Behavior of Poly(ε-caprolactone)s (PCLs) Coalesced from Their

Mar 19, 2012 - ... with a CCID-IRIS/RGB color video camera made by Sony Corp. ..... Thus our challenge, which is also offered to the reader, is: how c...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/Macromolecules

Behavior of Poly(ε-caprolactone)s (PCLs) Coalesced from Their Stoichiometric Urea Inclusion Compounds and Their Use as Nucleants for Crystallizing PCL Melts: Dependence on PCL Molecular Weights Alper Gurarslan, Jialong Shen, and Alan E. Tonelli* Fiber & Polymer Science Program, Department of Textile Engineering, Chemistry, and Science, North Carolina State University, College of Textiles, Campus Box 8301, Raleigh, North Carolina 27695-8301, United States

ABSTRACT: We have formed noncovalent inclusion compounds (ICs) between guest poly(ε-caprolactone) (PCL) chains, with molecular weights ranging from ∼2000 to 80 000 g/mol, and host urea (U). Upon careful removal of the U host, each of the guest PCL chains were coalesced from their U-IC crystals to produce coalesced samples (c-PCLs). As previously observed for PCL and other polymer guests when coalesced from their ICs formed with host cyclodextrins (CDs), upon cooling from their melts, PCLs coalesced from their U-ICs also show enhanced abilities to crystallize, regardless of their molecular weight. Also consistent with polymer guests, including PCL, that were coalesced from their CD-ICs, c-PCLs obtained from their U-ICs retain their enhanced abilities to crystallize even after spending long times (weeks or more) in the melt. Because, unlike CD hosts, U does not thread over guest polymer chains in their ICs, we conclude that the enhanced ability of c-PCLs to crystallize from their melts upon removal of either host from their ICs is solely a consequence of their coalesced conformations/structures/ morphologies, which are stable to prolonged melt-annealing. Furthermore, because c-PCLs with chain lengths well below and well above those corresponding to the entanglement molecular weight of PCL behave similarly, we conclude that their enhanced ability to crystallize from the melt is likely an exclusive consequence of the extended and unentangled arrangement of their coalesced chains. In addition, c-PCLs obtained from their U-ICs are observed to effectively act as self-nucleants, when added in small amounts to as-received PCLs, to produce nuc-PCLs, which like neat c-PCLs exhibit an enhanced ability to melt-crystallize.

■

INTRODUCTION Recently,1−5 we have observed that the bulk behaviors of polymers can be significantly altered from those normally processed from their solutions and melts. This is achieved by first forming noncovalent inclusion compounds (ICs) between guest polymers and certain small molecule hosts like cyclodextrins (CDs), where the guest polymer chains are confined to narrow channels (∼0.5−1.0 nm in diameter) and so are necessarily highly extended and separated from each other, as illustrated in Figure 1. Careful removal of the host crystalline lattice results in a coalesced polymer sample, as depicted there, where the coalesced guest polymer chains are suggested to be reorganized in a highly extended and unentangled manner carried over from their isolation and extension in their crystalline ICs. This manner of nanoprocessing polymers has been shown to significantly alter their behaviors. Prior to the present study, we © 2012 American Chemical Society

have coalesced polymers exclusively from their CD-ICs. Here we extend this approach to samples of poly(ε-caprolactone) (PCL) coalesced from ICs formed with the host urea (U) (see Figure 2) to learn if their behaviors are similar to or distinct from PCLs coalesced from CD-ICs, where the guest PCL chains are threaded through the cyclic host CDs. In addition, we examine the potential effect(s) of chain length/molecular weight on the behaviors of PCL samples coalesced from their U-ICs. PCL is a biodegradable/bioabsorbable polyester first synthesized in the 1930s by ring-opening polymerization of ε-caprolactone. Today, PCL is well established in the biomedical field, often acting as the main component for Received: February 8, 2012 Revised: March 8, 2012 Published: March 19, 2012 2835

dx.doi.org/10.1021/ma300270g | Macromolecules 2012, 45, 2835−2840

Macromolecules

Article

and found to be improved over those of films formed from asreceived PCL. Here we examine the effectiveness of PCLs coalesced from their U-ICs to also act as effective self-nucleants.

■

EXPERIMENTAL SECTION

Materials. Four types of poly(ε-caprolactone) having numberaverage molecular weights of 80 000, 42 500, 10 000, and 2000 g/mol (asr-PCL-80, -40, -10, and -2) (see Table 1) were obtained from Sigma-Aldrich, as were acetone, methanol, chloroform, dimethyl sulfoxide, and urea. It should be noted that each PCL contains a centrally located −(CH2−CH2−O−CH2−CH2−)− unit, as shown below. As a consequence, the two connected PCL chains run in opposite directions.

Table 1. PCL Samples

Figure 1. Formation of and coalescence of a polymer sample from its crystalline cyclodextrin inclusion complex.

sample ID PCL-2 PCL-10 PCL-40 PCL-80 a

commercial name polycaprolactone polycaprolactone polycaprolactone polycaprolactone

diol 189421 440752 181609 440744

Mna (g/mol)

n

2000 10000 42500 70−90000

8 43 186 350

Provided by the manufacturer.

Methods. Formation of PCL-U-ICs. 1 g of PCL was dissolved in 100 mL of acetone at 50 °C, and 32.5 g of urea (U) was dissolved in 200 mL of methanol also at 50 °C, and the U solution was slowly added to the PCL solution in a dropwise manner. The temperature was lowered after mixing was complete, but stirring continued at room temperature for an additional 30 min. A further 2 days was allowed for inclusion complex formation, after which the resulting precipitate was filtered and vacuum-dried. Coalescence of PCL from Its U IC. U can be removed from the PCL-U-IC by two techniques. Stirring 1 g of PCL-U-IC in 50 mL of methanol can completely remove the U. In a second route, 1 g of PCLU-IC was stirred in excess water for 1 day and then filtered and dried. Then the resulting material was stirred in 35 mL of methanol to completely remove the U and obtain coalesced PCL. Though the second route takes longer, it requires less methanol. Finally, in both cases the coalesced PCL samples were vacuum-dried for 1 day. PCL Films. Solvent cast PCL films were obtained by dissolving 1 g of PCL in 100 mL of acetone and then evaporating all the solvent. PCL films nucleated with c-PCL were obtained by solvent casting of 2.5% c-PCL with 97.5% as-received (asr)-PCL from acetone, which did not dissolve the c-PCL nucleant. All solvent cast films were vacuum-dried before further characterization. Melt-pressed films of asr and c-PCL-10 were prepared for polarized microscopy. FTIR. A Nicolet 510P FTIR spectrometer was employed in the range 4000−400 cm−1, with a resolution of 4 cm−1. FTIR data were analyzed by using Omnic software. 64 scans were collected for each spectrum. DSC. Differential scanning calorimetric (DSC) thermal scans were performed with a Perkin-Elmer Diamond DSC-7 instrument. Heating scans were started at 0 °C, and each sample was heated to 75 °C. After holding each sample in the melt for 5 min, they were cooled to 0 °C at the same rate they were previously heated (20 °C/min). Nitrogen was used as the purge gas. DSC data were analyzed with Pyris software, and melting and crystallization temperatures and enthalpies were calculated automatically by the software from the areas of their endothermic and exothermic peaks. Polarized Microscopy. Polarized optical microscopic observations of asr and c-PCL-10 thin films were performed on a Nikon Eclipse 50i POL optical microscope equipped with a CCID-IRIS/RGB color video camera made by Sony Corp. Photographs was taken by using crossed polarizers and a 1/4 λ plate.

Figure 2. Poly(ε-caprolactone) repeat unit (top left) and urea (top right), where carbons are gray, hydrogens white, oxygens red, and nitrogens blue. Space-filling drawing of a channel in the polyethyleneU-IC, which is isomorphic with that of the PCL-U-IC (bottom).6

bone and cartilage tissue scaffolding. Because of its relatively low melting temperature, PCL can be melt processed using very simple techniques and is often strengthened upon addition of fillers, such as fibers or particulates.7 Often times, PCL is used in combination with another biodegradable polyester, poly(L-lactic acid), and their blends have been shown to have good gas barrier properties.8 Despite PCL’s many unique properties, it is not a very strong material. Because of its low softening temperature (Tm ∼ 59− 64 °C and glass-transition temperature Tg ∼ −60 °C)9 and comparatively poor room temperature physical properties (tensile strength of 20−42 MPa and a tensile modulus of 0.2−0.44 GPa),10 PCL has not made great headway as a load bearing implantable material. If PCL could be strengthened in some fashion, then it might gain new biomedical applications, in particular for load-bearing implantables with superior compressive strengths. For example, if nucleating agents are used in the melt-crystallization of PCL to increase the homogeneity of its semicrystalline morphology and resultant strength, they must also be bioabsorbable and nontoxic for use as implantable materials. This was previously suggested for PCL coalesced from its α-CD-IC (c-PCL),3 where, in addition, a variety of its behaviors and properties as films were examined 2836

dx.doi.org/10.1021/ma300270g | Macromolecules 2012, 45, 2835−2840

Macromolecules

Article

Figure 3. FTIR spectra of asr-PCL (1), c-PCL (2), PCL-U-IC (3), and U (4). NMR. Dilute solutions of c-PCL and a mixture of 10 wt % urea and 90 wt % asr-PCL were prepared in 1:2 d6-DMSO:CDCl3 mixtures, and their 1H NMR spectra were recorded on a Bruker 700 MHz spectrometer equipped with 5 mm i.d. CPTCI (1H/13C/15N/D) Z-axis gradient cryoprobe. NMR data were analyzed with Topspin 2.1 software.

Table 2. Crystallization Temperatures of Different Molecular Weight asr- and c-PCLs Tc (°C)

■

RESULTS AND DISCUSSION IC formation was confirmed with FTIR and DSC. FTIR spectra of PCL-U-ICs exhibit peaks characteristic of both PCL and U,

sample

asr-PCLs

c-PCLs

Tc increments

PCL-2 PCL-10 PCL-40 PCL-80

17.0 16.8 18.8 11.7

26.5 32.5 35.0 33.3a

9.5 15.7 16.2 21.6

a For comparison, PCL-80 coalesced from its IC formed with host αCD crystallized at Tc = 36.8 °C from the melt upon cooling at −20.0 °C/min.

Table 3. Crystallization Temperatures of c-PCLs before and after Annealing 2 and 4 weeks at 80 °Ca Tc (°C) sample

initial

after annealing 2 weeks at 80 °C

c-PCL-2 c-PCL-10 c-PCL-40 c-PCL-80

26.5 32.5 35.0 33.3

26.0 30.9 34.1 30.1

after annealing 4 weeks at 80 °C 26.4 31.7 33.8 31.1

a

Because flow times observed in a Cannon-Ubbelhode viscometer for 1 g/dL CHCl3 solutions of asr- and c-PCLs, the latter both before and after annealing, were closely similar, long-term melt-annealing at 80 °C did not thermally degrade our PCL samples.

comparison, the melting temperatures of c- and asr-PCL-80 were 55 and 61 °C, respectively. Coalesced PCLs were obtained from their U-ICs after removing all of the host U. As seen in the FTIR spectra presented in Figure 3-(1),(2), it appears that very little if any U remains after coalescence. More sensitive 1H NMR observations also confirm the virtual absence of remnant host U, as demonstrated in Figure 5 by the absence of the urea proton peak at ∼5.35 ppm in the spectrum of c-PCL. Table 2 summarizes the melt-crystallization of as-received (asr-) and c-PCL samples. It is clear that all c-PCL melts crystallize at significantly higher temperatures than their corresponding asr-melts, with the difference between them increasing with their molecular weights. It is also apparent that

Figure 4. DSC thermograms of PCL-80-U-IC: first heating (A), second heating (B), first cooling (C).

as seen in Figure 3. However, this is insufficient to distinguish PCL-U-IC from a physical mixture of PCL and U, but in conjunction with DSC observations they can be distinguished. As seen in Figure 4, the first heating scan of PCL-U-IC does not show a melting peak for PCL. Instead, melting peaks for free U and PCL-U-IC are observed at 134 and 142 °C, since excess U was employed in forming the PCL-U-IC. After melting of the PCL-U-IC, in the second heating scan we see melting peaks for free PCL and U as well as that of the partially re-formed PCL-U-IC.6 All four U-IC’s formed using PCLs with different molecular weights were confirmed this way. For 2837

dx.doi.org/10.1021/ma300270g | Macromolecules 2012, 45, 2835−2840

Macromolecules

Article

Figure 5. The 700 MHz 1H NMR spectra of asr-PCL with 10 wt % of U (bottom) and c-PCL (top), which have been equally expanded.

Figure 6. First cooling DSC scans from the melts of neat asr-PCL-40, self-nucleated nuc-PCL-40 (2.5 wt % c-PCL-40 + 97.5 wt % asr-PCL-40), and c-PCL-40.

the assumed extended, unentangled natures of their constituent PCL chains (see Figure 1). The distinction between the Tcs of asr- and c-PCLs is maintained even after prolonged melt annealing of their c-melts at 80 °C, as can be seen in Table 3. This behavior suggested that c-PCLs would make excellent nucleation agents for enhancing the melt-crystallization of asr-PCLs. In Figure 6, the first cooling DSC scans of neat asr-, c-, and nuc-PCL-40 (97.5 wt % asr-PCL-40 self-nucleated with 2.5 wt % c-PCL-40) reveal that the resulting nuc-PCL-40 and neat cPCL-40 crystallize more readily than asr-PCL-40. The crystallization exotherms of c- and nuc-PCL-40 are much narrower/sharper, with crystallization ranges of ∼9−10 °C compared with ∼30 °C for asr-PCL-40, and occur at significantly higher temperatures (35 and 32.3 °C, respectively) compared with the crystallization exotherm of asr-PCL-40 (18.8 °C). Table 4 presents a comparison of melt-crystallization temperatures for asr-, c-, and nuc-samples of all four PCLs.

Table 4. Crystallization Temperatures of Different Molecular Weight asr-, nuc-PCLs, and PCLs Nucleated with c-PCL-10 Tc (°C) sample

PCL-2

PCL-10

PCL-40

PCL-80

as-received coalesced self-nucleated nucleated with c-PLC-10

17.0 26.5 18.8 21.3

16.8 32.5 27.9 27.9

18.8 35 32.3 29.5

11.7 33.3 29.6 28.1

the Tc of the highest molecular weight asr-sample (PCL-80) is significantly lower than the Tcs of the others, which are very similar. This may be due to the presumably highly entangled nature of the high molecular weight asr-PCL-80 melt. Most important is the similarity in the melt-crystallization temperatures of all the c-PCL samples, despite the vast differences in their molecular weights. This is consistent with 2838

dx.doi.org/10.1021/ma300270g | Macromolecules 2012, 45, 2835−2840

Macromolecules

Article

Figure 7. Optical microscopy images (500×, crossed polarizers, 1/4 λ plate) of melt-pressed asr-PCL-10 (left) and c-PCL-10 (right) films.

In addition to the self-nucleated samples, the Tcs of asr-PCL-2, -40, and -80 nucleated with c-PCL-10 are also given there, and they are seen to differ only minimally from the Tcs of their selfnucleated films. We can also see that the crystallization of nucPCL-2 is very similar to and not enhanced compared with asrPCL-2, independent of whether it is self-nucleated or nucleated with c-PCL-10. This is likely a result of its very low molecular weight, corresponding to just 8 repeat units (see Table 1), but not to the absence of entanglements, because all of the c-PCL melts are believed to crystallize from unentangled melts. CPCL-2 crystallizes less than 10 °C higher than asr-PCL-2, while the higher molecular weight c-PCLs crystallize from 16 to 22 °C higher in temperature than their asr-samples. The −20 °C/ min cooling rate employed may not allow time sufficient for the asr-PCL chains to fully crystallize even though the nucleating cPCL chains already have. Crystallinities calculated from second heating DSC scans, assuming ΔHm° = 135 J/g, are as follows: asr-, c-, and nucPCL80s are ∼34 ± 2% crystalline. If they have been melt pressed into film and are held at RT for a long time, all their crystallinities increase to ∼50%. PCL40: asr- and nuc- ∼ 49% crystallinities, PCL10: asr- and nuc- ∼ 47% crystallinities, and PCL2: asr- and nuc- have ∼60% crystallinities. To briefly summarize our observations, we can say that all four PCLs, with molecular weights ranging from well below to well above the reported11 entanglement molecular weight of PCL, ∼15 000 g/mol, when coalesced from their U-ICs are reorganized in a manner that significantly increases their meltcrystallization temperatures above those of their asr-samples. The difference in their thermal behaviors, and presumably their structures, is retained in their melts for extensive periods of time (see Table 3). In Figure 7, the optical microscope images of asr- and c-PCL-10 films observed under crossed polarizers are presented. There it is clear that the c-PCL-10 film has a finer scale, more homogeneous semicrystalline morphology (more uniform distribution of crystalline and amorphous regions), which apparently remains distinct from that of asrPCL-10 even after extensive melt annealing.12

their coalesced conformations/structures/morphologies, which are stable to prolonged melt-annealing. Furthermore, because cPCLs with chain lengths well below and well above those corresponding to the entanglement molecular weight of PCL behave similarly, we conclude that their enhanced ability to crystallize from the melt is likely a consequence of the extended and unentangled arrangement of their coalesced chains. PCL coalesced from its U-IC, while chemically identical to asr-PCL, shows a significantly improved melt-crystallizability because of its extended and unentangled coalesced structure. Small amounts of c-PCL can be used as a nucleant for the meltprocessing of asr-PCL. As a consequence, c-PCL may eventually be useable as a fully compatible “stealth” nucleant in the formation of biodegradable/bioabsorbable, load-bearing, implantable medical devices made from PCL.14

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are grateful to Dr. Hanna Gracz for recording the 1H NMR spectra of asr- and c-PCLs and to the National Textile Center (US Commerce Department) and the TECS Department in the College of Textiles at NC-State University for partially supporting this work.

■

REFERENCES

(1) Tonelli, A. E. Adv. Polym. Sci. 2009, 222, 115. (2) Tonelli, A. E. J. Polym. Sci., Part A2: Polym. Phys. Ed. 2009, 1543. (3) Williamson, B. R.; Krishnaswamay, R.; Tonelli, A. E. Polymer 2011, 52, 4517. (4) Mohan, A.; Gurarslan, A.; Joyner, X.; Child, R.; Tonelli, A. E. Polymer 2011, 52 (4), 1055−1062. (5) Gurarslan, A.; Tonelli, A. E. Macromolecules 2011, 44 (10), 3851. (6) Choi, C.; Davis, D. D.; Tonelli, A. E. Macromolecules 1993, 26, 1468. (7) Liu, H.; Han, C.; Dong, L. Polym. Compos. 2010, 31 (9), 1653. (8) Chiellini, E. Environmentally Compatible Food Packaging; Woodhead Publishing in Food Science, Technology and Nutrition; CRC: Boca Raton, FL, 2008. (9) Middleton, J. C.; Tipton, A. J. Biomaterials 2000, 21, 2335. (10) Mano, J. F.; Sousa, R. A.; Boesel, L. F.; Neves, N. M.; Reis, R. L. Compos. Sci. Technol. 2004, 64, 789. (11) Grosvenor, M. P.; Staniforth, J. N. Int. J. Pharm. 1996, 135, 103. (12) We have recently addressed the issue of why the organization/ structure/morphologies/behaviors of polymers coalesced from their noncovalently bonded ICs, formed with both cyclodextrin and urea hosts, differ from those obtained by solution and melt processing.13 What do we know: (A) They behave differently. (i) They crystallize

■

CONCLUSIONS As previously observed for PCL and other polymer guests when coalesced from their ICs formed with host CDs,1−5 upon cooling from their melts, PCLs coalesced from their U-ICs also show enhanced abilities to crystallize, regardless of their molecular weight. Also consistent were retention of the enhanced abilities of PCLs coalesced from their U-ICs to crystallize even after spending long times (weeks or more) in the melt. Because, unlike CD hosts, U does not thread over guest polymer chains when forming their ICs, we conclude that their enhanced ability to crystallize from their melts upon removal of the host U molecules is solely a consequence of 2839

dx.doi.org/10.1021/ma300270g | Macromolecules 2012, 45, 2835−2840

Macromolecules

Article

more readily, with higher Tcs, over a narrower temperature range, sometimes in unusual polymorphs, and apparently without chain folding. (ii) They have elevated Tgs. (iii) Their blends are intimately mixed. (iv) Their amorphous regions are more dense. (v) Their melts have much lower zero shear viscosities. (vi) They produce stronger, less extensible films and fibers. (vii) They are less permeable to gases (CO2). (B) All the above behaviors remain, even after extensive periods (weeks) spent in their melts, and are independent of their molecular weights and their IC host (CD or U). (C) Because polymerICs are generally obtained as crystalline powders, we anticipate that upon coalescence from them the resulting coalesced polymer samples will consist of small unoriented regions (smaller than the sizes of their IC crystals) of extended and unentangled chains (see Figure 1). [Please note that, though the guest polymer chains in each crystal of the CD-IC powder are expected to be coalesced somewhat as indicated in Figure 1, we do not mean to imply that this results in the initial overall macroscopic orientation of all extended and unentangled chains. Rather, we suggest that the initial macroscopic organization of polymer chains in the melt might resemble something like a collection of small “nematic-like” regions randomly arranged without a preferred orientation of their directors. (See: Beers, D. E.; Ramirez, J. E. J. Text. Inst. 1990, 81, 561 for a discussion of Vectra, a liquid-crystalline ester/ arylate copolymer, that evidences a macroscopically anisotropic melt, much like that suggested above locally for coalesced polymers). ] (D) This anticipated structure(s) is consistent with their behaviors noted in (A), but we have not yet connected them to the long-time, hightemperature stability noted in (B). Thus our challenge, which is also offered to the reader, is: how can we directly observe the structure(s)/ organization of coalesced polymers in their melts? (13) Gurarslan, A.; Joijode, A. S.; Tonelli, A. E. J. Polym. Sci., Part-A2: Polym. Phys. Ed. 2012. (14) Gurarslan, A.; Tonelli, A. E. Self-Reinforcement of Polymer Composites, ACS National Meeting, Anaheim, 52, 1, 186, March 2011.

2840

dx.doi.org/10.1021/ma300270g | Macromolecules 2012, 45, 2835−2840