Ping Song*a,
Lin Sangb,
Liuchun Zhengc,
Chao Wanga,
Kankan Liud and
Zhiyong Wei
*e
aSchool of Materials Science and Engineering, North University of China, Taiyuan 030051, China. E-mail: songping@nuc.edu.cn
bSchool of Automotive Engineering, Dalian University of Technology, Dalian 116024, China
cBeijing National Laboratory for Molecular Sciences, Key Laboratory of Engineering Plastics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China
dSchool of Chemical and Environmental Engineering, North University of China, Taiyuan 030051, China
eSchool of Chemical Engineering, Dalian University of Technology, Dalian 116024, China. E-mail: zywei@dlut.edu.cn
First published on 23rd May 2017
To gain an insight into the role of bound water of a nucleating agent in polymer nucleation, a biobased nucleating agent, orotic acid (OA), was selected as a model to investigate the effects on crystallization of poly(L-lactic acid) (PLLA). In such a context, two commercially available types of OA in anhydrous (OA-a) and monohydrated (OA-m) forms were melt mixed with PLLA, and their nucleation effectiveness on nonisothermal and isothermal melt crystallization of PLLA was comparatively studied. Results indicate that both forms of OA can significantly improve nonisothermal crystallization temperature and degree of crystallinity, overall isothermal crystallization rate, as well as nucleation density of PLLA. Interestingly, OA-a shows more prominent nucleation efficiency than OA-m. That is, the bound water of OA-m and its dehydration transition play a negative role in nucleation effects on PLLA crystallization. It is attributed to the deteriorated dispersion and the reduced active concentration of dehydration-transformed OA-a in PLLA/OA-m blends, as compared with pristine OA-a in PLLA/OA-a blends. Furthermore, an epitaxial mechanism is proposed to explain the nucleation phenomenon of PLLA/OA blends.
As a semicrystalline polyester with high melting point, however, PLLA has very slow crystallization rate and a low degree of crystallinity in traditional processing such as injection molding and thus poor thermal resistance, with heat deflection temperature (HDT) near its glass transition temperature (∼55 °C).10 This shortcoming has limited the wider utilization of PLLA, especially for durable articles. Therefore, manipulation of crystallization towards high performance is essential for the practical applications of PLLA.11,12 Adding nucleating agents (NAs) is a simple and efficient strategy to reduce the free energy barrier for critical nuclei formation and induce the high crystallization temperature at low supercooling. Until now, diverse inorganic and organic species have been introduced as NAs for PLLA, and their effects on crystallization kinetics and final properties have been investigated in literatures.13–31 In general, inorganic NAs have relative low cost and easy availability, but less efficient nucleation effects due to their poor compatibility with PLLA, as a consequence, often require additional surface treatment via organic modification. In contrast, organic NAs can be well dispersed in PLLA matrix and usually show better nucleation effects, nevertheless most of these chemicals have complicated preparation processes, and even are not biodegradable, bioabsorbable or nontoxic, resulting in that they are undesirable in the utilization related to food, human bodies and ecological environment demand.
In order to promote the crystallization of PLLA without compromising its bio-advantages, several biobased NAs with biodegradability, biocompatibility and nontoxicity like myo-inositol,32 thermoplastic starch,33 orotic acid,34 nucleobases,35 xylan esters,36 cyclodextrin complex,37 lignin,38 nanocelluloses,39,40 amino acids and poly(amino acids)41 have been developed. Among of these biobased NAs, it appears that orotic acid (OA), a naturally occurring substance (e.g., in milk products), is most effective in light of the significant enhancement on crystallization kinetics of PLLA with only 0.3 wt% dosage (crystallization peak temperature is 123.9 °C and crystallization enthalpy is 34.1 J g−1, when nonisothermally crystallized at a cooling rate of 10 °C min−1; crystallization half-time is 0.64 min, when isothermally crystallized at 120 °C).34 Besides, OA has also been shown to be a good nucleating agent for poly(3-hydroxybutyrate) and its copolyesters.42–44
In fact, there are two commercially available types for orotic acid, that is, anhydrous and monohydrated forms, which possess different structural features and properties,45 and may result in discrepant nucleation ability on polymer crystallization. However, previous studies are mainly focused on the nucleation effects of monohydrated form.34,42,43 To our best knowledge, the role of bound water of a hydrated NA and its phase stability and transitions during melt blending process in the nucleation effectiveness for PLLA crystallization are not involved and still unknown, although these understandings are of importance for regulating the crystallization kinetics and the final physical properties of PLLA by reasonable use of orotic acid as an effective nucleating agent.
In this work, to gain insight into the role of bound water of nucleating agent in polymer nucleation, the effects of anhydrous orotic acid (OA-a) and monohydrated orotic acid (OA-m) on the crystallization of PLLA were comparatively investigated. This paper reported on the results about the nonisothermal/isothermal melt crystallization and melting behavior, spherulite morphology and crystal structure. Moreover, their nucleation ability and mechanism were compared and discussed. Some of them are interesting and important, not only for the understanding of the bound water effects in OA, but also for the understanding of the bound water effects in other NAs on polymer crystallization in general. The research reported herein is expected to shed some light on fabricating high-performance PLLA biobased materials.
The Fourier transform infrared (FTIR) spectra were collected by using a Nicolet 6700 spectrometer (Thermo Fisher Scientific, America) equipped with a hot stage. The collection was at a resolution of 2 cm−1 by averaging 16 scans. The samples of pure PLLA and its blends with 1 wt% OA-a and OA-m were prepared by solution casting on potassium bromide plates. Each sample was held at 200 °C for 3 min to erase thermal history, and then it was cooled to 140 °C at a rate of 60 °C min−1. When the temperature reached 140 °C, data collection begun and the spectra were collected at 5 min intervals for pure PLLA and 1 min intervals for OA-nucleated PLLA samples.
The crystal morphologies of pure PLLA and PLLA/OA blends were observed by a polarized optical microscope (Leica DM4500P, Germany) equipped with a digital camera and a hot stage (Linkam THMS600, England). The samples were placed between two microscope cover glasses, melted at 200 °C for 3 min, and then quickly cooled down to a predetermined crystallization temperature at 60 °C min−1. The evolution of spherulites during isothermal crystallization was recorded.
Wide-angle X-ray diffraction (WAXD) analysis was taken on an X-ray diffractometer (Rigaku, Japan) with Ni-filtered Cu Kα radiation (λ = 0.154 nm) at a voltage of 40 kV and a current of 40 mA. Pure PLLA and its blends were isothermally melt-crystallized at 130 °C for 1 h before measurements. The diffraction angle (2θ) covered a range of 3–50° for OA powders and 10–30° for PLLA/OA blends at the scanning speed of 1.2° min−1.
Scanning electron microscopy was conducted using a field emission scanning electron microscope (FE-SEM, Hitachi SU8010, Japan) at an accelerating voltage of 10 kV. The samples were coated with a layer of gold beforehand.
Sample | Tc (°C) | ΔHc (J g−1) | Tcc (°C) | ΔHcc (J g−1) | Tm (°C) | ΔHm (J g−1) | Xc (%) | NE (%) |
---|---|---|---|---|---|---|---|---|
Pure PLLA | — | 0 | 111.4 | 34.4 | 160.2, 165.7 | 34.5 | 0.1 | 0 |
OA-a0.2% | 121.5 | 35.6 | — | 0 | 163.4 | 35.9 | 38.7 | 81.4 |
OA-a0.5% | 123.0 | 35.8 | — | 0 | 164.4 | 35.9 | 38.8 | 83.3 |
OA-a1% | 124.2 | 35.6 | — | 0 | 163.9 | 35.8 | 38.9 | 84.2 |
OA-a2% | 124.3 | 35.5 | — | 0 | 163.8 | 35.6 | 39.1 | 83.9 |
OA-m0.2% | 117.3 | 35.2 | — | 0 | 162.3 | 35.3 | 38.0 | 75.0 |
OA-m0.5% | 120.5 | 35.5 | — | 0 | 163.3 | 35.4 | 38.3 | 78.6 |
OA-m1% | 120.8 | 35.4 | — | 0 | 163.1 | 35.6 | 38.7 | 79.7 |
OA-m2% | 122.2 | 35.6 | — | 0 | 163.2 | 35.5 | 39.0 | 80.8 |
As seen in Fig. 1, the melt-crystallization peak of pure PLLA is almost undetected upon cooling at 10 °C min−1, and a cold-crystallization peak with comparable area to melting peak appears at 111.4 °C after the devitrification at 57.6 °C in the subsequent heating scan, demonstrating that the crystallization of pure PLLA is very slow. Obviously, both the two forms of orotic acid can accelerate the crystallization of PLLA. With the addition of OA-a or OA-m, a sharp crystallization peak is present in the DSC cooling curve and shifts to a higher temperature with increasing the content. The endothermic step of devitrification becomes obscure and the cold-crystallization peak disappears in the following heating process, indicating that these samples crystallize completely during the cooling process. From Fig. 1(b), two melting peaks are observed in the heating scan for pure PLLA, while only one for PLLA/OA blends. The melting behavior of PLLA depends strongly on the crystallization temperature, and the double endothermic peaks can be elucidated by the melt-recrystallization mechanism.47,48 For pure PLLA, the temperature range of cold crystallization (ca. 100–120 °C) is relatively low and less perfect crystals are formed, thus the sample undergoes melting-recrystallization-remelting upon heating. The low-temperature peak is ascribed to the melting of primary crystals produced in cold-crystallization process, and the high-temperature peak is attributed to the melting of reorganized crystals. In the case of OA-containing PLLAs, the Tc is about 10 °C higher than the Tcc of pure PLLA, and relatively perfect crystals are formed. The crystals melt directly without recrystallization, leading to the appearance of a single endothermic peak.
The melt-crystallization temperature (Tc) and degree of crystallinity (Xc) are important indicators to assess the nucleation effects of a nucleating agent on the nonisothermal crystallization of polymers. Generally, a high Tc reflects fast crystallization and a high Xc means strong crystallizability. For clarity, Fig. 2 displays the variation of Tc and Xc with the concentration of anhydrous and monohydrated orotic acid. It is shown that both Tc and Xc increase gradually with an increase of OA content irrespective of its type, and the rising trend becomes inconspicuous when the content is in excess of 1 wt%, especially for OA-a. Besides, the Tc and Xc of PLLAs nucleated by OA-a are on average ∼3 °C and ∼0.4% higher than the ones by OA-m at all compositions, respectively, manifesting that the nucleation ability of OA-a is more efficient than OA-m for the nonisothermal melt crystallization of PLLA.
![]() | ||
Fig. 2 Dependence of (a) crystallization temperature and (b) degree of crystallinity on OA content for PLLA/OA blends in nonisothermal melt-crystallization processes. |
In order to quantitatively analyze the effectiveness of OA-a and OA-m as nucleating agents (NAs), nucleation efficiency (NE) based on the self-nucleation theory is calculated by the following equation:49,50
![]() | (1) |
![]() | ||
Fig. 3 Evolution of relative crystallinity with crystallization time for (a) PLLA/OA-a blends and (b) PLLA/OA-m blends isothermally melt-crystallized at 130 °C. |
The isothermal crystallization kinetics of pure PLLA and PLLA/OA blends was analyzed in light of the well-known Avrami equation using the double logarithmic form: ln[−ln(1 − Xt)] = lnk + n
ln
t, where k is the overall rate constant involving contributions from both nucleation and growth, and n is the Avrami exponent associated to the nature of nucleation and the dimensionality of crystal growth.51,52 The values of k and n can be obtained from the intercept and slope of fitted straight line, respectively, by plotting ln[−ln(1 − Xt)] versus ln
t. Fig. 4 shows such plots for pure PLLA and PLLAs containing two forms of orotic acid isothermally crystallized at 130 °C. To reduce the fitting error, only the data of Xt between 5% and 30% were employed, since initial nucleation and secondary crystallization processes always lead to deviation of Avrami plot from linearity in the segments related to the beginning and the ending of crystallization. The obtained n, k, and R2 (square of correlation coefficient) are listed in Table 2. The values of R2 are approximately equal to 1, indicative of very good linear fittings. The values of n are very close to 3 regardless of the addition of OA-a or OA-m, suggesting a three-dimensional crystal growth with simultaneous nucleation.11 Since the unit of k is min−n and n is not identical for different samples, the values of k are not employed directly to compare the overall crystallization rate.
![]() | ||
Fig. 4 Avrami plots for (a) PLLA/OA-a blends and (b) PLLA/OA-m blends isothermally melt-crystallized at 130 °C. |
Sample | n | k (min−n) | R2 | t1/2 (min) |
---|---|---|---|---|
Pure PLLA | 2.64 | 6.05 × 10−5 | 0.99998 | 34.1 |
OA-a0.2% | 2.99 | 0.62 | 0.99993 | 1.03 |
OA-a0.5% | 2.96 | 1.03 | 1.00000 | 0.87 |
OA-a1% | 2.86 | 1.76 | 1.00000 | 0.73 |
OA-a2% | 2.86 | 1.70 | 1.00000 | 0.73 |
OA-m0.2% | 3.13 | 0.16 | 0.99998 | 1.61 |
OA-m0.5% | 3.09 | 0.47 | 0.99989 | 1.13 |
OA-m1% | 2.90 | 0.57 | 0.99979 | 1.05 |
OA-m2% | 3.11 | 0.82 | 0.99993 | 0.93 |
The crystallization half-time t1/2, which is defined as the time needed to attain half of the final crystallinity (i.e., Xt = 50%), is introduced to distinguish the nucleation effects of OA-a from OA-m on the isothermal crystallization kinetics of PLLA. Usually, a short t1/2 indicates large overall crystallization rate. The value of t1/2 can be obtained directly from the curve of time dependence of relative crystallinity (as shown in Fig. 3), and the results are also listed in Table 2. For ease of comparison, the t1/2 is plotted as a function of the OA content as shown in Fig. 5. It is seen that the t1/2 decreases from 34.1 min for pure PLLA to 1.03 min and 1.61 min for the blends with 0.2 wt% OA-a and OA-m, respectively, at the isothermal crystallization temperature of 130 °C. The t1/2 further decreases, in other words, the overall crystallization rate increases with additional loading of two forms of OA. Note that the t1/2 levels off for the PLLA blends with OA-a when its content exceeds 1 wt%. Combined the nonisothermal crystallization results as shown in Fig. 2, it is drawn that the optimum dosage of OA-a is around 1 wt%. In comparison, the t1/2 of PLLAs containing OA-a is shorter than the one of OA-m for all contents, confirming that anhydrous orotic acid has a higher nucleation effectiveness.
![]() | ||
Fig. 5 Variation of crystallization half-time with OA content for PLLA/OA blends melt-crystallized at 130 °C. |
Fig. 6 shows the DSC heating curves of pure PLLA and OA-containing PLLAs at different loadings after isothermally melt-crystallized at 130 °C. It is shown that the addition of OA-a or OA-m has little effect on the melting behavior of PLLA and all the samples have similar melting temperatures at about 166 °C, implying that the crystal stability of PLLA is not influenced significantly by the presence of orotic acid under this crystallization condition.
![]() | ||
Fig. 6 DSC thermograms recorded in heating scans at a rate of 10 °C min−1 for pure PLLA and its blends with two forms of orotic acid after isothermal melt crystallization at 130 °C. |
Fig. 8 shows the normalized intensities of the bands at 920, 1211 and 1458 cm−1 over the course of crystallization for PLLA blends with 1 wt% OA-a and OA-m. The normalized intensities of the band at 920 cm−1 is used to probe the kinetics of 103 helix formation, while the bands at 1211 and 1458 cm−1 are used to follow the conformational changes in the backbone. The related results of pure PLLA are not given in the present study, because it is quite difficult to crystallize for pure PLLA at such a high crystallization temperature (140 °C) and almost no peak changes can be observed even if the time is increased to 240 min (see Fig. S3 in ESI†). In Fig. 8, it is clear that the induction time and t1/2 of OA-a1% is much shorter than the ones of OA-m1%, proving that OA-a is more effective as compared with OA-m. On the other hand, for either OA-a1% or OA-m1%, the sequence of intensity changes for the three bands at the early nucleation stages is as follows: 1458 cm−1 > 1211 cm−1 > 920 cm−1. This means that the interchain interactions corresponding to skeletal vibrations (δasCH3 and νasCOC + rasCH3) precede the intrachain 103 helix formation, which is similar to the situation of pure PLLA as reported in literatures.23,53,54 Therefore, OA-a and OA-m do not alter the sequence of interchain and intrachain interactions, but accelerate the skeletal conformational ordering and 103 helix formation of PLLA.
![]() | ||
Fig. 8 Normalized peak intensities at 920, 1211, and 1458 cm−1 as a function of crystallization time for (a) OA-a1% and (b) OA-m1%. |
The effects of two forms of OA on the crystalline structure of PLLA were investigated by wide-angle X-ray diffraction (WAXD) measurements. Fig. 10 shows the WAXD profiles of pure PLLA and its blends with 0.5 wt% OA-a or OA-m melt-crystallized at 130 °C for 1 h. For all the three samples, the main diffraction peaks appear at the 2θ of 14.7°, 16.6°, 19.0°, and 22.2°, which are assigned to the (010), (200)/(110), (203), and (015) reflections of the PLLA α crystal.55 The locations of characteristic diffraction peaks are not shifted, implying that the crystalline structure of PLLA is not altered by the incorporation of OA. The intensity of diffraction peaks of pure PLLA is relatively weak, because the isothermal crystallization has not been completed within 1 h at 130 °C, as revealed in the inset of Fig. 3(a).
The structural features of OA-a and OA-m were then characterized by WAXD. Fig. 12 shows the resulting diffraction profiles. A strong diffraction peak is observed at 2θ = 28.8° for both of the two samples, which corresponds to the stacking reflection of layer structures.45 In contrast to OA-m, a broadened diffraction peak appears at 2θ around 20° for OA-a, indicating that OA-a has a lower degree of long-range order. Recently, it has been found that orotic acid and water molecules are linked by strong hydrogen bonds in nearly perfectly planar layers in OA-m, and the stacked layers are maintained in the dehydration product (i.e., OA-a), whereas their lattice parameters are different (a = 5.456 Å, b = 5.964 Å, c = 9.043 Å, α = 95.41°, β = 94.28°, γ = 97.18° for OA-a; a = 5.949 Å, b = 6.836 Å, c = 9.480 Å, α = 73.62°, β = 72.16°, γ = 68.93° for OA-m). Notably, this literature has also pointed out that OA-m loses the bound water and dissociates into OA-a when it is heated above 135 °C under normal or hermetically sealed high-pressure atmospheric conditions.45 In addition, it has been reported that an endothermic peak, which is attributed to the evaporation of the water associated with OA-m, appears at 144 °C before the melting peak of poly(hydroxybutyrate-co-hydroxyvalerate) (PHBV) at 171 °C in the first heating DSC curve of PHBV solvent-cast film containing 5 wt% OA-m.44 In this work, considering the blended samples were prepared in the temperature range of 170–190 °C for 6 min, it is reasonable to conclude that OA-m is transformed into OA-a during the melt extrusion processes, and finally, OA-a acts as the true nucleating agent in PLLA/OA-m blends.
Fig. 13 shows the POM images of PLLA blends with 1 wt% OA-a and OA-m in the molten state. The bright spots in OA-a1% and OA-m1% indicate the pristine OA-a and dehydration-transformed OA-a, respectively, and their size distribution histograms are displayed in Fig. S4.† In contrast to the pristine OA-a in OA-a1% with a particle size of 2.67 ± 1.39 μm, the dehydration-transformed OA-a in OA-m1% tends to agglomerate together and has a larger particle size of 4.95 ± 4.33 μm, showing a uneven dispersion in the PLLA melt. On the other hand, from Fig. 13, it can be seen that the amount of dehydration-transformed OA-a in OA-m1% is smaller than that of pristine OA-a in OA-a1%. This fact is due to the loss of 1 mol water per mol of OA-m. The final concentration of dehydration-transformed OA-a is theoretically ∼10.3% lower than that of pristine OA-a at the same loading level. It is worth stressing that the poor dispersion and the less nucleation sites for PLLA/OA-m blends can also be evidenced by Fig. 9. Consequently, the negative role of bound water of OA-m and its dehydration transition in nucleation effects on PLLA crystallization is attributed to the deteriorated dispersion and the reduced active concentration of dehydration-transformed OA-a in PLLA/OA-m blends, as compared with pristine OA-a in PLLA/OA-a blends.
Up to date, the nucleation phenomena of polymer crystallization induced by NAs have been successfully elucidated by chemical or epitaxial nucleation mechanisms. With regard to the chemical nucleation, NA dissolves in the polymer melt and reacts with it through chain scission, giving rise to the formation of a new nucleating substance which can induce the nucleation of other polymer chains.56 According to the POM results illustrated in Fig. 13, it is observed that the pristine OA-a and dehydration-transformed OA-a almost do not dissolve in the PLLA melt. What's more, it has been identified that no chemical reaction occurs between OA-m and poly(3-hydroxybutyrate), a bacterial polyester which has similar chain structure to PLLA.43 Therefore, the nucleation of PLLA/OA blends should not be caused by chemical nucleation.
In the case of epitaxial nucleation, the existence of crystallographic relationships and the ability of forming favorable interactions between polymer and NA are the two key factors for the epitaxial growth of polymer chains on the NA surface.19 Taking into account the PLLA spherulite growing along the a-axis of unit cell,57 it thus appears that the b-axis and/or c-axis of PLLA may have the structural relation with OA unit cell. In view of the lattice parameters of PLLA α-form crystal: a = 10.78 Å, b = 6.04 Å, c = 28.73 Å, α = β = γ = 90°,55 indeed, a good matching can be found between the b-axis of PLLA (6.04 Å) and the b-axis of OA-a (5.964 Å) with a misfit of only 1.3%, suggesting a possible crystallographic relationship of [010]PLLA//[010]OA. On the other hand, earlier studies have reported that uracil and cyanuric acid can also act as efficient NAs for PLLA.26,35 Judging from the same groups of the three NAs as shown in Scheme 1, it can be deduced that the potential hydrogen bond interaction between the carbonyl groups in PLLA chains and the imino groups (in keto form) or the hydroxyl groups (in enol form) of OA molecules could stabilize the epitaxial growth process. Based on the above analysis, epitaxial nucleation is likely responsible for the accelerated effects of OA-a and OA-m on the crystallization of PLLA, but more in-depth research is still required to understand the exact mechanism.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7ra02617j |
This journal is © The Royal Society of Chemistry 2017 |