- Email: [email protected]
Near-infrared (NIR) imaging analysis of polylactic acid (PLA) nanocomposite by multiple-perturbation two-dimensional (2D) correlation spectroscopy
Journal of Molecular Structure 1069 (2014) 171–175
Contents lists available at ScienceDirect
Journal of Molecular Structure journal homepage: www.elsevier.com/locate/molstruc
Near-infrared (NIR) imaging analysis of polylactic acid (PLA) nanocomposite by multiple-perturbation two-dimensional (2D) correlation spectroscopy Hideyuki Shinzawa a,⇑, Takurou N. Murakami b, Masakazu Nishida a, Wataru Kanematsu a, Isao Noda c a b c
Research Institute of Instrumentation Frontier, National Institute of Advanced Industrial Science and Technology (AIST), Nagoya 463-8560, Japan Research Institute for Innovation in Sustainable Chemistry, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba 305-8565, Japan Department of Materials Science and Engineering, University of Delaware, Newark, DE 19716, USA
h i g h l i g h t s Two-dimensional (2D) correlation analysis for spectroscopic imaging data. UV-induced variation of polylactic acid including nanoparticles. Identification of similar or dissimilar intensity variation on image plane.
a r t i c l e
i n f o
Article history: Received 25 December 2013 Received in revised form 11 March 2014 Accepted 11 March 2014 Available online 18 March 2014 Keywords: Multiple-perturbation two-dimensional (2D) correlation spectroscopy Near-infrared (NIR) imaging Polylactic acid (PLA) Nanocomposite
a b s t r a c t Multiple-perturbation two-dimensional (2D) correlation spectroscopy was applied to sets of near-infrared (NIR) imaging data of polylactic acid (PLA) nanocomposite samples undergoing UV degradation. Incorporation of clay nanoparticles substantially lowers the surface free energy barrier for the nucleation of PLA and eventually increases the frequency of the spontaneous nucleation of PLA crystals. Thus, when exposed to external stimuli such as UV light, PLA nanocomposite may show different structure alternation depending on the clay dispersion. Multiple-perturbation 2D correlation analysis of the PLA nanocomposite samples revealed different spatial variation between crystalline and amorphous structure of PLA, and the phenomenon especially becomes acute in the region where the clay particles are coagulated. The incorporation of the clay leads to the cleavage-induced crystallization of PLA when the sample is subjected to the UV light. The additional development of the ordered crystalline structure then works favorably to restrict the initial degradation of the polymer, providing the delay in the weight loss of the PLA. Ó 2014 Elsevier B.V. All rights reserved.
Introduction This paper extends our work on the spectroscopic imaging data analysis based on multiple-perturbation two-dimensional (2D) correlation spectroscopy [1–3]. We have recently reported several novel concepts on the multiple-perturbation two-dimensional (2D) correlation spectroscopy which can handle more than two perturbations in the correlation analysis [1–3]. By using multiple-perturbation 2D correlation technique, it becomes possible to effectively sort out key information underlying spectroscopic imaging data. For example, in a separated article, we demonstrated the central idea of imaging data analysis based on multiple-perturbation 2D correlation spectroscopy with a simple example of ⇑ Corresponding author. Tel.: +81 527367563. E-mail address: [email protected] (H. Shinzawa). http://dx.doi.org/10.1016/j.molstruc.2014.03.014 0022-2860/Ó 2014 Elsevier B.V. All rights reserved.
Raman imaging data of polymer blends [4]. The chemically meaningful variation of spectral feature depending on two spatial variables (e.g., x- and y-coordinates) was readily elucidated as a form of 2D correlation spectra. In this communication, we will explore a practical case of imaging data analysis based on the multiple-perturbation 2D correlation technique. An examples is provided with near-infrared (NIR) imaging data of polylactic acid (PLA) nanocomposite films undergoing UV-induced variation to show how the multiple-perturbation 2D correlation technique can be utilized in analyzing the real-world material. Poly(lactic acid) (PLA) is a novel class of bio-based biodegradable plastics that can be produced from renewable feed stocks without directly depending on petroleum [5,6]. The inclusion of nanoparticles in PLA matrix often induces molecular-level interaction with polymer matrix and eventually offers improvement of physical or chemical property of the polymer [7,8]. Such polymer
172
H. Shinzawa et al. / Journal of Molecular Structure 1069 (2014) 171–175
nanocomposites are generally formed by the addition of a small amount of nanoclay dispersion. Typical form of the nanocomposite is intercalated nanocomposite, in which the unit cells of clay structure are expanded by the insertion of polymer into the interlayer spacing, while the periodicity of the clay crystal structure is maintained [9,10]. The development of the intercalated structure often brings additional variation of supermolecular structure of PLA. For example, the clay substantially lowers the surface free energy barrier for nucleation and eventually increases the frequency of the spontaneous nucleation of the PLA crystals. When exposed to external stimuli, such as UV light, PLA noacomposite may show different structure variation (e.g., degradation) depending on the clay dispersion. Thus, the analysis of the NIR spectra of PLA nanocomposite may reveal effects of the clay on the UV-induced variation, which in turn brings pertinent background information on polymer deformation systems in terms of molecular level structure. Methods Doubly 2D correlation spectra We assume that a set of spectra A(v, p, q) of a system under multiple perturbations are collected as a function of an appropriate spectral variable, like wavenumber v. The two additional variables, p = 1, 2, . . ., P and q = 1, 2, . . ., Q, represent different perturbation variables, e. g., x- and y-axis coordinates, respectively. Multipleperturbation 2D correlation spectra can be derived from the three way data matrix A(v, p, q) by lumping two perturbation variables p and q [4]. Collective 2D correlation spectra are derived directly from the three way data matrix A(v, p, q) by lumping two perturbation variables p and q together. Reference and dynamic spectra over p and q are described as
Aðv Þ ¼
Q P X 1 X Aðv ; p; qÞ PQ p¼1 q¼1
e v ; p; qÞ ¼ Aðv ; p; qÞ Aðv Þ Að
ð1Þ
ð2Þ
Doubly synchronous and asynchronous correlation spectra for p and q planes can be described as
Upq ðv 1 ; v 2 Þ ¼
Q P X 1 X e v 1 ; p; qÞ Að e v 2 ; p; qÞ Að PQ 1 p¼1 q¼1
ð3Þ
Wpq ðv 1 ; v 2 Þ ¼
Q P X 1 X e v 1 ; p; qÞ A e – ðv 2 ; p; qÞ Að pq PQ 1 p¼1 q¼1
ð4Þ
e – ðv 2 ; p; qÞ represents the Hilbert–Noda transformation where A pq given by Q P X X e – ðv 2 ; p; qÞ ¼ e v 2 ; r; sÞ Npr N qs Að A pq r¼1
Npr ¼
0
ð5Þ
s¼1
for p ¼ r
1=ðr pÞp otherwise
constructed with v1 and v2. The determination of the sequential order is generally less significant in actual practice of the 2D correlation analysis of spectroscopic imaging data. There are several options in the selection of reference point on the microimage. If the reference point is switched to the point on the opposite side of the image, the sign of asynchronous correlation becomes opposite providing the different sequential order of the events. It is thus convenient to estimate the disrelation spectrum as an approximation for the doubly asynchronous spectrum by circumventing the Hilbert–Noda transformation, which is computationally somewhat demanding [1,11–13].
K2pq ðv 1 ; v 2 Þ ¼ Upq ðv 1 ; v 1 ÞUpq ðv 2 ; v 2 Þ U2pq ðv 1 ; v 2 Þ
ð7Þ
The disrelation intensity can be seen as a portion of the total joint variance of signal fluctuations measured at v1 and v2 that is not accounted for by covariance. In other words, it corresponds to the disvariance component of the total variance, where fluctuations are not coincidental with each other but occur separately along with perturbation directions, i.e., asynchronousity [1,13,14]. The intensity of a disrelation correlation spectrum K2pq ðv 1 ; v 2 Þ can be used as an index to estimate dissimilar change of the patterns on the images constructed with v1 and v2, while it provides only positive value. Experimental procedure Material A commercial PLA (Mn = 11,000, Mw = 65,000, Tm = 170 °C, Tg = 70 °C) supplied by Mitsui Chemicals Company and organically modified clay (S-BEN WÒ, Hojun, Aichi, Japan) were used. The PLA and clay were thoroughly dried in vacuum at 80 °C overnight before the manipulation. The PLA pellets of 1 g and the clay of 3 wt% were dissolved into 100 ml chloroform and sonicated with an ultrasonic homogenizer USS-1 (Nihonseki Kaisha Ltd., Tokyo, Japan) for 5 min. The solutions were then dried in vacuum at 30 °C overnight to remove the chloroform. Total 0.5 g nanocomposite samples were first hot-pressed at 190 °C for 10 min under 5 MPa pressure and quenched in the ice water. The sample was finally aged at 25 °C for 12 h. The thicknesses of the films were approximately 80 lm. A part of the PLA film was exposed to UV light with a UV-irradiator for 30 h. In our geometry, the UV light intensity was approximately 2.0 mW/cm2. Change in the haze induced by the UV irradiation was measured by a Haze meter NDH5000 (Nippon Denshoku Industries Co., Ltd., Tokyo, Japan). NIR spectra Sets of NIR imaging spectra of the PLA films were collected in transmittance mode with a PerkinElmer Spotlight 400 by coadding 32 scans. NIR spectra were collected approximately over the 500 400 lm region with a spatial resolution 6 lm per pixel. Visible images of the samples are illustrated in Fig. 1.
ð6Þ
In the spatial 2D correlation analysis, the intensity of a synchronous 2D correlation spectrum Upq(v1, v2) represents the simultaneous or coincidental changes of spectral intensity variations at v1 and v2 caused by the increase in distance from a reference point on image [4,12,13]. On the other hand, the intensity of an asynchronous spectrum Wpq(v1, v2) indicates the sequential or successive changes of spectral intensities at v1 and v2, caused by the change of the spatial coordinate [4,11–13]. In other words, the doubly correlated 2D correlation spectra derived from imaging data reveal the difference in the change of pattern on the images
Results and discussions Sample opacity Fig. 2 shows the opacity of the PLA films measured by the haze meter. The obvious change in the opacity of the samples caused by the UV irradiation can be observed. Measuring the haze involves separating scattered light from directly transmitted light. For example, the transmitted light which deviates more than 2.5° from the incident beam by forward scattering is considered to be haze. Indeed, such phenomenon is characteristic of semicrystalline
H. Shinzawa et al. / Journal of Molecular Structure 1069 (2014) 171–175
173
Fig. 3. Average spectra of PLA-nanocomposite samples.
Fig. 1. Visible images of PLA nanocomposite samples subjected to UV-irradiation for (A) 0 and (B) 30 h.
Fig. 2. Change in opacity of PLA nanocomposite samples.
polymers, where small crystals scatter light. The change in the haze may indicate the possible compositional change of PLA, i.e., cleavage-induced crystallization stemming from the scissoring of the polymer chain [15]. Thus, detailed analysis of the NIR spectra of specimens becomes important to derive pertinent information on the UV-induced variation of the molecular structure. NIR spectra Average spectra of the PLA-nanocomposite samples are shown in Fig. 3. A peak observed around 5810 cm1 is assignable to antisymmetric CH2 stretching band reflecting the quantity of crystalline structure of PLA [16]. It is noted that the raw spectra show a downward shift with irradiation time. Such a decrease in the apparent spectral intensity is probably explained as the weight loss of the PLA caused by the hydrolysis [17,18]. Fig. 4 represents twodimensional images constructed by the spectral intensity at 5810 cm1 of the samples subjected to the UV-irradiation for (A) 0 and (B) 30 h, respectively. Note that the image is pseudocolored with a common color scale. Although clearly distinguishable lumps
Fig. 4. Pseudocolored images constructed by spectral intensity at 5810 cm1 of samples subjected to UV-irradiation for (A) 0 and (B) 30 h.
of massively aggregated clay particles can be seen in Fig. 1(A), no significant traces are identified in Fig. 4(A). On the other hand, entire image plane of Fig. 4(B) becomes darker than that of Fig. 4(A), mostly reflecting the decrease in the quantity of the PLA caused by the weight loss. It is also noted the relatively greater intensities can be observed in the regions where the clay particles are massively aggregated. Such pattern may suggest possible interaction between the PLA and clay induced by the UV-irradiation. The practical utility of the this type of visualization will be found in the area of identifying distribution of the components by virtually exploring the image as shown in Fig. 4. On the other hand, 2D correlation analysis provides additional opportunity to probe even more detailed information on different trend in intensity variation observed on image. (A) Doubly synchronous and (B)
174
H. Shinzawa et al. / Journal of Molecular Structure 1069 (2014) 171–175
Fig. 6 represents (A) doubly synchronous and (B) doubly disrelation spectra calculated from the UV-irradiated sample. The doubly synchronous correlation spectrum results in the similar pattern to that of Fig. 5(A), indicating simultaneous change of all the PLA bands. On the other hand, the emergence of the disrelation peaks in Fig. 6(B) become important. They appear only if the patterns of the spectral intensity changes observed at different spectral variables are not identical. This feature becomes especially useful in sorting out subtle and complex changes in spectral feature and establishing unambiguous assignments of bands. A cross peak at the coordinate (5810, 5940) suggests that different trends in the change in the spectral intensity at these spectral variables. In other words, when we move from the reference point to the other side of the image, the spectral intensities at these wavenumbers vary in different manners. Such different trends in the change of pattern reveal the generation of species originating from different spatial dependences. In fact, the peak observed around 5940 cm1 is assigned to the antisymmetric CH2 stretching of the amorphous component [16]. Consequently, the development of the apparent disrelation peak reveals that the crystalline and amorphous structures of PLA essentially undergoes different variation when exposed to the UV light. It is important to point out here that this feature of 2D correlation spectroscopy especially becomes important to identify pertinent variation of spectral intensities and establish unambiguous assignments. It will provide additional insight into imaging data which is not derived by the conventional image construction based on spectral intensity of single wavenumber. Now that we have identified the bands having different spatial dependences, additional information on the spatial distribution
Fig. 5. (A) Doubly synchronous and (B) doubly disrelation correlation spectra derived from sample without UV-irradiation.
doubly disrelation correlation spectra derived from the sample without UV-irradiation are shown in Fig. 5. The bottom left corners in Fig. 1 were used as reference point for the calculation. Note that the doubly disrelation spectrum K2pq ðv 1 ; v 2 Þ represents the squared disrelation intensity defined by Eq. (7), which provides only positive value. Doubly correlated 2D correlation spectra represent similar or dissimilar spectral intensity variation induced by increase in geometric distance from a reference point on the image [4]. For example, entire spectral plane of the doubly synchronous correlation spectrum is covered with positive correlation peaks. This result can be interpreted to mean that the correlation peaks in this spectrum all arise from bands associated with PMMA, i.e., they are all varying together in the same direction. Such coincidental changes mostly suggest the variation of sample thickness or PLA content caused by the inclusion of the clay. On the other hand, the disrelation spectrum K2pq ðv 1 ; v 2 Þ comprises with many small correlation peaks over the entire plane, mostly indicating the absence of significantly dissimilar intensity variations among the wavenumbers. For example, when we move from the reference point to the other side of the image, we encounter the change in the spectral intensity of PLA bands varying simultaneously without any delay. In general, the incorporation of the clay often brings additional development of crystalline structure of PLA due to the surface interaction between PLA and clay [10,18]. This compositional change causes in the different intensity variation of crystalline and amorphous band of PLA and it then leads to the development of asynchronous correlation peaks [10,18]. The absence of significant correlation peak in our case may be related to the fact that the sample was quickly quenched without annealing.
Fig. 6. (A) Doubly synchronous and (B) doubly disrelation correlation spectra calculated from UV-irradiated sample.
H. Shinzawa et al. / Journal of Molecular Structure 1069 (2014) 171–175
175
amorphous component tends to undergo structural variation when exposed to UV-irradiation [15]. Consequently, additional development of crystalline structure by the interaction between the PLA and clay potentially works favorably to restrict the initial degradation of the polymer. Conclusion An application of spectroscopic imaging data analysis by multiple-perturbation 2D correlation technique was present in this study. Similar or dissimilar variation between the spectral intensities measured at two different wavenumbers was clearly identified when the PLA-nanocomposite sample was subjected to the UVirradiation. For example, the significant asynchronous correlation intensity at the coordinate (5810, 5940) indicated that different trends in the change of the spectral intensity of these bands, revealing the generation of the species originating from different spatial dependences. This feature becomes especially useful in sorting out subtle and complex changes in spectral feature underlying the image data and establishing unambiguous assignments of bands. Further analysis of the NIR spectra provided a possible mechanistic picture of the system. Interaction between the PLA and clay causes additional crystallization of PLA when the PLA undergoes the molecular scission of the polymer chain by the UV-irradiation. Such cleavage-induced crystallization then brings substantial delay in the weight loss of the PLA. Accordingly, conclusive evidence on the accelerated hydroids of PLA nonocomposite is provided by the analysis with an aid of the doubly correlated 2D correlation spectra. Acknowledgements Fig. 7. Images of samples subjected to UV-irradiation for (A) 0 and (B) 30 h constructed with ratio of spectral intensities at 5810 and 5940 cm1.
can be drawn in a very intuitively understandable manner. For example, Fig. 7 represents the images of the samples subjected to the UV-irradiation for (A) 0 and (B) 30 h constructed with ratio of the spectral intensities at 5810 and 5940 cm1, i.e., kcrystaline/kamorphous. The value of kcrystaline/kamorphous can be seen as an index representing relative quantity of the crystalline and amorphous components of PLA. In Fig. 7(A), it is noted that no significant feature can be observed. The absence of any specific pattern on the image reflects the simultaneous change in the spectral intensities of the crystalline and amorphous bands. On the other hand, notable pattern is now clearly visible in Fig. 7(B). The development of such pattern suggests increase in the relative content of the crystalline structure of PLA. Namely, the decrease in the spectral intensity at crystalline band is delayed compared with that at amorphous band and the alternation is especially acute in the region where the clay particles are aggregated. All the results put together, it provides a possible picture of the system. UV-irradiation substantially causes the molecular scission of the polymer chain. The scision can lead to the cleavage-induced crystallization of PLA, which predominantly occurs around the interfaces between the PLA and clay. Then the additional development of the crystalline structure brings the delay in the weight loss of the PLA. This result is consistent with the view that less ordered
TNM would like to thank New Energy and Industrial Technology Development Organization (NEDO). References [1] H. Shinzawa, S. Morita, K. Awa, M. Okada, I. Noda, Y. Ozaki, H. Sato, Appl. Spectrosc. 63 (2009) 501. [2] H. Shinzawa, K. Awa, I. Noda, Y. Ozaki, Appl. Spectrosc. 67 (2) (2013) 163. [3] H. Shinzawa, T. Genkawa, W. Kanematsu, Vib. Spectrosc. 1028 (2012) 164. [4] H. Shinzawa, K. Hashimoto, H. Sato, W. Kanematsu, I. Noda, J. Mol. Struct (2014). [5] G. Scott, D. Gilead, Biodegradable polymers, Principles and Applications, Chapman & Hall, London, 1995. [6] D. Garlotta, J. Polym. Environ. 9 (2001) 63. [7] S.S. Ray, P. Maiti, M. Okamoto, K. Yamada, K. Ueda, Macromolecules 35 (2002) 2104. [8] S.S. Ray, K. Yamada, M. Okamoto, A. Ogami, K. Ueda, Chem. Mater. 15 (2003) 1456. [9] D.R. Paul, L.M. Robeson, Polymer 48 (2008) 3187. [10] H. Shinzawa, M. Nishida, W. Kanematsu, T. Tanaka, K. Suzuki, I. Noda, Analyst 137 (2012) 1913. [11] I. Noda, Appl. Spectrosc. 47 (1993) 1329. [12] I. Noda, Appl. Spectrosc. 54 (2000) 994. [13] I. Noda, Y. Ozaki, Two-dimensional Correlation Spectroscopy, Willey & Sons, Chichester, West Sussex, 2004. [14] T. Isaksson, Vib. Spectrosc. 36 (2004) 25. [15] H. Shinzawa, M. Nishida, T. Tanaka, W. Kanematsu, Appl. Spectrosc. 66 (2012) 470. [16] H. Shinzawa, M. Nishida, A. Tsuge, D. Ishikawa, Y. Ozaki, S. Morita, W. Kanematsu, Appl. Spectrosc. 6 (2014) 371. [17] H. Tsuji, Y. Kawashima, H. Takikawa, S. Tanaka, Polymer 48 (2007) 4213. [18] Y. Ando, H. Sato, H. Shinzawa, M. Okamoto, I. Noda, Y. Ozaki, Vib. Spectrosc. 60 (2012) 158.
Contents lists available at ScienceDirect
Journal of Molecular Structure journal homepage: www.elsevier.com/locate/molstruc
Near-infrared (NIR) imaging analysis of polylactic acid (PLA) nanocomposite by multiple-perturbation two-dimensional (2D) correlation spectroscopy Hideyuki Shinzawa a,⇑, Takurou N. Murakami b, Masakazu Nishida a, Wataru Kanematsu a, Isao Noda c a b c
Research Institute of Instrumentation Frontier, National Institute of Advanced Industrial Science and Technology (AIST), Nagoya 463-8560, Japan Research Institute for Innovation in Sustainable Chemistry, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba 305-8565, Japan Department of Materials Science and Engineering, University of Delaware, Newark, DE 19716, USA
h i g h l i g h t s Two-dimensional (2D) correlation analysis for spectroscopic imaging data. UV-induced variation of polylactic acid including nanoparticles. Identification of similar or dissimilar intensity variation on image plane.
a r t i c l e
i n f o
Article history: Received 25 December 2013 Received in revised form 11 March 2014 Accepted 11 March 2014 Available online 18 March 2014 Keywords: Multiple-perturbation two-dimensional (2D) correlation spectroscopy Near-infrared (NIR) imaging Polylactic acid (PLA) Nanocomposite
a b s t r a c t Multiple-perturbation two-dimensional (2D) correlation spectroscopy was applied to sets of near-infrared (NIR) imaging data of polylactic acid (PLA) nanocomposite samples undergoing UV degradation. Incorporation of clay nanoparticles substantially lowers the surface free energy barrier for the nucleation of PLA and eventually increases the frequency of the spontaneous nucleation of PLA crystals. Thus, when exposed to external stimuli such as UV light, PLA nanocomposite may show different structure alternation depending on the clay dispersion. Multiple-perturbation 2D correlation analysis of the PLA nanocomposite samples revealed different spatial variation between crystalline and amorphous structure of PLA, and the phenomenon especially becomes acute in the region where the clay particles are coagulated. The incorporation of the clay leads to the cleavage-induced crystallization of PLA when the sample is subjected to the UV light. The additional development of the ordered crystalline structure then works favorably to restrict the initial degradation of the polymer, providing the delay in the weight loss of the PLA. Ó 2014 Elsevier B.V. All rights reserved.
Introduction This paper extends our work on the spectroscopic imaging data analysis based on multiple-perturbation two-dimensional (2D) correlation spectroscopy [1–3]. We have recently reported several novel concepts on the multiple-perturbation two-dimensional (2D) correlation spectroscopy which can handle more than two perturbations in the correlation analysis [1–3]. By using multiple-perturbation 2D correlation technique, it becomes possible to effectively sort out key information underlying spectroscopic imaging data. For example, in a separated article, we demonstrated the central idea of imaging data analysis based on multiple-perturbation 2D correlation spectroscopy with a simple example of ⇑ Corresponding author. Tel.: +81 527367563. E-mail address: [email protected] (H. Shinzawa). http://dx.doi.org/10.1016/j.molstruc.2014.03.014 0022-2860/Ó 2014 Elsevier B.V. All rights reserved.
Raman imaging data of polymer blends [4]. The chemically meaningful variation of spectral feature depending on two spatial variables (e.g., x- and y-coordinates) was readily elucidated as a form of 2D correlation spectra. In this communication, we will explore a practical case of imaging data analysis based on the multiple-perturbation 2D correlation technique. An examples is provided with near-infrared (NIR) imaging data of polylactic acid (PLA) nanocomposite films undergoing UV-induced variation to show how the multiple-perturbation 2D correlation technique can be utilized in analyzing the real-world material. Poly(lactic acid) (PLA) is a novel class of bio-based biodegradable plastics that can be produced from renewable feed stocks without directly depending on petroleum [5,6]. The inclusion of nanoparticles in PLA matrix often induces molecular-level interaction with polymer matrix and eventually offers improvement of physical or chemical property of the polymer [7,8]. Such polymer
172
H. Shinzawa et al. / Journal of Molecular Structure 1069 (2014) 171–175
nanocomposites are generally formed by the addition of a small amount of nanoclay dispersion. Typical form of the nanocomposite is intercalated nanocomposite, in which the unit cells of clay structure are expanded by the insertion of polymer into the interlayer spacing, while the periodicity of the clay crystal structure is maintained [9,10]. The development of the intercalated structure often brings additional variation of supermolecular structure of PLA. For example, the clay substantially lowers the surface free energy barrier for nucleation and eventually increases the frequency of the spontaneous nucleation of the PLA crystals. When exposed to external stimuli, such as UV light, PLA noacomposite may show different structure variation (e.g., degradation) depending on the clay dispersion. Thus, the analysis of the NIR spectra of PLA nanocomposite may reveal effects of the clay on the UV-induced variation, which in turn brings pertinent background information on polymer deformation systems in terms of molecular level structure. Methods Doubly 2D correlation spectra We assume that a set of spectra A(v, p, q) of a system under multiple perturbations are collected as a function of an appropriate spectral variable, like wavenumber v. The two additional variables, p = 1, 2, . . ., P and q = 1, 2, . . ., Q, represent different perturbation variables, e. g., x- and y-axis coordinates, respectively. Multipleperturbation 2D correlation spectra can be derived from the three way data matrix A(v, p, q) by lumping two perturbation variables p and q [4]. Collective 2D correlation spectra are derived directly from the three way data matrix A(v, p, q) by lumping two perturbation variables p and q together. Reference and dynamic spectra over p and q are described as
Aðv Þ ¼
Q P X 1 X Aðv ; p; qÞ PQ p¼1 q¼1
e v ; p; qÞ ¼ Aðv ; p; qÞ Aðv Þ Að
ð1Þ
ð2Þ
Doubly synchronous and asynchronous correlation spectra for p and q planes can be described as
Upq ðv 1 ; v 2 Þ ¼
Q P X 1 X e v 1 ; p; qÞ Að e v 2 ; p; qÞ Að PQ 1 p¼1 q¼1
ð3Þ
Wpq ðv 1 ; v 2 Þ ¼
Q P X 1 X e v 1 ; p; qÞ A e – ðv 2 ; p; qÞ Að pq PQ 1 p¼1 q¼1
ð4Þ
e – ðv 2 ; p; qÞ represents the Hilbert–Noda transformation where A pq given by Q P X X e – ðv 2 ; p; qÞ ¼ e v 2 ; r; sÞ Npr N qs Að A pq r¼1
Npr ¼
0
ð5Þ
s¼1
for p ¼ r
1=ðr pÞp otherwise
constructed with v1 and v2. The determination of the sequential order is generally less significant in actual practice of the 2D correlation analysis of spectroscopic imaging data. There are several options in the selection of reference point on the microimage. If the reference point is switched to the point on the opposite side of the image, the sign of asynchronous correlation becomes opposite providing the different sequential order of the events. It is thus convenient to estimate the disrelation spectrum as an approximation for the doubly asynchronous spectrum by circumventing the Hilbert–Noda transformation, which is computationally somewhat demanding [1,11–13].
K2pq ðv 1 ; v 2 Þ ¼ Upq ðv 1 ; v 1 ÞUpq ðv 2 ; v 2 Þ U2pq ðv 1 ; v 2 Þ
ð7Þ
The disrelation intensity can be seen as a portion of the total joint variance of signal fluctuations measured at v1 and v2 that is not accounted for by covariance. In other words, it corresponds to the disvariance component of the total variance, where fluctuations are not coincidental with each other but occur separately along with perturbation directions, i.e., asynchronousity [1,13,14]. The intensity of a disrelation correlation spectrum K2pq ðv 1 ; v 2 Þ can be used as an index to estimate dissimilar change of the patterns on the images constructed with v1 and v2, while it provides only positive value. Experimental procedure Material A commercial PLA (Mn = 11,000, Mw = 65,000, Tm = 170 °C, Tg = 70 °C) supplied by Mitsui Chemicals Company and organically modified clay (S-BEN WÒ, Hojun, Aichi, Japan) were used. The PLA and clay were thoroughly dried in vacuum at 80 °C overnight before the manipulation. The PLA pellets of 1 g and the clay of 3 wt% were dissolved into 100 ml chloroform and sonicated with an ultrasonic homogenizer USS-1 (Nihonseki Kaisha Ltd., Tokyo, Japan) for 5 min. The solutions were then dried in vacuum at 30 °C overnight to remove the chloroform. Total 0.5 g nanocomposite samples were first hot-pressed at 190 °C for 10 min under 5 MPa pressure and quenched in the ice water. The sample was finally aged at 25 °C for 12 h. The thicknesses of the films were approximately 80 lm. A part of the PLA film was exposed to UV light with a UV-irradiator for 30 h. In our geometry, the UV light intensity was approximately 2.0 mW/cm2. Change in the haze induced by the UV irradiation was measured by a Haze meter NDH5000 (Nippon Denshoku Industries Co., Ltd., Tokyo, Japan). NIR spectra Sets of NIR imaging spectra of the PLA films were collected in transmittance mode with a PerkinElmer Spotlight 400 by coadding 32 scans. NIR spectra were collected approximately over the 500 400 lm region with a spatial resolution 6 lm per pixel. Visible images of the samples are illustrated in Fig. 1.
ð6Þ
In the spatial 2D correlation analysis, the intensity of a synchronous 2D correlation spectrum Upq(v1, v2) represents the simultaneous or coincidental changes of spectral intensity variations at v1 and v2 caused by the increase in distance from a reference point on image [4,12,13]. On the other hand, the intensity of an asynchronous spectrum Wpq(v1, v2) indicates the sequential or successive changes of spectral intensities at v1 and v2, caused by the change of the spatial coordinate [4,11–13]. In other words, the doubly correlated 2D correlation spectra derived from imaging data reveal the difference in the change of pattern on the images
Results and discussions Sample opacity Fig. 2 shows the opacity of the PLA films measured by the haze meter. The obvious change in the opacity of the samples caused by the UV irradiation can be observed. Measuring the haze involves separating scattered light from directly transmitted light. For example, the transmitted light which deviates more than 2.5° from the incident beam by forward scattering is considered to be haze. Indeed, such phenomenon is characteristic of semicrystalline
H. Shinzawa et al. / Journal of Molecular Structure 1069 (2014) 171–175
173
Fig. 3. Average spectra of PLA-nanocomposite samples.
Fig. 1. Visible images of PLA nanocomposite samples subjected to UV-irradiation for (A) 0 and (B) 30 h.
Fig. 2. Change in opacity of PLA nanocomposite samples.
polymers, where small crystals scatter light. The change in the haze may indicate the possible compositional change of PLA, i.e., cleavage-induced crystallization stemming from the scissoring of the polymer chain [15]. Thus, detailed analysis of the NIR spectra of specimens becomes important to derive pertinent information on the UV-induced variation of the molecular structure. NIR spectra Average spectra of the PLA-nanocomposite samples are shown in Fig. 3. A peak observed around 5810 cm1 is assignable to antisymmetric CH2 stretching band reflecting the quantity of crystalline structure of PLA [16]. It is noted that the raw spectra show a downward shift with irradiation time. Such a decrease in the apparent spectral intensity is probably explained as the weight loss of the PLA caused by the hydrolysis [17,18]. Fig. 4 represents twodimensional images constructed by the spectral intensity at 5810 cm1 of the samples subjected to the UV-irradiation for (A) 0 and (B) 30 h, respectively. Note that the image is pseudocolored with a common color scale. Although clearly distinguishable lumps
Fig. 4. Pseudocolored images constructed by spectral intensity at 5810 cm1 of samples subjected to UV-irradiation for (A) 0 and (B) 30 h.
of massively aggregated clay particles can be seen in Fig. 1(A), no significant traces are identified in Fig. 4(A). On the other hand, entire image plane of Fig. 4(B) becomes darker than that of Fig. 4(A), mostly reflecting the decrease in the quantity of the PLA caused by the weight loss. It is also noted the relatively greater intensities can be observed in the regions where the clay particles are massively aggregated. Such pattern may suggest possible interaction between the PLA and clay induced by the UV-irradiation. The practical utility of the this type of visualization will be found in the area of identifying distribution of the components by virtually exploring the image as shown in Fig. 4. On the other hand, 2D correlation analysis provides additional opportunity to probe even more detailed information on different trend in intensity variation observed on image. (A) Doubly synchronous and (B)
174
H. Shinzawa et al. / Journal of Molecular Structure 1069 (2014) 171–175
Fig. 6 represents (A) doubly synchronous and (B) doubly disrelation spectra calculated from the UV-irradiated sample. The doubly synchronous correlation spectrum results in the similar pattern to that of Fig. 5(A), indicating simultaneous change of all the PLA bands. On the other hand, the emergence of the disrelation peaks in Fig. 6(B) become important. They appear only if the patterns of the spectral intensity changes observed at different spectral variables are not identical. This feature becomes especially useful in sorting out subtle and complex changes in spectral feature and establishing unambiguous assignments of bands. A cross peak at the coordinate (5810, 5940) suggests that different trends in the change in the spectral intensity at these spectral variables. In other words, when we move from the reference point to the other side of the image, the spectral intensities at these wavenumbers vary in different manners. Such different trends in the change of pattern reveal the generation of species originating from different spatial dependences. In fact, the peak observed around 5940 cm1 is assigned to the antisymmetric CH2 stretching of the amorphous component [16]. Consequently, the development of the apparent disrelation peak reveals that the crystalline and amorphous structures of PLA essentially undergoes different variation when exposed to the UV light. It is important to point out here that this feature of 2D correlation spectroscopy especially becomes important to identify pertinent variation of spectral intensities and establish unambiguous assignments. It will provide additional insight into imaging data which is not derived by the conventional image construction based on spectral intensity of single wavenumber. Now that we have identified the bands having different spatial dependences, additional information on the spatial distribution
Fig. 5. (A) Doubly synchronous and (B) doubly disrelation correlation spectra derived from sample without UV-irradiation.
doubly disrelation correlation spectra derived from the sample without UV-irradiation are shown in Fig. 5. The bottom left corners in Fig. 1 were used as reference point for the calculation. Note that the doubly disrelation spectrum K2pq ðv 1 ; v 2 Þ represents the squared disrelation intensity defined by Eq. (7), which provides only positive value. Doubly correlated 2D correlation spectra represent similar or dissimilar spectral intensity variation induced by increase in geometric distance from a reference point on the image [4]. For example, entire spectral plane of the doubly synchronous correlation spectrum is covered with positive correlation peaks. This result can be interpreted to mean that the correlation peaks in this spectrum all arise from bands associated with PMMA, i.e., they are all varying together in the same direction. Such coincidental changes mostly suggest the variation of sample thickness or PLA content caused by the inclusion of the clay. On the other hand, the disrelation spectrum K2pq ðv 1 ; v 2 Þ comprises with many small correlation peaks over the entire plane, mostly indicating the absence of significantly dissimilar intensity variations among the wavenumbers. For example, when we move from the reference point to the other side of the image, we encounter the change in the spectral intensity of PLA bands varying simultaneously without any delay. In general, the incorporation of the clay often brings additional development of crystalline structure of PLA due to the surface interaction between PLA and clay [10,18]. This compositional change causes in the different intensity variation of crystalline and amorphous band of PLA and it then leads to the development of asynchronous correlation peaks [10,18]. The absence of significant correlation peak in our case may be related to the fact that the sample was quickly quenched without annealing.
Fig. 6. (A) Doubly synchronous and (B) doubly disrelation correlation spectra calculated from UV-irradiated sample.
H. Shinzawa et al. / Journal of Molecular Structure 1069 (2014) 171–175
175
amorphous component tends to undergo structural variation when exposed to UV-irradiation [15]. Consequently, additional development of crystalline structure by the interaction between the PLA and clay potentially works favorably to restrict the initial degradation of the polymer. Conclusion An application of spectroscopic imaging data analysis by multiple-perturbation 2D correlation technique was present in this study. Similar or dissimilar variation between the spectral intensities measured at two different wavenumbers was clearly identified when the PLA-nanocomposite sample was subjected to the UVirradiation. For example, the significant asynchronous correlation intensity at the coordinate (5810, 5940) indicated that different trends in the change of the spectral intensity of these bands, revealing the generation of the species originating from different spatial dependences. This feature becomes especially useful in sorting out subtle and complex changes in spectral feature underlying the image data and establishing unambiguous assignments of bands. Further analysis of the NIR spectra provided a possible mechanistic picture of the system. Interaction between the PLA and clay causes additional crystallization of PLA when the PLA undergoes the molecular scission of the polymer chain by the UV-irradiation. Such cleavage-induced crystallization then brings substantial delay in the weight loss of the PLA. Accordingly, conclusive evidence on the accelerated hydroids of PLA nonocomposite is provided by the analysis with an aid of the doubly correlated 2D correlation spectra. Acknowledgements Fig. 7. Images of samples subjected to UV-irradiation for (A) 0 and (B) 30 h constructed with ratio of spectral intensities at 5810 and 5940 cm1.
can be drawn in a very intuitively understandable manner. For example, Fig. 7 represents the images of the samples subjected to the UV-irradiation for (A) 0 and (B) 30 h constructed with ratio of the spectral intensities at 5810 and 5940 cm1, i.e., kcrystaline/kamorphous. The value of kcrystaline/kamorphous can be seen as an index representing relative quantity of the crystalline and amorphous components of PLA. In Fig. 7(A), it is noted that no significant feature can be observed. The absence of any specific pattern on the image reflects the simultaneous change in the spectral intensities of the crystalline and amorphous bands. On the other hand, notable pattern is now clearly visible in Fig. 7(B). The development of such pattern suggests increase in the relative content of the crystalline structure of PLA. Namely, the decrease in the spectral intensity at crystalline band is delayed compared with that at amorphous band and the alternation is especially acute in the region where the clay particles are aggregated. All the results put together, it provides a possible picture of the system. UV-irradiation substantially causes the molecular scission of the polymer chain. The scision can lead to the cleavage-induced crystallization of PLA, which predominantly occurs around the interfaces between the PLA and clay. Then the additional development of the crystalline structure brings the delay in the weight loss of the PLA. This result is consistent with the view that less ordered
TNM would like to thank New Energy and Industrial Technology Development Organization (NEDO). References [1] H. Shinzawa, S. Morita, K. Awa, M. Okada, I. Noda, Y. Ozaki, H. Sato, Appl. Spectrosc. 63 (2009) 501. [2] H. Shinzawa, K. Awa, I. Noda, Y. Ozaki, Appl. Spectrosc. 67 (2) (2013) 163. [3] H. Shinzawa, T. Genkawa, W. Kanematsu, Vib. Spectrosc. 1028 (2012) 164. [4] H. Shinzawa, K. Hashimoto, H. Sato, W. Kanematsu, I. Noda, J. Mol. Struct (2014). [5] G. Scott, D. Gilead, Biodegradable polymers, Principles and Applications, Chapman & Hall, London, 1995. [6] D. Garlotta, J. Polym. Environ. 9 (2001) 63. [7] S.S. Ray, P. Maiti, M. Okamoto, K. Yamada, K. Ueda, Macromolecules 35 (2002) 2104. [8] S.S. Ray, K. Yamada, M. Okamoto, A. Ogami, K. Ueda, Chem. Mater. 15 (2003) 1456. [9] D.R. Paul, L.M. Robeson, Polymer 48 (2008) 3187. [10] H. Shinzawa, M. Nishida, W. Kanematsu, T. Tanaka, K. Suzuki, I. Noda, Analyst 137 (2012) 1913. [11] I. Noda, Appl. Spectrosc. 47 (1993) 1329. [12] I. Noda, Appl. Spectrosc. 54 (2000) 994. [13] I. Noda, Y. Ozaki, Two-dimensional Correlation Spectroscopy, Willey & Sons, Chichester, West Sussex, 2004. [14] T. Isaksson, Vib. Spectrosc. 36 (2004) 25. [15] H. Shinzawa, M. Nishida, T. Tanaka, W. Kanematsu, Appl. Spectrosc. 66 (2012) 470. [16] H. Shinzawa, M. Nishida, A. Tsuge, D. Ishikawa, Y. Ozaki, S. Morita, W. Kanematsu, Appl. Spectrosc. 6 (2014) 371. [17] H. Tsuji, Y. Kawashima, H. Takikawa, S. Tanaka, Polymer 48 (2007) 4213. [18] Y. Ando, H. Sato, H. Shinzawa, M. Okamoto, I. Noda, Y. Ozaki, Vib. Spectrosc. 60 (2012) 158.