 5). Moreover, the following size measurements of NP5-FITC and NP6-FITC showed no major increase after modification with the dye. O'Reilly and coworkers synthesized alkyne containing shell-cross-linked polymeric nanoparticle with a size around 40 nm and labeled the surface with an azide-modified FITC via CuAAC. The concentration was determined with c = 2.5 × 10⁻⁶ mol L⁻¹ and the degree of modification was calculated to 25% related to the functional groups on the surface.⁸³ In comparison our modification approach of using an addition reaction between the isothiocyanate and primary amines resulted in a higher FITC amount on the particle surface and no metal catalyst was used. Zeta-potential was measured for the pure nanoparticles NP4–NP6 and after surface modification (see Fig. S9†). Whereas the smaller NP4 indicated a positive zeta-potential of 14.84 ± 0.94 mV, the zeta-potential was reduced to 6.83 ± 0.88 mV for NP5 and and −9.08 ± 0.50 mV for NP6. Accordingly, NP4 displayed also after modification with FITC or the RGD-peptide still a positive zeta-potential of 11.84 ± 0.17 and 8.30 ± 0.48 mV, respectively while folic acid modification changed the zeta-potential to negative values of −12.64 ± 0.79 mV. In comparison the larger particle NP6 (d = 70.25 nm in H₂O) indicated a slightly negative zeta-potential of −9.08 ± 0.50 mV which remained nearly unchanged after surface modification with folic acid, FITC or RGD-peptide in the range of −5.70 to −9.09 nm (see Fig. S9 and Table S2†). A clear trend can be seen between particle size and zeta-potential on the one side and surface modification on the other side.

Conclusion

Here, we described the successful synthesis and characterization of a bifunctional, amphiphilic macromonomer based on poly(2-oxazolines) with multiple acrylate groups in their side chains (P1) and azide end group that can be easily converted to primary amine moieties (P2) via Staudinger reaction. The polymers P1 and P2 were applied as bifunctional surfactants in the synthesis of nanoparticles (NP1–NP6) by using a microemulsion approach. Particle sizes of 20–75 nm were obtained by using different amounts of 1,6-hexanediol dimethacrylate (HDDMA) as a core cross-linker and were characterized by DLS and TEM measurements. Finally, successful surface modification of the nanoparticle series NP4–NP6 derived from the amino-functional polymer P2 was demonstrated with folic acid and a RGD-peptide for potential active targeting application and FITC as a fluorescence probe. The degree of surface modification was determined by UV/Vis spectroscopy depend on the coupling chemistry and steric demand of the reactant and was 1.74–2.0 × 10⁴ mol L⁻¹ for folic acid modification, 0.11–0.18 × 10⁴ mol L⁻¹ for RGD-peptide modification and 1.04–1.25 × 10⁴ mol L⁻¹ for FITC modification.

Experimental

Materials

Fluorescein isothiocyanate (FITC) and triethylamine (TEA) were purchased from Sigma-Aldrich (Steinheim, Germany), folic acid hydrate (FA) was ordered from TCI (Tokyo, Japan), diisopropylethylamine (DIPEA) was supplied from Acros (Nidderau, Germany), 1-ethyl-3-(3′-dimethylaminopropyl)carbodiimide hydrochloride (EDC·HCl) was purchased from ChemPep (Welligton, U.S.), N,N,N′,N′-bis(tetramethylen)fluoroformamidiniumhexafluorophosphate (TFFH) was bought from Carbolution Chemicals (Saarbrücken, Germany) and were used as received unless otherwise stated. The peptide sequence GRGDS6Ahx6AhxF was synthesized by Omar Sallouh. Solvents were dried and purified by using a purification system from M Braun Glovebox Technology PLC 800. Methyl triflate (MeOTf), 2-methyl-2-oxazoline (MOx), 2-heptyl-2-oxazoline (synthesized according to Seelinger et al.)⁸⁴ and acetonitrile (ACN) for polymer preparation were dried by refluxing over CaH₂ under a dry argon atmosphere. The dialysis membrane were composed of regenerated cellulose from ZelluTrans/Roth V-Series with a MWCO = 1000 or 5000.

Measurements

The NMR spectra were recorded on 500 MHz spectrometer AVANCE-III HDX-500 with 5 mm nitrogen cooled Prodigy H(C,N) probe (Bruker BioSpin GmbH) or on a 400 MHz NMR spectrometer Nanobay AVANCE-III HD-400 with 5 mm BBFOsmart probe (Bruker BioSpin GmbH). The spectra were calibrated using the signals of the deuterated solvent CDCl₃ at 7.26 ppm. The FT-IR spectra were recorded on a Bruker Tensor 27 Platinum ATR equipped with OPUS software. For collecting spectra, a total number of 32 scans was used. The size exclusion chromatography (SEC) was performed on a Smartline 2300 (KNAUER) equipped with a refractive index (RI) detector (tempered to 60 °C) using a PSS GRAM analytical column set (1 × pre-column + 1× 1000A + 1× 30A). N,N-Dimethylformamide (HPLC-grade) was used as eluent (+5 g L⁻¹ LiBr) at a flow rate of 1 mL min⁻¹ at 60 °C. SEC columns were calibrated with PMMA standards (from PSS). Prior to each measurement, the polymer samples were purified by using a 0.2 μm Teflon filter (VWR) to remove larger particles. Dynamic light scattering experiments were performed using a Malvern Zetasizer Nano S (ZEN 1600). A 4 mW He–Ne laser (633 nm wavelength) with a fixed detector angle of 173° was used. About 1 mL of dust-free sample was transferred to a special light scattering cell without filtration. The polymer samples were dissolved in methanol or water equilibrated at 25 °C for one minute before the data acquisition started. The measurements were repeated five times with 10 runs. For further interpretation, the peak average of histograms from the number distributions of 50 accumulations was reported as the average diameter of the particles.

Transmission electron microscopy (TEM) images were obtained by using an energy filter transmission electron microscope (Philips CM200) to study the morphology of the prepared particles. The samples with c = 0.05 mg mL⁻¹ methanol were stained with uranylacetate. For calculation of the particle sizes imageJ software was used.

Monomer synthesis

Synthesis of 2-(5-azidopentyl)-2-oxazoline (4) was synthesized according to Lav et al.⁵⁶

6-Azidohexanoic acid (1). 8.84 g 6-bromohexanoic acid (45.3 mmol, 1 eq.) and 14.73 g sodium azide (226.6 mmol, 5 eq.) were dissolved in 100 mL of dry DMSO and stirred for 24 h at 60 °C. The reaction was stopped by adding 100 mL water and the mixture was extracted with 75 mL of dichloromethane. Then, the combined organic layer was washed with 50 mL water twice. The organic phase was dried over magnesium sulfate and the organic solvent was removed. The product was dried under high pressure (6.02 g, yield: 85%).

¹H-NMR (500.13 MHz, CDCl₃): δ (ppm) = 11.60 (s, 1H, COOH), 3.28 (2H, N₃CH₂), 2.37 (2H, CH₂COOH), 1.65 (4H, CH₂CH₂CH₂CH₂CH₂), 1.43 (2H, CH₂CH₂CH₂CH₂CH₂). ¹³C-NMR (100.64 MHz, CDCl₃): δ (ppm) = 24.1 (CCH₂CH₂), 26.1 (CCH₂CH₂CH₂), 28.4 (CH₂CH₂N₃), 33.8 (CCH₂), 51.1 (CH₂N₃), 179.9 (COOH).

N-Succinimidyl-6-azidohexanoate (2). 6.02 g 6-azidohexanoic acid (38.3 mmol, 1 eq.), 7.03 g NHS (61.1 mmol, 1.6 eq.) and 8.78 g EDC·HCl (45.8 mmol, 1.2 eq.) were dissolved in 100 mL of dry dichloromethane. The reaction was stopped after 24 h stirring at room temperature by removing the solvent. The residue was dissolved in a mixture of diethyl ether and water (150 mL/50 mL) and extracted six times with 30 mL of water. The organic layer was dried over magnesium sulfate and the solvent was removed. The product was dried under vacuum (8.59 g, yield: 88%).

¹H-NMR (500.13 MHz, CDCl₃): δ (ppm) = 3.26 (2H, N₃CH₂), 2.79 (4H, C=OCH₂CH₂C=O), 2.59 (2H, CH₂COON), 1.75 (2H, N₃CH₂CH₂), 1.61 (2H, CH₂CH₂COON), 1.48 (2H, CH₂CH₂CH₂CH₂CH₂). ¹³C-NMR (100.64 MHz, CDCl₃): δ (ppm) = 24.08, 25.54, 25.82, 28.33, 30.73, 51.04, 168.36, 169.13.

N-(2-Chloroethyl)-6-azidohexanamide (3). 8.59 g N-succinimidyl-6-azidohexanoate (33.7 mmol, 1 eq.) and 3.99 g 2-chloroethylamine hydrochloride (33.7 mmol, 1 eq.) were dissolved in 100 mL of dry dichloromethane. The mixture was ice-cooled and 11.9 mL triethylamine (84.3 mmol, 2.5 eq.) was added dropwise. After the mixture was stirred for 30 minutes at 0 °C, it was allowed to equilibrate to room temperature and was stirred additional 72 h. The reaction was stopped by adding 50 mL of water. The organic layer was washed two times with 25 mL of water. The organic phase was dried over magnesium sulfate and the solvent was evaporated. The product was dried under vacuum (5.85 g, yield: 80%).

¹H-NMR (500.13 MHz, CDCl₃): δ (ppm) = 6.12 (1H, NH), 3.78 (2H, CH₂Cl), 3.58 (2H, NCH₂), 3.24 (2H, N₃CH₂), 2.19 (2H, CH₂CON), 1.64 (2H, N₃CH₂CH₂), 1.59 (4H, CH₂CH₂COO), 1.41 (2H, N₃CH₂CH₂CH₂). ¹³C-NMR (100.64 MHz, CDCl₃): δ (ppm) = 25.01, 26.24, 28.55, 36.28, 41.10, 44.12, 51.20, 168.27.

2-(5-Azidopentyl)-2-oxazoline (4). 5.85 g N-(2-chloroethyl)-6-azidohexanamide (26.7 mmol) and 1.5 g KOH were dissolved in 7.2 mL of dry methanol. The reaction mixture was stirred at 50 °C for 72 h. The salt was removed by filtration and the organic solvent was removed. Diethylether was added to dissolve the product and to allow filtration of salt. After the solvent was removed, the product was dried under vacuum (4.09 g, yield: 83%).

¹H-NMR (500.13 MHz, CDCl₃): δ (ppm) = 4.20 (2H, OCH₂), 3.80 (2H, NCH₂), 3.26 (2H, N₃CH₂), 2.27 (2H, CH₂CON), 1.64 (4H, CH₂CH₂CH₂CH₂CH₂), 1.42 (2H, N₃CH₂CH₂CH₂). ¹³C-NMR (100.64 MHz, CDCl₃): δ (ppm) = 25.3 (CH₂CH₂C(O)N), 26.1 (CH₂CH₂CH₂C(O)N), 27.6 (CH₂CH₂CH₂N₃), 28.4 (CH₂C(O)N), 51.1 (CH₂N), 54.2 (CH₂CH₂CH₂N₃), 67.0 (NCH₂CH₂O), 168.1 (C(O)N).

FTIR (ATR mode): 2090 cm⁻¹ (N=N=N).

2-(5-Pentyl-[(1,2,3-triazol)-4-yl-methacrylat)]oxazoline (5). In a 1 : 1 solution of THF : H₂O 2 g 2-(5-azidopentyl)-2-oxazoline (10.97 mmol, 1 eq.) was dissolved and 1.36 g propargylmethacrylate (10.97 mmol, 1 eq.) was added. The mixture was stirred for 5 minutes at room temperature. Then 137 mg CuSO₄·5H₂O (0.55 mmol, 0.05 eq.) and 317 mg sodium ascorbate (1.10 mmol, 0.1 eq.) were added and the solution was allowed to stir overnight at room temperature. The reaction was stopped by adding a saturated Na-EDTA-solution and was extracted three times with dichloromethane. The organic layer was dried over MgSO₄ and the solvent was removed. After the product was dried under vacuum, a yellow, viscous oil was obtained (2.55 g, yield: 76%).

¹H-NMR (500.13 MHz, CDCl₃): δ (ppm) = 1.34 (t, 2H, CH₂CH₂CH₂CH₂CH₂N), 1.65 (t, 2H, CH₂CH₂CH₂CH₂CH₂N), 1.91 (s, 5H, CH₂CH₂CH₂CH₂CH₂N, CH₃), 2.24 (t, 2H, CH₂CH₂CH₂CH₂CH₂N), 3.79 (t, 2H, NCH₂CH₂O), 4.18 (t, 2H, NCH₂CH₂O), 4.33 (t, 2H, CH₂CH₂CH₂CH₂CH₂N), 5.26 (s, 2H, CH₂OCO), 5.55/6.10 (s, 2H, CH₃C=CH₂), 7.59 (s, 1H, triazol-CH). ¹³C-NMR (100.64 MHz, CDCl₃): δ (ppm) = 18.14 (CH₃), 25.08 (CH₂CH₂CH₂CH₂CH₂N), 25.83 (CH₂CH₂CH₂CH₂CH₂N), 27.44 (CH₂CH₂CH₂CH₂CH₂N), 29.77 (CH₂CH₂CH₂CH₂CH₂N), 50.00 (CH₂CH₂CH₂CH₂CH₂N), 54.20 (NCH₂CH₂O), 57.79 (CH₂OCO), 67.09 (NCH₂CH₂O), 123.61 (CH₃C=CH₂), 126.13 (NNNC=CH), 135.79 (CH₃C=CH₂), 142.78 (NNNC=CH), 167.14 (CH₂OCO), 167.92 (NCO).

MS (ESI) m/z calcd for C₁₅H₂₂N₄O₃ ([H]⁺), 307.1770; found, 307.1772.

Polymerization

Poly[{(2-heptyl-2-oxazoline)₅-co-(2-(5-pentyl-[(1,2,3-triazol)-4-yl-meth-acrylat)]oxazoline)₃}_stat-block-(2-methyl-2-oxazoline)₃₀]-N₃ (P1). The polymerization and workup procedures were carried out following a general procedure. In a Schlenk tube, 198 μL 2-heptyl-2-oxazoline (HOx, 4 eq.), 357 mg 2-(5-pentyl-[(1,2,3-triazol)-4-yl-methacrylat)]oxazoline (AOx, 4 eq.), 33 μL methyl triflate (1 eq.) and 5 mL dry acetonitrile were mixed under inert conditions (argon). The reaction mixture was stirred at 110 °C for 4 h. Subsequently, 875 μL 2-methyl-2-oxazoline (MOx, 30 eq.) was added and polymerized at 120 °C for 2 h. At room temperature, 98 mg sodium azide as a terminating agent was added and the reaction mixture was stirred over night at room temperature. The solid residue was filtered off and the solvent was removed at reduced pressure. Then water was added and the crude product was dialyzed (MWCO = 1000) and afterwards lyophilized. Further the polymer was purified by precipitation in cold diethyl ether. The precipitated polymer was removed by centrifugation and dried under high pressure (750 mg, yield: 52%).

¹H-NMR (400.25 MHz, CDCl₃): δ (ppm) = 0.85 (s, 15H, CH_3,HOx), 1.25 (brs, 46H, 4 × CH_2,HOx, CH_2,AOx), 1.57 (brs, 18H, CH_2,HOx, CH_2,AOx), 1.92 (s, 17H, CH_3,AOx, CH_2,AOx), 2.06–2.13 (m, 98H, CH_3,MOx), 2.19–2.35 (m, 19H), 3.00/2.93 (m, 3H, CH_3,I), 3.44 (m, 152H, CH₂–CH_2,backbone), 4.34 (brs, 6H, CH_2,AOx), 5.26 (s, 6H, OCH_2,AOx), 5.57/6.11 (s, 6H, C=CH₂), 7.65 (s, 3H, C=CHN). SEC: PDI = 1.13, M_n = 8478 g mol⁻¹. FTIR (ATR mode): N=N=N 2112 cm⁻¹.

Poly[{(2-heptyl-2-oxazoline)₅-co-(2-(5-pentyl-[(1,2,3-triazol)-4-yl-meth-acrylat)]oxazoline)₃}_stat-block-(2-methyl-2-oxazoline)₃₀]-NH₂ (P2). 400 mg of P1 (0.1 mmol, 1 eq.) was dissolved in dry DCM. Then 75 mg triphenylphosphine (0.3 mmol, 3 eq.) was added. The reaction mixture was allowed to stir at room temperature for 18 h. It was stopped by adding 8 mL water and stirring additionally for 30 min. Afterwards the organic solvent was removed and the aqueous phase was filtered and afterwards lyophilized. The white solid obtained after lyophilization was dissolved in 5 mL acetonitrile and filtered (0.2 μm, PP) two times. For further purification the polymer was precipitated in cold diethyl ether, centrifuged and dried under vacuum (270 mg, yield: 67%).

¹H-NMR (500.08 MHz, CDCl₃): δ (ppm) = 0.86 (s, 15H, CH_3,HOx), 1.27 (brs, 44H, 4 × CH_2,HOx, CH_2,AOx), 1.58 (brs, 17H, CH_2,HOx, CH_2,AOx), 1.93 (s, 19H, CH_3,AOx, CH_2,AOx), 2.06–2.13 (m, 106H, CH_3,MOx), 2.19–2.35 (m, 21H), 3.00/2.93 (m, 3H, CH_3,I), 3.44 (m, 159H, CH₂–CH_2,backbone), 4.35 (brs, 6H, CH_2,AOx), 5.27 (s, 5H, OCH_2,AOx), 5.58/6.12 (s, 6H, C=CH₂), 7.67 (s, 3H, C=CHN). SEC: PDI = 1.25, M_n = 7975 g mol⁻¹.

Microemulsion-polymerization

A 1 mM polymer solution in water was prepared, then different amounts of 1,6-hexanedioldimethacrylate, AIBN and heptadecane relative to the amount of the polymer were added. In Table 2 the amounts were given in detail. After 30 minutes degassing with argon, the mixture was sonicated for 5 minutes. The temperature was then increased to 65 °C and polymerization proceeded overnight. After the polymerization mixture was cooled down, the milky solution was centrifuged for 30 minutes at 4400 rpm (2×). The aqueous layer was removed and lyophilized. The white solid was dissolved in chloroform, precipitated in cold diethyl ether and dried under vacuum.

Folic acid functionalization (NP4-NP6-FA)

2 mg NP (4.59 × 10⁻⁷ mol, 1 eq.) was dissolved in 5 mL dry DCM. Then 81 μL of 5 mg mL⁻¹ folic acid in DCM (9.18 × 10⁻⁷ mol, 2 eq.) and 36 μL of 5 mg mL⁻¹ EDC·HCl in DCM (9.18 × 10⁻⁷ mol, 2 eq.) were added and the reaction mixture was cooled down to −78 °C. After adding of 1.6 μL of 100 μL mL⁻¹ TEA in DCM at this temperature, the reaction was stirred for 24 h. Afterwards the solution was filtered and the solvent was removed at reduced pressure. To remove the excess of folic acid, the residue was dissolved in water and dialyzed against a 0.1 M NaHCO₃ solution for 24 h (MWCO 1000). After lyophilization, the solid was dissolved in chloroform, precipitated in cold diethyl ether, centrifuged and dried under high pressure. The amount of coupled folic acid was calculated via UV/Vis spectroscopy at λ_max = 363 nm with ε_{363 nm} = 1800 L mol⁻¹ cm⁻¹ of 1 mg mL⁻¹ NP in 0.1 M NaOH solution.

Functionalization with GRGDS6Ahx6AhxF-peptide-(NP4-NP6-P)

2 mg NP (4.59 × 10⁻⁷ mol, 1 eq.) and 0.16 μL DIPEA (9.18 × 10⁻⁷ mol, 2 eq.) was dissolved in 5 mL dry DMF. A mixture of 1.3 mg GRGDS6Ahx6AhxF (9.18 × 10⁻⁷ mol, 2 eq.), 0.29 mg TFFH (9.18 × 10⁻⁷ mol, 2 eq.) and 0.16 μL DIPEA in DMF were added to the NP solution. The reaction was stirred for 24 h at room temperature. The organic solvent was removed and the residual solid dialyzed against water for 48 h (MWCO 5000). After lyophilization the functionalized NPs were dissolved in chloroform and precipitated in cold diethyl ether and dried at high pressure. The amount of coupled RGD-peptide was calculated via UV-Vis spectroscopy at λ_max = 254 nm with ε_{254 nm} = 3300 L mol⁻¹ cm⁻¹ of 1 mg mL⁻¹ NP in an aqueous solution.

FITC-functionalization (NP4-NP6-FITC)

2 mg NP (4.59 × 10⁻⁷ mol, 1 eq.) was dissolved in 5 mL ethanol. Then 233 μL of 1 mg mL⁻¹ FITC in ethanol (5.97 × 10⁻⁷ mol, 1.3 eq.) was added and the reaction mixture was stirred in the dark at room temperature. After 24 h the reaction was stopped, the solvent was removed at reduced pressure. To remove the excess of FITC, the residue was dissolved in chloroform and precipitated in cold diethyl ether. This procedure was repeated until no FITC was observed in the diethyl ether phase. The functionalized NPs were dried under high vacuum. The amount of coupled FITC was calculated via UV-Vis spectroscopy at λ_max = 495 nm with ε_{495 nm} = 3400 L mol⁻¹ cm⁻¹ of 1 mg mL⁻¹ NP in 0.1 M NaHCO₃ solution.

Acknowledgements

The authors gratefully acknowledge Frau Meuris from the group of Prof. Dr J. Tiller (BCI) for preparing the TEM images. We thank also the group of Prof. H. Rehage (CCB/Physikalische Chemie) and coworkers for providing the DLS device.

References

1. F. Masood, Mat. Sci. Eng., C, 2016, 60, 569–578.
2. M. Elsabahy and K. L. Wooley, Chem. Soc. Rev., 2012, 41, 2545–2561.
3. Y. Jiang, et al., Biomaterials, 2014, 35, 4969.
4. D.-E. Lee, et al., Chem. Soc. Rev., 2012, 41, 2656–2672.
5. R. Toy, et al., Adv. Drug Delivery Rev., 2014, 76, 79–97.
6. J. P. Raoa, et al., Prog. Polym. Sci., 2011, 36, 887–913.
7. D. Lemoine, et al., J. Controlled Release, 1998, 54, 15–27.
8. R. Gurny, et al., Drug Dev. Ind. Pharm., 1981, 7, 1–25.
9. R. K. O'Reilly, et al., Chem. Soc. Rev., 2006, 35, 1068–1083.
10. B. Thurmond, et al., J. Am. Chem. Soc., 1997, 119, 6656–6665.
11. V. Butun, et al., J. Am. Chem. Soc., 1998, 120, 12135–12136.
12. A. Munoz-Bonilla, et al., Macromolecules, 2010, 43, 2721–2731.
13. M. Antonietti and K. Landfester, Prog. Polym. Sci., 2002, 27, 689–757.
14. F. Yan and J. Texter, Soft Matter, 2006, 2, 109–118.
15. M. Antonietti, et al., Macromolecules, 1991, 24, 6636–6643.
16. J. M. Asua, Prog. Polym. Sci., 2002, 27, 1283–1346.
17. K. Landfester and V. Mailaender, Expert Opin. Drug Delivery, 2013, 10, 593–609.
18. R. Munoz-Espi, et al., Curr. Opin. Colloid Interface Sci., 2012, 17, 212–224.
19. M. Wua, et al., Colloids Surf., B, 2009, 69, 147–151.
20. W. Li, et al., Macromolecules, 2008, 41, 6387–6392.
21. W. Li, et al., J. Am. Chem. Soc., 2010, 132, 7823–7824.
22. C. Cheng, et al., Prog. Chem., 2011, 23, 791–799.
23. Y. Li, et al., Macromolecules, 2012, 45, 4623–4629.
24. A. V. Fuchs and K. J. Thurecht, Macromol. Chem. Phys., 2015, 216, 1271–1281.
25. F. Candau, et al., J. Polym. Sci., Polym. Chem. Ed., 1985, 23, 193–214.
26. M. Antonietti, et al., Macromolecules, 1991, 24, 6636–6643.
27. C. Wu, Macromolecules, 1994, 27, 298–299.
28. F. Bléger, et al., Macromolecules, 1994, 27, 2559–2565.
29. M. Antonietti, Macromol. Chem. Phys., 1995, 196, 441–466.
30. S. Tomoeda, et al., Macromolecules, 2011, 44, 5599–5604.
31. R. G. Graff, et al., Macromolecules, 2015, 48, 2118–2126.
32. S. Liu, K. D. Hermanson and E. W. Kaler, Macromolecules, 2006, 39, 4345–4350.
33. J. M. O'Donnell, Chem. Soc. Rev., 2012, 41, 3061–3076.
34. K. Margulis-Goshen and S. Magdassi, Curr. Opin. Colloid Interface Sci., 2012, 17, 290–296.
35. A. V. Heydenreich, et al., Int. J. Pharm., 2003, 254, 83–87.
36. F. Yan and J. Texter, Soft Matter, 2006, 2, 109–118.
37. G. Riess, J. Nervo and D. Rogez, Polym. Eng. Sci., 1977, 17, 634–638.
38. G. Riess and J. Nervo, Inf. Chim., 1977, 170, 185.
39. T. Liu, H. Schuch, M. Gerst and B. Chu, Macromolecules, 1999, 32, 6031–6042.
40. G. David, et al., Eur. Polym. J., 2002, 38, 73–78.
41. B. Babaç, et al., Eur. Polym. J., 2004, 40, 1947–1952.
42. K. Aoi and M. Okada, Prog. Polym. Sci., 1996, 21, 151–208.
43. S. Kobayashi, Prog. Polym. Sci., 1990, 15, 751–823.
44. M. T. Zarka, et al., Chem.–Eur. J., 2003, 9, 3228–3234.
45. S. Cesana, et al., Macromol. Chem. Phys., 2006, 207, 183–192.
46. W. Hiller, et al., Macromolecules, 2015, 48, 4032–4045.
47. M. Bortenschlager, et al., Chem.–Eur. J., 2007, 13, 520–528.
48. R. Luxenhofer, et al., Macromol. Rapid Commun., 2012, 33, 1613–1631.
49. J. Ulbricht, et al., Biomaterials, 2014, 35, 4848–4861.
50. A. L. Kampmann, et al., Macromol. Chem. Phys., 2016, 217, 1704–1711.
51. G. Bissadi and R. Weberskirch, Polym. Chem., 2016, 7, 5157–5168.
52. S. Kobayashi, et al., Makromol. Chem., Rapid Commun., 1992, 13, 337–341.
53. S. Kobayashi, et al., J. Polym. Sci., Part A: Polym. Chem., 1993, 31, 3133–3139.
54. H. Uyama and S. Kobayashi, Polym. Int., 1994, 34, 339–344.
55. V. V. Jerca, et al., Polym. Bull., 2011, 66, 785–796.
56. T. X. Lav, et al., React. Funct. Polym., 2013, 73, 1001–1008.
57. K. Kempe, et al., Macromolecules, 2011, 44, 6424–6432.
58. H. Staudinger and J. Meyer, Helv. Chim. Acta, 1919, 2, 635–646.
59. W. Qian, et al., Polym. Chem., 2016, 7, 3300–3310.
60. B. Bahrami, et al., Tumor Biol., 2015, 36, 5727–5742.
61. K. Miyata, et al., React. Funct. Polym., 2011, 71, 227–234.
62. S. Khoee and A. Kavand, Eur. J. Med. Chem., 2014, 73, 18–29.
63. W. Scarano, et al., Biomacromolecules, 2013, 14, 962–975.
64. N. K. Nukolova, et al., Biomaterials, 2011, 32, 5417–5426.
65. L. Zhang, et al., Int. J. Pharm., 2004, 287, 155–162.
66. M. M. Jia, et al., ACS Appl. Mater. Interfaces, 2014, 6, 11413–11423.
67. D. Arosio and C. Casagrande, Adv. Drug Delivery Rev., 2016, 97, 111–143.
68. D. Cox, et al., Nat. Rev. Drug Discovery, 2010, 9, 804–820.
69. D. Kaufmann, et al., Macromol. Biosci., 2008, 8, 577–588.
70. X. B. Xiong, et al., Biomacromolecules, 2007, 8, 874–884.
71. J. A. Zupancich, et al., Biomacromolecules, 2009, 10, 1554–1563.
72. C. Y. Zhan, et al., J. Controlled Release, 2010, 143, 136–142.
73. H. T. T. Duong, et al., Polym. Chem., 2010, 1, 171–182.
74. M. Oba, et al., Mol. Pharm., 2008, 5, 1080–1092.
75. M. Reuter, et al., Biophys. Chem., 2009, 144, 27–37.
76. E. Fuentes-Paniagua, et al., Eur. Polym. J., 2015, 71, 61–72.
77. F. Oltolina, et al., Langmuir, 2015, 31, 1766–1775.
78. N. Engelhardt, PhD thesis, Technische Universität Dortmund, Dortmund, 2014.
79. R. Rossin, et al., J. Nucl. Med., 2005, 46, 1210–1218.
80. D. Pan, et al., Chem. Commun., 2003, 2400–2401.
81. W. Yao, et al., J. Colloid Interface Sci., 2015, 458, 14–21.
82. S. L. Mekuria, et al., J. Phys. Chem. B, 2016, 120, 123–130.
83. R. K. O'Reilly, et al., Chem. Mater., 2005, 17, 5976–5988.
84. H. Witte and W. Seelinger, Liebigs Ann. Chem., 1974, 996–1009.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra22896h

---