SEARCH

SEARCH BY CITATION

Keywords:

  • living radical polymerization (LRP);
  • lower critical solution temperature (LCST);
  • reversible addition fragmentation chain transfer (RAFT);
  • thiol-ene;
  • thiol-yne;
  • water-soluble polymers

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY AND CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES AND NOTES

Sequential thiol-ene/thiol-ene and thiol-ene/thiol-yne reactions have been used as a facile and quantitative method for modifying end-groups on an N-isopropylacrylamide (NIPAm) homopolymer. A well-defined precursor of polyNIPAm (PNIPAm) was prepared via reversible addition-fragmentation chain transfer (RAFT) polymerization in DMF at 70 °C using the 1-cyano-1-methylethyl dithiobenzoate/2,2′-azobis(2-methylpropionitrile) chain transfer agent/initiator combination yielding a homopolymer with an absolute molecular weight of 5880 and polydispersity index of 1.18. The dithiobenzoate end-groups were modified in a one-pot process via primary amine cleavage followed by phosphine-mediated nucleophilic thiol-ene click reactions with either allyl methacrylate or propargyl acrylate yielding ene and yne terminal PNIPAm homopolymers quantitatively. The ene and yne groups were then modified, quantitatively as determined by 1H NMR spectroscopy, via radical thiol-ene and radical thiol-yne reactions with three representative commercially available thiols yielding the mono and bis end functional NIPAm homopolymers. This is the first time such sequential thiol-ene/thiol-ene and thiol-ene/thiol-yne reactions have been used in polymer synthesis/end-group modification. The lower critical solution temperatures (LCST) were then determined for all PNIPAm homopolymers using a combination of optical measurements and dynamic light scattering. It is shown that the LCST varies depending on the chemical nature of the end-groups with measured values lying in the range 26–35 °C. © 2009 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 47: 3544–3557, 2009


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY AND CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES AND NOTES

The thiol-ene reaction1 (the hydrothiolation of a C[DOUBLE BOND]C bond) has recently attracted significant attention in the materials arena2 because it displays many of the attributes of click chemistry.3 Such additions can be accomplished under a variety of experimental conditions including acid/base catalysis,4 nucleophile-mediated,5, 6 radical-mediated (often induced photochemically),2, 7–12 and via a solvent promoted process.13, 14 However, the reaction is most commonly performed under radical conditions where it is applicable to many ene substrates, or under nucleophilic conditions with activated enes where the process proceeds via an anionic chain process, Scheme 1. A real synthetic strength of the thiol-ene reaction is its versatility with respect to ene substrates that can be used and include both activated and nonactivated species such as norbornenes, (meth)acrylates, maleimides, and allyl ethers to name but a few.2 Recent examples describing the application of the thiol-ene reaction in the polymer/materials field include the work of Killops et al.15 who described its use in dendrimer synthesis, Chan and coworkers5 highlighted its application in convergent star synthesis, Gress et al.16 used it to modify side chains in poly[2-(3-butenyl)-2-oxazoline], Rissing and Son17 detailed the synthesis of multifunctional branched organosilanes, Campos et al.18 described its use in the synthesis of materials for soft imprint lithography, Pounder and coworkers19 highlighted its application in the functionalization of poly(ester)s, Li et al.20 described the synthesis of functional telechelics and a modular approach to block copolymers, and Chen et al. have described the synthesis and temperature-induced self-assembly of neoglycopolymers.21

thumbnail image

Scheme 1. (A) Mechanism for the radical mediated thiol-ene reaction with a terminal ene, and (B) anionic chain mechanism for the amine/phosphine mediated nucleophilic thiol-ene reaction with an acrylate.

Download figure to PowerPoint

The radical-mediated addition of a thiol to an yne22, 23 has been widely studied. Importantly, it should be highlighted that the radical thiol-yne reaction can be considered as a sister reaction to the radical thiol-ene reaction, and possesses many similar characteristics both kinetically and mechanistically. As with the thiol-ene reaction, the thiol-yne reaction, in general, proceeds rapidly under a variety of experimental conditions selectively yielding the mono or bisaddition products.24–29 In the case of the double addition products formed under radical conditions, the reaction of two equivalents of thiol with a terminal alkyne is itself a two-step process, Scheme 2. The first, slower, step (1) involves the regiospecific anti-Markovnikov-like addition of a thiyl radical across the C[TRIPLE BOND]C bond yielding the intermediate vinylic radical as a mixture of the cis and trans diastereoiosomers. This vinylic radical subsequently undergoes chain transfer to additional thiol, yielding the cis/trans vinylthioether concomitantly generating a new thiyl radical. The vinylthioether product is able to undergo a second, faster, regiospecific (2), formally thiol-ene, reaction with the newly generated thiyl radical yielding the intermediate bisthioether radical that undergoes chain transfer with another thiol molecule quantitatively yielding the target bisthioether with exclusive 1,2-orientation.30 However, one important difference between the radical thiol-ene and radical thiol-yne reactions, at least in the case of the double addition products, is that the thiol-yne reaction of a terminal acetylene results in the generation of a stereocenter and thus the final 1,2-addition product is a mixture of (R) and (S) enantiomers, and as such the radical thiol-yne reaction is not stereoselective.

thumbnail image

Scheme 2. Proposed mechanism for the radical mediated thiol-yne reaction with a terminal alkyne.

Download figure to PowerPoint

However, like the radical-mediated thiol-ene reaction, the radical-initiated thiol-yne reaction can proceed extremely rapidly yielding products quantitatively under facile conditions and with little-to-no clean up required, that is, reactions can be easily performed in an air atmosphere at ambient temperature and humidity. Although also known since the first half of the last century, the thiol-yne reaction has been essentially overlooked in the polymer/materials field. However, researchers have recently begun to evaluate this potentially highly versatile reaction. Fairbanks et al.30 described radical-mediated thiol-yne step-growth polymerization as a route to highly crosslinked networks and conducted a detailed kinetic analysis of the process simultaneously noting the similarities with the radical thiol-ene process. Chan et al.31 reported the synthesis and physical properties of high refractive index polysulfide networks formed via the photopolymerization of a range of alkyldithiols with a series of alkyldiynes, and most recently, Chan, Hoyle, and Lowe6 described the first example of a sequential, quantitative nucleophilic thiol-ene/radical thiol-yne process for the synthesis of a range of branched structures including examples with potential biomedical significance.

Reversible addition-fragmentation chain transfer (RAFT) radical polymerization32–35 is a powerful synthetic tool mediated by thiocarbonylthio compounds that is simple to perform, highly functional group tolerant, applicable to a wide range of monomeric substrates and experimental conditions, facilitates the synthesis of (co)polymers with narrow molecular weight distributions, predetermined molecular weights, and advanced architectures.36–46 As a consequence of the degenerative chain transfer mechanism, (co)polymers prepared by RAFT bear very specific end-groups, the chemical nature of which is dependent on the structure of the chain transfer agent (CTA) and, to a lesser extent, the CTA/initiator pair. Assuming no undesirable side reactions then the α terminus of a (co)polymer chain will be chemically identical to the R-group of the RAFT CTA while the ω terminus bears a thiocarbonylthio functional group, the exact chemical structure of which will be dependent of the class of RAFT CTA used, that is, dithioester versus xanthate versus trithiocarbonate etc. As such, RAFT offers an attractive, and convenient, route to telechelic (co)polymers simply via the use of functional RAFT agents.47 In addition, the thiocarbonylthio end-groups are easily modified postpolymerization to a variety of species including cleavage to a thiol, typically with a primary or secondary amine,5, 48, 49 cleavage to a thiolate,50, 51 thermolysis to an ene,52, 53 or radical-induced reduction to a saturated species with, for example, Bu3SnH.54 Importantly, the presence of the reactive thiol, or thiolate, at the ω chain terminus allows for further modification reactions to be performed via the versatile thiol-ene chemistry.

Described herein is a facile sequential route to either mono or bis end-functionalized poly(N-isopropylacrylamide) (PNIPAm) via tandem nucleophilic thiol-ene/radical thiol-ene and nucleophilic thiol-ene/radical thiol-yne reactions. These reaction sequences take advantage of the click characteristics of the thiol-ene reaction and the similar, rapid and quantitative radical-mediated reaction of thiols with alkynes. This represents the first time such sequential thiol-ene/thiol-ene and thiol-ene/thiol-yne reactions proceeding by inherently different mechanistic pathways have been used in polymer synthesis/modification and represents a new highly efficient methodology for the site specific modification of RAFT-prepared (co)polymers.

EXPERIMENTAL

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY AND CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES AND NOTES

All reagents were purchased from the Aldrich Chemical Company Inc. at the highest available purity and used as received unless stated otherwise. 1-Cyano-1-methylethyl dithiobenzoate (CPDB) was prepared according to a literature procedure.32 2,2′-Azobis(2-methylpropionitrile) (AIBN) was recrystallized from methanol, and stored in a freezer at −20 °C until needed. N-Isopropylacrylamide (NIPAm) was purified by recrystallization from hexane and stored in a freezer at −20 °C until needed.

Homopolymerization of N-Isopropylacrylamide

A mixture of NIPAm (15.0 g, 133 mmol), CPDB (388 mg, 1.75 mmol), AIBN (58 mg, 0.353 mmol), and DMF (10.0 g) were added to a Schlenk flask equipped with a magnetic stirring bar. The mixture was stirred for at least 30 min while submerged in an ice bath to ensure complete dissolution of CPDB and AIBN. The flask was then purged by repeatedly evacuating and refilling with N2 at least three times before immersion in a preheated oil bath at 70 °C. After 48 h, the reaction was quenched by rapid cooling in liquid N2 and exposure to air. The homopolymer was isolated by precipitation into a large excess of hexanes cooled in an ice-bath followed by dialysis against methanol for 8 h with the solvent changed every 2 h. The purified polymer was then dried in a vacuum oven at room temperature overnight.

Synthesis of Propargyl End Functionalized Poly(N-isopropylacrylamide) (PNIPAm-S-PROPA)

A mixture of polyNIPAm (PNIPAm) (2.5 g, 0.424 mmol), propargyl acrylate (551 mg, 5.0 mmol), dimethylphenylphosphine (10 mg, 0.072 mmol), and CH2Cl2 (7.5 g) were added to a 50-mL round-bottomed flask equipped with a magnetic stirring bar. After purging the mixture with N2 for 15 min with moderate stirring, octylamine (969 mg, 7.5 mmol) was added via a 1-mL syringe. The mixture was purged for another 15 min, and then left to stir overnight at room temperature. The product was isolated by precipitation into a large excess of hexanes cooled in an ice-bath followed by dialysis against methanol for 8 h with the solvent changed every 2 h. The purified polymer was then dried in a vacuum oven at room temperature overnight.

Synthesis of Allyl End Functionalized Poly(N-isopropylacrylamide) (PNIPAm-S-ALMA)

The synthesis of PNIPAm-S-ALMA was achieved following exactly the same protocol as described above for the preparation of PNIPAm-S-PROPA except allyl methacrylate was used in place of propargyl acrylate.

General Procedure for the Photochemical, Radical Thiol-Ene and Thiol-Yne Reactions

To a 20-mL scintillation vial, either PNIPAm-S-ALMA or PNIPAm-S-PROPA (0.2 g), 100% excess of target thiol (based on the molarity of “ene” or “yne”), benzil dimethyl ketal (Irgacure 651) photoinitiator (15.0 mg), and the appropriate volume of THF were added to give a total weight of 1.5 g. The vial was gently swirled to ensure complete dissolution of added reagents before being placed in a Rayonet UV reactor under irritation of UV light (λ = 350 nm). The reaction was allowed to proceed for 2 h before the removal from the UV reactor. The sample was immediately dialyzed against MeOH for a period of 8 h with the MeOH being changed every 2 h. The product was isolated by drying overnight in vacuo at room temperature.

General Instrumentation

1H NMR spectra were recorded on a Bruker 300 53 mm spectrometer in appropriate deuterated solvents. Size exclusion chromatographic analysis was performed on a Waters system comprised of a Waters 515 HPLC pump, Waters 2487 Dual λ absorbance detector, and Waters 2410 RI detector equipped with a PolymerLabs PLgel 5 μm guard column and a PolymerLabs PLgel 5 μm MIXED-C column (molecular weight range: 200–2,000,000 g/mol), in THF stabilized with 281 ppm BHT at a flow rate of 1.0 mL/min. The column was calibrated with a series of narrow molar mass distribution poly(methyl methacrylate) standards. The digital thermometer used in solution studies was a Traceable® expanded range thermometer. pH measurements was performed on a Accument® AR15 pH meter calibrated with three pH buffers (pH = 4.00, 7.00, and 10.00). Photochemical reactions were performed on a Rayonet RPR-100 photochemical reactor with a 350 nm light source. Dynamic light scattering (DLS) experiments were conducted on a Malvern Instruments Zetasizer Nano-ZS (red badge) instrument operating with a 633 nm laser. Data was collected and processed with the Dispersion Technology software V5.10.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY AND CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES AND NOTES

The reversible addition-fragmentation chain transfer (RAFT) radical polymerization of N-isopropylacrylamide (NIPAm) was accomplished with 1-cyano-1-methylethyl dithiobenzoate (CPDB) as the RAFT chain transfer agent in conjunction with AIBN in DMF at 70 °C, Scheme 3, with a targeted Mn of 8800 at 100% conversion.

thumbnail image

Scheme 3. Reversible addition-fragmentation chain transfer polymerization of N-isopropylacrylamide and subsequent end-group modifications, via a combination of nucleophilic thiol-ene/radical thiol-ene and nucleophilic thiol-ene/radical thiol-yne pathways.

Download figure to PowerPoint

The homopolymerization was intentionally terminated early in an effort to preserve thiocarbonylthio end-groups and minimize the presence of terminated products derived from coupling reactions. The resulting homopolymer was characterized using a combination of size exclusion chromatography (SEC) and 1H NMR spectroscopy. SEC indicated an Mn of 3100 with a corresponding polydispersity index of 1.18 [Fig. 1(A)]. The low targeted molecular weight of the NIPAm homopolymer allowed for a more accurate determination of Mn via 1H NMR spectroscopic end-group analysis, which indicated a degree of polymerization of 50 (PNIPAm50), equivalent to an absolute Mn of 5880 assuming quantitative α,ω-functionalization of the PNIPAm50 with cyanoisopropyl and dithiobenzoate functional groups, respectively, Figure 1(B).

thumbnail image

Figure 1. (A) Size exclusion chromatographic (SEC) trace (RI signal) of the parent poly(N-isopropylacrylamide) homopolymer and (B) 1H NMR spectrum, recorded in CDCl3, of the parent poly(N-isopropylacrylamide) homopolymer highlighting the presence of the phenyl end-group and the calculation of the absolute degree of polymerization. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Download figure to PowerPoint

With the well-defined PNIPAm50 in-hand the thiocarbonylthio end-groups were modified to allylic and propargylic species. The transformations were accomplished via tandem aminolysis/thiol-Michael (thiol-ene) reactions with allyl methacrylate (ALMA) and propargyl acrylate (PROPA), respectively. This one-pot conversion was achieved using a combination of octylamine (OctAM) and dimethylphenylphosphine (Me2PPh). OctAM was specifically used to cleave the thiocarbonylthio end-groups yielding PNIPAm50-SH, while Me2PPh served two important roles. Firstly, it prevented the formation of macromolecular disulfide formed by the aerial oxidation of PNIPAm50-SH, and secondly, it served as an extremely potent catalyst for the subsequent nucleophilic thiol-ene reaction.5, 6 Indeed, the reducing capabilities of phosphines as well as their ability to serve as initiators/catalysts for Michael conjugate addition reactions is well documented.55–57 The products from the thiol-ene reaction, PNIPAm50-S-ALMA and PNIPAm50-S-PROPA, were characterized using a combination of 1H NMR spectroscopy and SEC, Figure 2 and Table 1. Figure 2(A) shows the 1H NMR spectrum of PNIPAm50-S-PROPA recorded in CDCl3 with the corresponding SEC trace shown inset. Several key features are evident. Firstly, when comparing the NMR spectrum in 2A with 1B, it is clear that the resonances of the phenyl hydrogens associated with the phenyl ring of the dithiobenzoate end-group of PNIPAm50 are completely absent indicating quantitative cleavage. Additionally, a new resonance, labeled a, is observed in 2A and is assigned to the propargylic methylene hydrogens of the end group in PNIPAm50-S-PROPA. Interestingly, a can be used to calculate the average degree of polymerization (DP), simultaneously verifying quantitative reaction. A ratio of a with b, the resonance associated with the methine group of the isopropyl side chains, indicates a DP of 52. This agrees almost exactly with the previously calculated value of 50 determined using the phenyl end-group as noted above, and is within the error associated with NMR spectroscopy. This excellent agreement not only confirms the high efficacy of the macromolecular, Me2PPh-mediated nucleophilic thiol-ene reaction but also serves to verify the beneficial effect of the Me2PPh in preventing disulfide formation because any significant oxidative coupling of the macromolecular thiols would result in a calculated DP that differs significantly from that calculated before end-group cleavage. The absence of any polymeric disulfide was also confirmed by SEC, see inset in Figure 2(A). The SEC trace is unimodal, symmetric, and narrow with no evidence of any coupled products that are most commonly observed as a shoulder at lower retention times on the main distribution. Indeed, the measured polydispersity index for PNIPAm50-S-PROPA is identical to the precursor PNIPAm50 with Mw/Mn = 1.18.

thumbnail image

Figure 2. (A) 1H NMR spectrum, recorded in CDCl3, of the propargyl-end functionalized PNIPAm50 verifying quantitative end-group modification with the SEC trace shown inset, and (B) 1H NMR spectrum, recorded in CDCl3, of the allyl-end functionalized PNIPAm50 verifying quantitative end-group modification with the resulting SEC trace shown inset. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Download figure to PowerPoint

Table 1. Summary of the Number Average Molecular Weights Determined by Size Exclusion Chromatography and NMR Spectroscopy, the Polydispersity Indices, Average Calculated Degrees of Polymerization, and Lower Critical Solution Temperatures as Determined Optically and by Dynamic Light Scattering for the Precursor PNIPAm50 Homopolymer and the Series of End-Modified PNIPAm Homopolymers
PolymerMn SECaMn NMRbMw/MnaCalculated DPbLCSTc (°C)LCSTd (°C)
  • a

    Determined by size exclusion chromatography with a MIXED-C column (molecular weight range: 200–2,000,000 g/mol), in THF stabilized with 281 ppm BHT at a flow rate of 1.0 mL/min. The column was calibrated with a series of narrow molecular weight distribution poly(methyl methacrylate) standards.

  • b

    Determined by 1H NMR spectroscopy.

  • c

    Determined visually with a temperature probe and gentle heating with a heat gun.

  • d

    Determined by dynamic light scattering with 1 wt % solutions and heating from 5–40 °C. Temperature was increased in 1 °C increments and allowed to equilibrate for 2 min. before a measurement being made.

  • e

    Insoluble, even at temperatures approaching 0 °C.

PNIPAm310058801.185029.029.0
PNIPAm-S-ALMA33001.225130.029.0
PNIPAm-S-PROPA32001.185229.034.0
PNIPAm-S-ALMA-(S-HexOH)30001.175333.026.0
PNIPAm-S-PROPA-(S-HexOH)228001.145632.0
PNIPAm-S-ALMA-(S-Hex)28001.1827.026.0
PNIPAm-S-PROPA-(S-Hex)227001.1727.026.0
PNIPAm-S-ALMA-(S-POSS)37001.224735.035.0
PNIPAm-S-PROPA-(S-POSS)238001.2358ee

Similar observations were made in the case of the ALMA-modified PNIPAm50, likewise prepared via an OctAM/Me2PPh-mediated nucleophilic thiol-ene reaction, Figure 2(B). Resonances associated with the allylic end-group are clearly present in the region δ ∼ 4.5–5.3 ppm and are labeled a and c. A simple ratio of a with b indicates a DP of 51, a value entirely consistent with those determined for PNIPAm50 and PNIPAm50-S-PROPA, and likewise indicates successful and quantitative formation of PNIPAm50-S-ALMA. SEC analysis of PNIPAm50-S-ALMA, Table 1, indicated a very small increase in the polydispersity index from 1.18 to 1.22. However, the chromatogram was unimodal and symmetric.

With the PNIPAm50-S-PROPA and PNIPAm50-S-ALMA polymers in-hand, the allylic and propargylic end-groups were modified with three different small molecule thiols via both radical-mediated thiol-ene and thiol-yne reactions. Specifically, the unsaturated end-groups were reacted with 6-mercaptohexan-1-ol (S-HexOH), hexane-1-thiol (S-Hex), and a 3-mercaptopropyl polyhedral oligomeric silsesquioxane (S-POSS) with R = isobutyl under photochemical conditions at ambient temperature, humidity, and under a normal air atmosphere (Fig. 3).

thumbnail image

Figure 3. Chemical structures of the three representative, commercially available, thiols used in the radical thiol-ene and radical thiol-yne end-group reactions.

Download figure to PowerPoint

These thiols were chosen as common commercially available species with hydrophobic and hydrophilic character. However, this chemistry is not limited to these molecules, and a range of other, functional thiols, including those of potential biomedical importance, can be employed in such thiol-ene and thiol-yne radical reactions.6, 8, 58 Figure 4(A–F) shows the 1H NMR spectra, recorded in CDCl3, of the products obtained from the reaction of PNIPAm50-S-PROPA and PNIPAm50-S-ALMA with the three thiols.

thumbnail image

Figure 4. 1H NMR spectra, recorded in CDCl3 for (A) PNIPAm50-S-ALMA-S-HexOH), (B) PNIPAm50-S-PROPA-(S-HexOH)2, (C) PNIPAm50-S-ALMA-S-Hex, (D) PNIPAm50-S-PROPA-(S-Hex)2, (E) PNIPAm50-S-ALMA-S-POSS, and (F) PNIPAm50-S-PROPA-(S-POSS)2 demonstrating the high efficiency of the radical mediated thiol-ene and thiol-yne reactions. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Download figure to PowerPoint

Figure 4(A,B) show the 1H NMR spectra for PNIPAm50-S-PROPA-(S-HexOH)2 and PNIPAm50-S-ALMA-(S-HexOH). In the case of the thiol-ene reaction on PNIPAm50-S-ALMA, two important differences between Figures 4(A) and 2(B) are evident. Firstly, the resonances associated with the allylic end-group, previously observed at δ ∼ 4.5–5.3 ppm, have completely disappeared while new resonances due to the S-HexOH are evident. The resonance of the [BOND]CH2[BOND] group alpha to the OH group is clearly visible at δ ∼ 3.6 ppm while that associated with the [BOND]CH2[BOND] alpha to S is visible at δ ∼ 2.7 ppm but overlaps with another, polymer-related, resonance. A ratio of the integrals of the signals labeled a and b indicates a DP of 53 which agrees almost exactly with the DP values calculated for the PNIPAm homopolymer and the ALMA end-modified species and, again, suggests a quantitative thiol-ene reaction. Similar observations were made in the case of the radical thiol-yne reaction of PNIPAM50-S-PROPA with S-HexOH. The resonance associated with the propargylic hydrogens, observed at δ ∼ 4.6 ppm [Fig. 2(A)], is completely absent while the resonance of the methylene hydrogens alpha to the OH group is again present at δ ∼ 3.6 ppm. In this instance, a ratio of a with b indicates a DP of 56, slightly higher than that determined for PNIPAm50-S-PROPA but still indicating essentially quantitative reaction. The reaction of PNIPAm50-S-ALMA and PNIPAm50-S-PROPA with hexane-1-thiol proved difficult to quantify because no distinct, nonoverlapping, resonances associated with S-Hex were observed [Fig. 4(C,D)]. However, after reaction the key allylic and propargylic resonances were completely absent, qualitatively indicating successful and complete reaction. Finally, the allylic and propargylic end-groups of the NIPAm homopolymer were modified by radical reaction with S-POSS yielding novel polymer-nanomaterial hybrid structures. As with the previous reactions, the reaction was monitored via a combination of 1H NMR spectroscopy and SEC. Figure 4(E,F) show the 1H NMR spectra, recorded in CDCl3, for the S-POSS-modified NIPAm homopolymers. Consistent with the previous two thiols, following the radical thiol-ene and thiol-yne reactions neither the allylic nor propargylic resonances are observed in the 1H NMR spectra qualitatively indicating complete consumption of the unsaturated bonds. In both instances resonances associated with the S-POSS moiety are clearly present although most overlap with polymer signals. One signal that can be integrated with reasonable accuracy is the doublet at δ ∼ 0.6 ppm that is assigned to the methylene group of the isobutyl alkyl groups on the POSS cage and is labeled b in Figure 4(E). In the case of the reaction with PNIPAm50-S-ALMA, a ratio of a with b indicates a DP of 47, again a value that agrees extremely well with that calculated for the precursor PNIPAm homopolymer. For the thiol-yne reaction on PNIPAm50-S-PROPA, a ratio of the same two resonances indicates a DP of 58 [Fig. 4(F)]. This value would seem to indicate that the radical thiol-yne reaction did not go to completion, although assuming that the initial DP of 50 for the NIPAm homopolymer is correct; this value indicates a yield ≥90%. Given the fact that the signals associated with b are not fully baseline resolved, the error inherently associated with these integral values, and the large bulky nature of the S-POSS isobutyl molecule, this value is still in reasonable agreement with the expected, ideal value of 50. In all instances, the polydispersity indices for the mono- and bisend-modified PNIPAm50 homopolymers remained essentially the same as the parent homopolymer with values in the range Mw/Mn = 1.14–1.23.

One reason PNIPAm attracts significant attention is due to its readily accessible lower critical solution temperature (LCST). It is well documented that the LCST of PNIPAm is ∼32 °C,59–62 close to physiological temperature. Although varying the degree of polymerization,63 and topology,64 offers some degree of control and tunability of the LCST in NIPAm homopolymers; for materials with low molecular weights, the effect on the LCST of end-groups can become important.65, 66 As such, and given the relatively low molecular weight of the parent NIPAm homopolymer, it was anticipated that the introduction of either one or two S-HexOH, S-Hex, or S-POSS groups, all with differing hydrophilicities/hydrophobicities, at the PNIPAm ω chain terminus would have an effect on the LCST, and general aqueous solution behavior of the PNIPAm50-based materials. Indeed, and as clearly noted by Kujawa et al.,63 the introduction of hydrophobic end-groups can affect the aqueous phase behavior of PNIPAm in two distinct ways. Firstly, the general miscibility of the homopolymer in water is worsened due to the interactions between water and the hydrophobic end-groups, and secondly, sufficiently large hydrophobic species at the chain end can lead to aggregation/micellization reducing the mixing entropy of the polymer chains, although this does not really affect the hydration of the main polymer chains. Both of these factors will tend to favor phase separation, and in the case of PNIPAm, a lowering of the LCST.

The LCSTs of the parent NIPAm50 homopolymer, PNIPAm50-S-ALMA, PNIPAm50-S-PROPA, and the six mono-/bis-modified materials were evaluated using a combination of simple optical measurements and dynamic light scattering (DLS). For the optical measurements, 1 wt % solutions were gently heated and the temperature at which the starting water-white solutions became noticeably opaque was recorded as the LCST. In the case of DLS measurements, samples were heated from 5 to 40 °C with a measurement taken at each increment of 1 °C after equilibration for 2 min. The temperature at which an abrupt change in the hydrodynamic diameter (Dh) was observed was taken as the LCST.

Consider the optical measurements. In the case of the NIPAm50 homopolymer, the cloud point (Tcp) was determined to be 29 °C [Fig. 5(A)]. This value is very close to the generally quoted value of 32 °C for the coil-to-globule-to-precipitate transition associated with PNIPAm. Given the low molecular weight of this sample (Mn ∼ 5900) this observed difference of three degrees may be a direct result of the hydrophobic phenyl dithioester end-group reducing the general solubility of the PNIPAm chains. However, as noted by Kujawa,63 there is a clear molecular weight dependence on the LCST for NIPAm homopolymers. For example, in their studies of octadecyl α,ω-end-functionalized PNIPAm's of varying molecular weights (with Mn values ranging from 12,000 to 49,000), at a constant concentration of 1.0 g/L, the LCST was observed to fall from 30.7 °C for the highest molecular weight species down to 25.1 °C for the lowest molecular weight species. Therefore, this observed value of 29 °C may, simply, be a direct result of the low molecular weight of the precursor PNIPAm homopolymer. Essentially, identical observations were made in the case of the PNIPAm50-S-ALMA and PNIPAm50-S-PROPA homopolymers with observed Tcp values of 30.0 and 29.0 °C, respectively [Fig. 5(B,C)]. In the case of the PNIPAm50-S-ALMA-S-HexOH and PNIPAm50-S-PROPA-(S-HexOH)2 polymers, the Tcp values were observed to be 33 and 32 °C, respectively, identical to the commonly reported LCST for PNIPAm [Fig. 5(D)]. Presumably, these elevated Tcp values are due to the presence of the terminal, hydrophilic, H-bonding OH functional groups that offset any hydrophobic effect of the precursor ALMA/PROPA segments as well as the C6 fragment associated with S-HexOH. There appears to be little, if any, effect of having two versus one OH groups. This proposed beneficial effect of the hydrophilic OH groups is supported by the observation that in the case of the PNIPAm50-S-ALMA-S-Hex and PNIPAm50-S-PROPA-(S-Hex)2 polymers, the measured Tcp values are both 27 °C, several degrees lower than for the precursor PNIPAm50-S-ALMA and PNIPAm50-S-PROPA species [Fig. 5(E)]. This lowering on the LCST is attributed to the increase in hydrophobicity associated with the introduction of the hexyl end-groups. Again, however, there appears to be no distinguishable effect upon the introduction of one versus two hexyl chains. Finally, in the case of the POSS-modified PNIPAm's, there was a clear effect with respect to the introduction of one or two POSS cages at the ω-terminus. In the case of the difunctional species, PNIPAm50-S-PROPA-(S-POSS)2 [Fig. 5(F)], the end-modified PNIPAm was not water-soluble at the examined concentration of 1 wt % even at temperatures approaching 0 °C. This is consistent with the highly bulky and hydrophobic nature of the POSS cage. In the case of the mono-POSS species, PNIPAm50-S-ALMA-S-POSS, the modified polymer was soluble at the specified concentration and at ambient temperature, and also clearly exhibited LCST behavior. However, repeated measurements indicated an LCST of ∼ 35 °C, a value higher than expected given the low molecular weight of the parent PNIPAm homopolymer and the hydrophobic nature of POSS. We currently have no explanation for this anonymously high value of Tcp.

thumbnail image

Figure 5. Digital pictures demonstrating the reversible temperature-induced phase separation and chemical structures of 1 wt % solutions of (A) PNIPAm50 (B) PNIPAm50-S-ALMA (C) PNIPAm50-S-PROPA (D) PNIPAm50-S-PROPA-(S-HexOH)2 (E) PNIPAm50-S-PROPA-(S-Hex)2 (F) PNIPAm50-S-PROPA-(S-POSS)2, and (G) PNIPAm50-S-ALMA-S-POSS.

Download figure to PowerPoint

The LCSTs as estimated by DLS measurements generally agreed very well with those determined optically. For example, in the case of PNIPAm50 and PNIPAm50-S-ALMA the determined Tcp values were, in both instances, 29 °C, identical to those measured optically. The Tcp value for PNIPAm50-S-PROPA was found to be several degrees higher than that determined optically. In all three instances, however, the DLS data, see Figure 6 for an example, indicated some type of aggregation process with the formation of species, before macroscopic phase separation, with hydrodynamic sizes in the micron range. Although the soluble-to-insoluble transition in PNIPAm, and other thermoresponsive polymers, is often very sharp, as highlighted in the cooperative dehydration mechanism,67 Wu and coworkers68 noted that in dilute solutions of hydrophobically modified PNIPAm it is possible for chains to collapse and associate to form a stable mesoglobular phase69 between single chain globules and macroscopic precipitate. Although the exact mode of “self-assembly” of these end-group modified PNIPAms' is unclear, there certainly appears to be a colloidally stable species present before macroscopic phase separation. However, given the micron size of these aggregates, it is highly unlikely that these species are well-defined micelle-like structures that can, and have, been observed with end-group modified PNIPAms bearing larger hydrophobic end-group species.63 Good agreement between the Tcp values obtained by DLS and optical measurements was also observed in the case of the PNIPAm50-S-ALMA-S-Hex, PNIPAm50-S-PROPA-(S-Hex)2, and PNIPAm50-S-ALMA-S-POSS polymers with the S-Hex-based polymers yielding Tcp values of 26 °C by DLS, versus 27 °C optically. DLS indicated an identical, but still anonymously high value, of 35 °C for the S-POSS derivative. In the case of PNIPAm50-S-ALMA-S-HexOH, the DLS Tcp was significantly lower, at 26 °C, than that determined optically (33 °C).

thumbnail image

Figure 6. Plot of Z-average hydrodynamic diameter versus temperature for a 1 wt % aqueous solution of PNIPAm50 demonstrating the determination of the LCST.

Download figure to PowerPoint

SUMMARY AND CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY AND CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES AND NOTES

The synthesis of a range of well-defined, end-group modified poly(N-isopropylacrylamide)s (PNIPAm's) is described and the effect of end-group modification on the lower critical solution temperature noted. The end-group modifications were accomplished on a RAFT-prepared PNIPAm via a sequential process involving tandem amine-mediated dithioester end-group cleavage/phosphine-mediated thiol-ene reactions with allyl methacrylate and propargyl acrylate followed by radical-mediated thiol-ene and thiol-yne reactions to yield the mono and bisend functionalized species. This is the first time the sequential thiol-ene/thiol-ene and thiol-ene/thiol-yne reactions, proceeding via inherently different mechanism, have been applied in polymer synthesis. We highlight the high efficiency of the radical and nucleophile click thiol-ene reactions as well as its sister radical-mediated thiol-yne reaction. End group modifications are shown to be quantitative yielding novel-modified PNIPAms. The lower critical solution temperatures of the precursor PNIPAm and the end-group-modified species were examined using a combination of dynamic light scattering and optical measurements, yielding cloud point values that were generally in good agreement. The observed differences in the cloud points was rationalized in terms of the hydrophilicity/hydrophobicity of the introduced end-groups.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY AND CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES AND NOTES

A. B. Lowe would like to thank the Materials Research Science and Engineering Center (MRSEC, DMR-0213883) at USM for funding this research in the form of a stipend for B. Yu. Prof. Françoise Winnik is kindly thanked for her input and suggestions.

REFERENCES AND NOTES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL
  5. RESULTS AND DISCUSSION
  6. SUMMARY AND CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES AND NOTES