- Email: [email protected]
orientation states of polymer thin films revealed by polarization-dependent infrared absorption
European Polymer Journal 63 (2015) 247–254
Contents lists available at ScienceDirect
European Polymer Journal journal homepage: www.elsevier.com/locate/europolj
Gradiently varied chain packing/orientation states of polymer thin films revealed by polarization-dependent infrared absorption Xiaolin Lu a,⇑, Yongli Mi b,⇑ a b
State Key Laboratory of Bioelectronics, School of Biological Science & Medical Engineering, Southeast University, Nanjing 210096, Jiangsu Province, PR China Department of Chemical and Biomolecular Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong
a r t i c l e
i n f o
Article history: Received 8 November 2014 Received in revised form 22 December 2014 Accepted 22 December 2014 Available online 30 December 2014 Keywords: RA-FTIR Polarization-dependent Polyacrylamide Thin films
a b s t r a c t Gradiently varied chain packing/orientation states of the polyacrylamide (PAL) thin films spin-coated on the gold (Au) substrates were found via the polarized reflection–absorption Fourier transform infrared spectroscopy (RA-FTIR). As the film thickness increases, the splitted amide I bands provide a direct evidence that the PAL thin films are of a gradiently varied bi-layered structure. In the bottom layer, most of the PAL molecules show random orientation which is induced by the non-favorable interaction from the adjacent Au surface. In the top layer, most of PAL molecules show parallel orientation to the Au surface which is induced by spin coating, evidenced by the enhanced low-frequency splitted amide I band (1658 cm 1) and N–H stretching modes of the amino groups when the light electric field vector is adjusted to be parallel to the Au surface. The observation reported in this study should be of universal significance for polymer thin films on the supported substrates, where the interfacial interaction as well as spin coating could vary the polymer packing/orientation states substantially. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Polymer thin films are a kind of important materials due to their broad applications in coating, adhesive and microelectronic industries [1,2]. The long-term reliability of a polymer thin film to a large extent relies on whether the polymer chains in the film are in the relaxed state or not because the evolution of the chain conformation during usage may have negative effect on its stability. For instance, when the thickness of a polymer thin film decreases down to or lower than certain critical length, i.e. a molecular size in the scale of the mean-square endto-end distance (Ree) or radius of gyration (Rg), chain conformation of the polymer could notably be different from ⇑ Corresponding authors. Tel.: +86 25 83791810 (X. Lu). E-mail addresses: [email protected] (X. Lu), [email protected] (Y. Mi). http://dx.doi.org/10.1016/j.eurpolymj.2014.12.030 0014-3057/Ó 2014 Elsevier Ltd. All rights reserved.
that of its bulk state [3–6]. Therefore the residue stress or the entropic force may evidently exist and gradually deteriorate the property of the polymer thin film [4,6–8]. It is thus of great importance to investigate the chain packing/orientation state and the related dynamic behavior for a polymer thin film. Different techniques have been applied to study this issue in the past, such as fluorescence microscopy [9–12], ellipsometry [13–16], Brillouin light scattering [16–18], X-ray reflectivity [3,19,20] and sum frequency generation vibrational spectroscopy [21–25]. It has generally been recognized that a polymer thin film is a metastable material [4] which has distinct physicochemical properties from those of the bulk polymer. For a polymer thin film on the supported substrate, besides the chain confinement effect, the interaction between the polymer and the substrate, whether attractive or nonattractive, can also have significant effect which can alter
248
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
arrangement of the polymer chains near the polymer/substrate interface. A preliminary report on the supported polymer thin films prepared by spin coating has suggested the interaction from the substrate and spin coating can remarkably alter the chain packing states of the polymer thin films [26]. In this study, complete experimental results were present to explore the chain orientation states by analyzing the vibrational signals of the polyacrylamide (PAL) molecular groups using the polarized reflection– absorption Fourier transform infrared spectroscopy (RAFTIR) with the grazing-angle incidence. A gradiently varied bi-layered structure of the PAL thin films on the Au substrates was confirmed. The result indicates the PAL chains in the bottom layer (adjacent to the Au substrate) and the top layer have undoubtedly different orientational orderings which lead to the polarization-dependent infrared absorption. 2. Experimental section The Au substrates were prepared by depositing a 200nm Au layer on top of the silicon wafers (Silicon polished wafer, N-type, Grinm, Inc.) using the high vacuum evaporator (High Vacuum Evaporator, DV-502A, Denton Inc.). The silicon wafers were pretreated by piranha solution (a mixed solution with 3:7 volume ratio of 30 wt.% H2O2 solution and 98 wt.% H2SO4) to eliminate the possible surface contamination. PAL (Mw = 5–6 106) were purchased from Polysciences, Inc. The mean square end-to-end distance is estimated to be around 102 nm. The PAL thin films were prepared by spin coating PAL solutions on top of the Au substrates (Spin Coater, P6700, Specialty Coating Systems Inc.). Solution concentration and spinning speed were adjusted to obtain the PAL thin films with desired thicknesses. The PAL films with thicknesses of 9.8 nm, 39 nm, 73 nm, 104 nm, 170 nm, 269 nm, 375 nm, and 559 nm were prepared. All the PAL film samples were annealed in a vacuum oven (Shel lab 1410, Shalton Mfg Inc.) at 60 °C for 24 h. Before further characterization, the samples
were put in a desiccator. The Variable-Angle Spectroscopic Ellipsometer (J.A. Woollam Co., Inc.) was used to measure the thicknesses of the PAL thin films on the Au substrates. The RA-FTIR spectroscopy (RA-FTIR, FTS-6000, Bio-Rad Inc.) were applied to capture the polarized infrared spectra of the PAL thin films on the Au substrates. The polarizer was mounted before the infrared light shot at the sample surface. The s- and p-polarized lights with the grazingangle (85°) incidence were used, as shown in Fig. 1. With this grazing-angle incidence, the transition dipole moments of molecular groups parallel and normal to the Au surface can be separately detected simply by using sand p-polarized lights respectively, as demonstrated by the selection rules for RA-FTIR on the reflected metal surfaces [27–29]. The ZnSe polarizer (Tydex J.S. Co., spectral range from 1.5 lm to 14 lm or 6667 cm 1 to 714 cm 1) was used to obtain the light with >95% polarization. The orientation of a certain PAL molecular group can thus be analyzed based on its polarization-dependent infrared absorption. During the infrared experiment, the sample chamber of the infrared spectrometer was vacuumized for at least 1 h before collecting the background and the sample spectra. Otherwise the interference from humidity and carbon dioxide in the sample chamber would lead to the low signal-to-noise ratio. 3. Results and discussion 3.1. Polarization-dependent spectra and splitting of amide I band The normal infrared spectrum for a PAL bulk sample has been reported previously [30]. The amide I mode and the N–H stretching modes from the amide groups (ACONH2) are the strongest modes. Here as shown in Fig. 2, using RA-FTIR, the collected spectra for a series of PAL thin films spin-coated onto the Au substrates show strong characteristics depending on the light polarization and film thickness. Fig. 2 shows the RA-FTIR spectra of the PAL thin film
Fig. 1. Optics of RA-FTIR with a grazing-angle incidence (left panel). For a certain molecular group, with this grazing angle incidence, the infrared transition dipole moment parallel to the Au substrate surface is detected when the light polarization is adjusted to be ‘‘s’’ and the infrared transition dipole moment normal to the Au substrate surface is detected when the light polarization is adjusted to be ‘‘p’’. The monomer unit of PAL is also shown (right panel).
249
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
0.6
Es
9.8 nm 39 nm 73 nm 104 nm 170 nm 269 nm 375 nm 559 nm
NH2
Absorbance
0.4
A
Ep
0.8
B
9.8 nm 39 nm 73 nm 104 nm 170 nm 269 nm 375 nm 559 nm
0.6
Amide I
0.4
Amide I
NH2
0.2
Thickness increase
Thickness increase
0.2
0.0
0.0 3500
3000
2500
Wavenumber
2000
1500
3500
(cm-1)
3000
2500
Wavenumber
2000
1500
(cm-1)
Fig. 2. The s- (Panel A) and p-polarized (Panel B) RA-FTIR spectra of the PAL thin films with different thicknesses spin-coated on the Au surface with the grazing angle (85°) incidence. From the bottom to top, the film thicknesses are 9.8 nm, 39 nm, 73 nm, 104 nm, 170 nm, 269 nm, 375 nm, and 559 nm, respectively.
samples with eight different thicknesses (from 9.8 nm to 559 nm) collected using the s- (Panel A) and p-polarized (Panel B) lights, respectively. Two remarkable features can be read from those spectra. Firstly, as the film thickness increases, the intensities for the N–H stretching modes (3000–3600 cm 1) of the amino groups using the s-polarized light increase more significantly than those using the p-polarized light. Secondly, the splitting of the amide I mode is clearly observed when the film thickness is above 73 nm and such band splitting appears polarization-dependent. Fig. 3 shows the rescaled spectra in the range of amide
Es
Amide I -1 ~1658 cm
I band. It is pretty clear that the splitting of the amide I band gradually happens as the film thickness increases, with one band at 1690 cm 1 and the other band at 1658 cm 1. Besides, using the s-polarized light, the band splitting is more significant than that using the p-polarized light, as shown in the left panel of Fig. 3. Such spectral characteristics strongly indicate this is an interface-induced phenomenon, i.e. the chain packing state at or near the interface is different from that away from the interface. Previous studies on infrared spectroscopy of protein molecules have proved that different protein secondary
A
Ep
0.8
B Amide I -1 ~1690 cm
0.3 -1
-1
Absorbance
~1690 cm
0.2
~1658 cm
0.6
Amide II
Thickness increases
Thickness increases
0.4
Amide II 0.1 0.2
0.0 1800
0.0 1700
1600
Wavenumber
1500
1800
(cm-1)
Fig. 3. The rescaled RA-FTIR spectra in the amide I range from 1800 cm 73 nm, 104 nm, 170 nm, 269 nm, 375 nm, and 559 nm, respectively.
1700
1600
Wavenumber 1
to 1500 cm
1
1500
(cm-1)
. From the bottom to top, the film thicknesses are 9.8 nm, 39 nm,
250
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
structures result in different absorption positions for amide I bands [31], demonstrating different chain conformation states can be differentiated from the band positions. Although there is not direct correlation between protein amide I band and PAL amide I band, for example, the two resolvable bands of PAL cannot be directly correlated to any specific conformation states of PAL, the two resolvable bands with the polarization dependence for the thick PAL thin films can still suggest the PAL chains in the thick films at least have two different packing states. Meanwhile, in Fig. 3, splitting of the amide I band show strong polarization-dependence, indicating the amide groups in these two packing states have different orientations on the Au substrate. As shown in Fig. 2, the intensities for the N–H stretching modes increases significantly with increasing the thickness when the s-polarized light is used (Panel A) compared to those when the p-polarized light is used (Panel B). As shown in Fig. 3, when the films are thicker than 73 nm, the splitted low-frequency band at 1658 cm 1 also increases more significantly using the s-polarized light (Panel A) compared to that using the p-polarized light (Panel B). These results suggest the side amide groups (including amino and carbonyl groups) show parallel orientation to the Au surface for the PAL molecules in the top layer of the film. Furthermore, these results also suggest that most of the PAL molecules in this packing state show parallel orientation to the Au surface since spin coating can render the expanded conformation for the polymer chains during preparation. The parallel orientation of the side amide groups indicates the polar interaction between the PAL molecules mainly lie along the surface plane instead of being normal to the surface. In Fig. 3, for spectra of the PAL films thinner than 73 nm, no splitting of amide I band can be directly observed whether using the s- or p-
polarized light, suggesting only one homogeneous state exists or the fraction of the new packing state indicated by the splitted band at 1658 cm 1 is relatively small. In the next section, we will see the latter argument is correct from fitting these spectra. Up to now, a qualitative picture for the PAL thin films spin-coated on top of the Au substrates can be drawn. As the PAL film thickness increases, two packing/orientation states for the PAL molecules inside the films can be recognized, the PAL molecules on the top part of the film show parallel orientation to the surface plane while the PAL molecules on the bottom part of the film do not show this parallel orientation, possibly induced by the non-favorable interaction from the Au surface (compared to the strong polar interaction among the PAL molecules themselves). 3.2. Further evidences by spectral fitting and a new experimental scheme Since the splitted amide I bands show strong polarization and thickness dependent (Fig. 3), the orientation of the amide groups related to the film thickness needs to be further analyzed in order to acquire an accurate structural picture at the molecular level. In this study, the collected experimental spectra were normalized with respect to that of the bare Au substrate. Since the splitting of the amide I band was observed, deconvolution of the two amide I bands is needed for the thickness-dependent analysis. A combined Lorentzian–Gaussian deconvolution method was used to fit the spectra and obtain the integrated intensity for each deconvoluted band. The fitted results for the spectra in Fig. 3 were shown in Table 1. Fig. 4 shows the example deconvoluted spectra of the 39-nm and 269-nm thick PAL films (for deconvoluted spectra of other thicknesses, please see the supporting
Table 1 The fitting results for the spectra shown in Fig 3. Sample
Polarization
Amide I (first)
Amide I (second)
Center (cm 1)
FWHM (cm 1)
Area
Center (cm 1)
FWHM (cm 1)
Amide II Area
Center (cm 1)
FWHM (cm 1)
Area
9.8 nm
Es Ep
1686.5 1685.9
39.2 36.6
0.30 3.49
1661.9 1655.8
34.8 34.4
0.10 0.78
1608.7 1613.8
49.8 41.6
0.10 0.96
39 nm
Es Ep
1686.1 1685.2
37.6 36.1
0.87 11.95
1658.7 1657.8
34.9 32.3
0.31 2.18
1611.3 1614.9
43.8 43.3
0.12 2.57
73 nm
Es Ep
1688.0 1688.3
36.6 36.2
1.55 17.30
1658.6 1659.6
34.6 34.9
0.46 3.92
1616.0 1616.1
43.2 42.0
0.55 4.41
104 nm
Es Ep
1689.0 1691.2
35.7 34.6
2.11 25.44
1658.6 1660.0
34.8 33.3
0.97 7.79
1612.3 1614.9
45.9 46.2
0.89 6.85
170 nm
Es Ep
1691.1 1692.7
34.5 39.3
4.45 27.83
1657.5 1656.8
32.9 27.8
3.33 6.03
1612.1 1615.7
51.3 56.2
2.67 9.39
269 nm
Es Ep
1692.8 1693.7
32.3 35.1
4.48 27.97
1657.7 1658.3
36.1 30.2
3.91 7.99
1611.1 1614.5
47.3 46.8
2.62 7.97
375 nm
Es Ep
1692.9 1694.1
34.7 32.9
5.76 24.50
1654.1 1658.8
34.3 35.3
6.06 11.26
1612.3 1612.1
50.3 45.6
6.42 8.36
559 nm
Es Ep
1693.2 1694.0
35.1 33.9
5.78 26.5
1653.9 1657.0
36.2 37.7
9.64 13.7
1608.8 1611.1
49.3 48.6
9.53 10.7
Note: during fitting, three band centers were set as 1690 ± 5 cm was within 10%.
1
, 1658 ± 5 cm
1
, and 1612 ± 5 cm
1
. Error of FWHMs was within 5 cm
1
and error of area
251
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
0.03
Absorbance
39 nm, E p
0.3
39 nm, E s 0.02
0.2
0.01
0.1
0.0
0.00 1750
1700
1650
1600
1550
1750
1700
1650
1600
1550
0.8
269 nm, E s
Absorbance
0.15 0.10
269 nm, E p
0.6 0.4
0.05
0.2
0.00
0.0 1750
1700
1650
Wavenumber
1600
(cm-1)
1550
1750
1700
1650
Wavenumber
1600
1550
(cm-1)
Fig. 4. The deconvoluted spectra of the 39-nm and 269-nm thick PAL films in the amide I range. Original spectra are shown as black thin lines; fitted spectra are shown as black thick lines; deconvoluted bands are shown as gray lines.
information). Therefore, the area ratio of the splitted amide I band (A1658/A1690) is plotted in terms of the film thickness for the s- and p-polarized spectra, respectively. Clearly shown in Fig. 5, as the film thickness increases, A1658/A1690 using the s-polarized light increases more significantly than that using the p-polarized light, evidenced by the two different slopes by linearly fitting the data, i.e. 0.0024 and 0.0006. This again suggests the two packing states indicated by the splitted bands at 1690 cm 1 and 1658 cm 1 show different orientations on the Au substrate surfaces. The carbonyl groups in the PAL chains related to the 1658 cm 1 band preferentially show parallel orientation to the Au surface.
Ratio of splitted amide I band
2.0
A~1658 /A ~1690
Es
1.5
Slope=0.0024
1.0
Ep Slope=0.0006
0.5
0.0 0
100
200
300
400
500
600
Thickness (nm) Fig. 5. Intensity ratio of the splitted amide I bands (A1658/A1690) in terms of the film thickness, for the s- (Es) and p-polarized (Ep) spectra, respectively.
Furthermore, the preferred orientation of the amino groups in term of the film thickness also needs to be evaluated. Similarly, the spectra in the N–H stretching range (3550–3000 cm 1) were deconvoluted for the s- and ppolarized spectra. Based on the previous studies [26,30], there are three discernable bands in this range, i.e. the N–H symmetric stretching (ss) of the amino groups at 3195 cm 1, the N–H anti-symmetric stretching (as) mode of the amino groups at 3340 cm 1, and a hydrogen-bond related vibrational mode at 3450 cm 1. As an example, the deconvoluted spectra of the 39-nm and 269-nm thick PAL thin films are shown in Fig. 6. The fitting results are shown in Table 2. We also plotted the intensity areas of the N–H ss and as modes of amino groups in terms of the film thickness, as shown in Fig. 7. It can be seen, whether for the N–H ss or as modes, the integrated area increases more significantly for the s-polarized spectra than those for the p-polarized spectra, suggesting the amino groups show certain parallel orientation to the surface plane as the film thickness increases. Moreover, a sharp increase of the intensity areas of the N–H ss and as modes is remarkably observed when the film thickness increases from 300 nm to 400 nm, when the s-polarized light is used, suggesting the amino groups show strong parallel orientation to the surface plane when the film thickness is above this length scale. We speculate that the parallel orientation of the amino groups is correlated to the parallel orientation of the carbonyl groups, indicated by the lowfrequency splitted amide I band at 1658 cm 1 in Fig. 3 as the film thickness increases. Therefore, a gradiently varied bi-layered structure in the spin-coated PAL thin films should be responsible for the splitted amide I bands and the parallel orientation of the amino and carbonyl groups discussed above. The layer adjacent to the Au substrate, with most amino and carbonyl groups randomly
252
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
0.08
39 nm, E p
39 nm, Es
0.010
Absorbance
0.06 0.04 0.005
0.02 0.00 3500 3400 3300 3200 3100 3000
3500 3400 3300 3200 3100 3000 0.20
269 nm, E s
Absorbance
0.15
269 nm, E p
0.15
0.10 0.10 0.05
0.05
0.00
0.00 3500 3400 3300 3200 3100 3000
Wavenumber
3500 3400 3300 3200 3100 3000
(cm-1)
Wavenumber (cm-1)
Fig. 6. The deconvoluted spectra of the 39-nm and 269-nm thick PAL films in the amino stretching range. Original spectra are shown as black thin lines; fitted spectra are shown as black thick lines; deconvoluted bands are shown as gray lines.
orientated, is believed to be induced by the non-favorable interaction between the PAL molecules and the Au surface. The other layer on top with most of the PAL molecules orientating parallel to the Au surface (also correlated to the parallel orientation of amide groups), is believed to be induced by spin-coating and stabilized by the polar interaction (hydrogen bonding) between the amino and carbonyl groups. Because of the interplay between the interfacial and spin-coating effects, the chain orientation
induced by spin coating is more remarkable when the film thickness reaches 300 nm. Finally, in order to further confirm the gradiently bi-layered structure, the RA-FTIR spectra using the p-polarized light with different incidence angles were collected for the spin-coated PAL films of 73 nm and 297 nm. The scheme of applying different incidence angles and the ppolarized light for RA-FTIR is shown in Fig. 8. The angles between the E vectors and the Au surface are 10°, 30°,
Table 2 The fitting results for the amino stretching range in Fig. 2. Sample
Polarization
H-bonded
NH2 as
Center (cm 1)
FWHM (cm 1)
Area
NH2 ss
Center (cm 1)
FWHM (cm 1)
Area
Center (cm 1)
FWHM (cm 1)
Area
9.8 nm
Es Ep
3470.3 3449.4
62.6 63.5
0.03 0.34
3347.3 3341.4
179.3 180.0
0.36 3.01
3197.8 3193.1
94.6 91.5
0.20 1.42
39 nm
Es Ep
3449.0 3452.6
72.2 77.2
0.19 2.16
3349.5 3350.0
139.1 150.8
0.77 7.66
3197.8 3201.0
98.8 101.1
0.81 6.40
73 nm
Es Ep
3448.0 3447.6
71.6 90.1
0.34 3.90
3346.6 3347.5
153.9 142.9
2.17 14.35
3196.9 3202.4
90.4 101.8
1.28 8.51
104 nm
Es Ep
3448.4 3448.1
65.5 89.6
0.47 5.03
3343.4 3342.7
153.7 151.7
2.96 16.68
3193.2 3193.8
90.3 96.5
1.75 11.33
170 nm
Es Ep
3449.5 3450.8
86.3 84.2
2.88 5.08
3338.1 3342.5
163.2 159.7
10.42 19.84
3188.2 3191.8
96.2 95.7
7.26 14.06
269 nm
Es Ep
3446.2 3451.2
98.2 82.8
5.78 4.44
3339.6 3342.8
161.3 161.9
18.27 20.24
3191.6 3190.4
92.5 96.1
12.20 13.70
375 nm
Es Ep
3454.5 3446.6
124.1 117.5
37.47 14.91
3336.8 3331.8
177.4 154.1
68.05 27.89
3186.2 3185.9
152.4 110.0
56.25 23.43
559 nm
Es Ep
3454.5 3450.3
120.5 110.4
43.70 19.35
3341.6 3334.6
158.4 157.7
72.66 43.03
3192.8 3186.0
141.9 103.3
69.69 30.18
Note: during fitting, three band centers were set as 3450 ± 5 cm was within 10 cm 1 and error of area was within 10%.
1
(except for the 9.8-nm thick film), 3340 ± 10 cm
1
, and 3195 ± 10 cm
1
. Error of FWHMs
253
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
Area of amino stretching band
100
N-H stretching mode of amino groups
Es, ss
80
Es, as Ep, ss
60
Ep, as
40
20
0 0
100
200
300
400
500
600
Thickness (nm) Fig. 7. The intensities of the amino ss and as modes in terms of the PAL film thickness.
Fig. 8. The schematic of applying different incidence angles and the ppolarized light for RA-FTIR. When the incidence angle changed from 85° to 10°, the E vector direction changes from nearly perpendicular to nearly parallel to the surface.
0.2
A
Absorbance
PAL on Au, 73 nm 85 o p-polarized light o
60 30 o, 10 o
0.1
0.2
PAL on Au, 73 nm p-polarized light
85 o 60 o 30 o, 10 o
0.1
A'
0.0
0.0 3500
3000
2500
2000
1500
60 30 o, 10 o
0.4
1800
B
PAL on Au, 297 nm o p-polarized light 85 o
Absorbance
60°, and 85°, respectively. The directions of the E vectors, as indicated by E10°, E30°, E60°, and E85° in Fig. 8, is changed from approximately parallel to perpendicular to the Au surface in the above sequence. Thus the molecular groups with transition dipole moments from parallel to perpendicular orientation can be detected with respect to the surface plane. Fig. 9 shows the RA-FTIR spectra for the 73-nm and 297-nm thick PAL films on the Au surfaces based on this scheme. It can be seen, for the 73-nm thick PAL film, all the spectra show the similar feature. For the 297-nm thick PAL film, when the incidence angle changes from 85° to 10°, intensities of the N–H ss and as modes of the amino groups increase significantly; in the meantime, the splitting of the amide I band gradually appears with increasing the low-frequency band at 1658 cm 1 and decreasing the high frequency band at 1690 cm 1. Such experimental results further solidify our argument that the spin-coated PAL thin films on top of the Au substrates are of the gradiently bi-layered structure. When the PAL thin film is thick enough, the bi-layered structure is more distinct. Most of the PAL molecules in the bottom layer attached to the Au surface is of random orientation while most of the PAL molecules in the top layer show strong parallel orientation to the surface plane, as evidenced by the increased N–H stretching vibrational modes of the amino groups and the low-frequency band at 1658 cm 1 when the s-polarized is used (Fig. 3) or the E vector is adjusted prone to be parallel to the Au surface (Fig. 9). It should be noted, inside the PAL thin films of the gradiently bi-layered structure, there should be no sharp interface between the bottom random layer and the top orientated layer. On the contrary, there should exist a smooth gradient in between, as suggested by the data in Figs. 5, 9 (the new scheme) and Table 1, where for all the PAL thin films two amide I bands can be deconvoluted, even for the very thin films.
1750
1700
1650
1600
PAL on Au, 297 nm o ~1658 85o p-polarized light
1550
B'
60 30o, 10o
0.4
0.2
0.2
0.0
0.0 3500
3000
2500
2000
Wavenumber (cm-1)
1500
1800
~1690 1750
1700
1650
1600
1550
Wavenumber (cm-1)
Fig. 9. The p-polarized spectra of the 73-nm and 297-nm thick PAL films collected with different incidence angles of 10°, 30°, 60°, and 85°.
254
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
4. Conclusions Using the polarized RA-FTIR spectroscopy with the grazing angle incidence and the scheme depicted in Fig. 8, we explicitly reveal there exists a gradiently varied bi-layered structure for the spin-coated PAL thin films on top of the Au substrates. The splitting of the amide I band and the polarization-dependent infrared absorption with respect to the PAL film thickness indicate the two packing states are orientationally different. The bottom layer adjacent to the Au substrate is mainly composed of the PAL molecules randomly orientated, induced by the non-favorable interaction from the Au surface. The top layer is mainly composed of the PAL molecules with parallel orientation to the Au surface indicated by the polarization-dependent absorption of amide I band and amino stretching modes, which is induced by spin coating. Splitting of the amide I band is more significant when the film thickness is more than 102 nm, suggesting the interfacial interaction could affect orientation of the PAL molecules over the length scale of at least a molecular size. The observation described in this study is a distinct phenomenon for polymer thin films spin-coated on the solid substrates. The orientated PAL molecules induced by spin coating are supposed to be quasi-stabilized by the strong polar interaction between the PAL chain units. In this respect, the polymer thin films are metastable and the chain evolution could still happen in the long run. Conflicts of interest The authors declare no competing financial interest. Acknowledgments This study was supported by the National Natural Science Foundation of China (51173169), Scientific Research Foundation for the Returned Overseas Chinese Scholars, Ministry of Education, and Human Resources and Social Security Bureau of Zhejiang Province, China. X.L. is also grateful for the Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions, China (PAPD). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.eurpolymj.2014.12.030. References [1] Kheshgi HS, Scriven LE. Dewetting: nucleation and growth of dry regions. Chem Eng Sci 1991;46:519–26. [2] Tsui OKC. In: Tsui OKC, Russell TP, editors. Polymer thin films. Singapore: World Scientific; 2008. p. 267–94 [chapter 11]. [3] Reiter G. Mobility of polymers in films thinner than their unperturbed size. Europhys Lett 1993;23:579–84. [4] Reiter G, de Gennes PG. Spin-cast, thin, glassy polymer films: highly metastable forms of matter. Eur Phys J E 2001;6:25–8.
[5] Richardson H, Carelli C, Keddie L, Sferrazza M. Structural relaxation of spin-cast glassy polymer thin films as a possible factor in dewetting. Eur Phys J E 2003;12:437–41. [6] Reiter G, Hamieh M, Damman P, Cclavons S, Gabriele S, Vilmin T, et al. Residual stresses in thin polymer films cause rupture and dominate early stages of dewetting. Nat Mater 2005;4:754–8. [7] Gu X, Reghavan D, Douglas JF, Karim A. Hole-growth instability in the dewetting of evaporating polymer solution films. J Polym Sci, Part B: Polym Phys 2002;40:2825–32. [8] Green PF, Ganesan V. Dewetting of polymeric films: unresolved issues. Eur Phys J E 2003;12:449–54. [9] Ellison CJ, Torkelson JM. The distribution of glass-transition temperatures in nanoscopically confined glass formers. Nat Mater 2003;2:695–700. [10] Tomczak N, Vallee RAL, van Dijk EMHP, Kuipers L, van Hulst NF, Vancso GJ. Segment dynamics in thin polystyrene films probed by single-molecule optics. J Am Chem Soc 2004;126:4748–9. [11] Priestley RD, Ellison CJ, Broadbelt LJ, Torkelson JM. Structural relaxation of polymer glasses at surfaces, interfaces, and in between. Science 2005;309:456–9. [12] Habuchi S, Oba T, Vacha M. Multi-beam single-molecule defocused fluorescence imaging reveals local anisotropic nature of polymer thin films. Phys Chem Chem Phys 2011;13:7001–7. [13] Keddie JL, Jones RAL, Cory RA. Interface and surface effects on the glass-transition temperature in thin polymer films. Faraday Discuss 1994;98:219–30. [14] Keddie JL, Jones RAL, Cory RA. Europhys Lett 1994;27:59–64. [15] Sharp JS, Forrest JA. Free surfaces cause reductions in the glass transition temperature of thin polystyrene films. Phys Rev Lett 2003;91. 235701-1–4. [16] Forrest JA, Dalnok-Veress K, Dutcher JR. Interface and chain confinement effects on the glass transition temperature of thin polymer films. Phys Rev E 1997;56:5705–16. [17] Forrest JA, Dalnoki-Veress K, Stevens JR, Dutcher JR. Effect of free surfaces on the glass transition temperature of thin polymer films. Phys Rev Lett 1996;77:2002–5. [18] Forrest JA, Mattsson J. Reductions of the glass transition temperature in thin polymer films: probing the length scale of cooperative dynamics. Phys Rev E 2000;61:R53–6. [19] Wallace WE, van Zanten JH, Wu WL. Influence of an impenetrable interface on a polymer glass-transition temperature. Phys Rev E 1995;52:R3329–32. [20] van Zanten JH, Wallace WE, Wu WL. Effect of strongly favorable substrate interactions on the thermal properties of ultrathin polymer films. Phys Rev E 1995;53:R2053–6. [21] Gracias DH, Zhang D, Lianos L, Ibach W, Shen YR, Somorjai GA. A study of the glass transition of polypropylene surfaces by sumfrequency vibrational spectroscopy and scanning force microscopy. Chem Phys 1999;245:277–84. [22] Zhang C, Hong SC, Ji N, Wang YP, Wei KH, Shen YR. Sum-frequency vibrational spectroscopic study of surface glass transition of poly(vinyl alcohol). Macromolecules 2003;36:3303–6. [23] Harp GP, Rangwalla H, Yeganeh MS, Dhinojwala A. Infrared-visible sum frequency generation spectroscopic study of molecular orientation at polystyrene/comb-polymer interfaces. J Am Chem Soc 2003;125:11283–90. [24] Li Q, Hua R, Cheah IJ, Chou KC. Surface structure relaxation of poly(methyl methacrylate). J Phys Chem B 2008;112:694–7. [25] Li B, Zhou J, Xu X, Yu J, Shao W, Fang Y, et al. Solvent quality affects chain conformational order at the polymer surface revealed by sum frequency generation vibrational spectroscopy. Polymer 2013;54:1853–9. [26] Lu X, Cheng I, Mi Y. Discovery of the bi-layered structure in spincoated polyacrylamide films on the gold surface. Polymer 2007;48:682–6. [27] Greenler RG. Infrared study of adsorbed molecules on metal surfaces by reflection techniques. J Chem Phys 1966;44:310–5. [28] Allara DL, Baca A, Pryde CA. Distortions of band shapes in external reflection infrared spectra of thin polymer films on metal substrates. Macromolecules 1978;11:1215–20. [29] Buffeteau T, Desbat B, Turlet JM. Polarization modulation FT-IR spectroscopy of surfaces and ultra-thin films: experimental procedure and quantitative analysis. Appl Spectrosc 1991;45:380–9. [30] Lu X, Mi Y. Characterization of the interfacial interaction between polyacrylamide and silicon substrate by Fourier transform infrared spectroscopy. Macromolecules 2005;38:839–43. [31] Tamm LK, Tatulian SA. Infrared spectroscopy of proteins and peptides in lipid bilayers. Q Rev Biophys 1997;30:365–429.
Contents lists available at ScienceDirect
European Polymer Journal journal homepage: www.elsevier.com/locate/europolj
Gradiently varied chain packing/orientation states of polymer thin films revealed by polarization-dependent infrared absorption Xiaolin Lu a,⇑, Yongli Mi b,⇑ a b
State Key Laboratory of Bioelectronics, School of Biological Science & Medical Engineering, Southeast University, Nanjing 210096, Jiangsu Province, PR China Department of Chemical and Biomolecular Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong
a r t i c l e
i n f o
Article history: Received 8 November 2014 Received in revised form 22 December 2014 Accepted 22 December 2014 Available online 30 December 2014 Keywords: RA-FTIR Polarization-dependent Polyacrylamide Thin films
a b s t r a c t Gradiently varied chain packing/orientation states of the polyacrylamide (PAL) thin films spin-coated on the gold (Au) substrates were found via the polarized reflection–absorption Fourier transform infrared spectroscopy (RA-FTIR). As the film thickness increases, the splitted amide I bands provide a direct evidence that the PAL thin films are of a gradiently varied bi-layered structure. In the bottom layer, most of the PAL molecules show random orientation which is induced by the non-favorable interaction from the adjacent Au surface. In the top layer, most of PAL molecules show parallel orientation to the Au surface which is induced by spin coating, evidenced by the enhanced low-frequency splitted amide I band (1658 cm 1) and N–H stretching modes of the amino groups when the light electric field vector is adjusted to be parallel to the Au surface. The observation reported in this study should be of universal significance for polymer thin films on the supported substrates, where the interfacial interaction as well as spin coating could vary the polymer packing/orientation states substantially. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Polymer thin films are a kind of important materials due to their broad applications in coating, adhesive and microelectronic industries [1,2]. The long-term reliability of a polymer thin film to a large extent relies on whether the polymer chains in the film are in the relaxed state or not because the evolution of the chain conformation during usage may have negative effect on its stability. For instance, when the thickness of a polymer thin film decreases down to or lower than certain critical length, i.e. a molecular size in the scale of the mean-square endto-end distance (Ree) or radius of gyration (Rg), chain conformation of the polymer could notably be different from ⇑ Corresponding authors. Tel.: +86 25 83791810 (X. Lu). E-mail addresses: [email protected] (X. Lu), [email protected] (Y. Mi). http://dx.doi.org/10.1016/j.eurpolymj.2014.12.030 0014-3057/Ó 2014 Elsevier Ltd. All rights reserved.
that of its bulk state [3–6]. Therefore the residue stress or the entropic force may evidently exist and gradually deteriorate the property of the polymer thin film [4,6–8]. It is thus of great importance to investigate the chain packing/orientation state and the related dynamic behavior for a polymer thin film. Different techniques have been applied to study this issue in the past, such as fluorescence microscopy [9–12], ellipsometry [13–16], Brillouin light scattering [16–18], X-ray reflectivity [3,19,20] and sum frequency generation vibrational spectroscopy [21–25]. It has generally been recognized that a polymer thin film is a metastable material [4] which has distinct physicochemical properties from those of the bulk polymer. For a polymer thin film on the supported substrate, besides the chain confinement effect, the interaction between the polymer and the substrate, whether attractive or nonattractive, can also have significant effect which can alter
248
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
arrangement of the polymer chains near the polymer/substrate interface. A preliminary report on the supported polymer thin films prepared by spin coating has suggested the interaction from the substrate and spin coating can remarkably alter the chain packing states of the polymer thin films [26]. In this study, complete experimental results were present to explore the chain orientation states by analyzing the vibrational signals of the polyacrylamide (PAL) molecular groups using the polarized reflection– absorption Fourier transform infrared spectroscopy (RAFTIR) with the grazing-angle incidence. A gradiently varied bi-layered structure of the PAL thin films on the Au substrates was confirmed. The result indicates the PAL chains in the bottom layer (adjacent to the Au substrate) and the top layer have undoubtedly different orientational orderings which lead to the polarization-dependent infrared absorption. 2. Experimental section The Au substrates were prepared by depositing a 200nm Au layer on top of the silicon wafers (Silicon polished wafer, N-type, Grinm, Inc.) using the high vacuum evaporator (High Vacuum Evaporator, DV-502A, Denton Inc.). The silicon wafers were pretreated by piranha solution (a mixed solution with 3:7 volume ratio of 30 wt.% H2O2 solution and 98 wt.% H2SO4) to eliminate the possible surface contamination. PAL (Mw = 5–6 106) were purchased from Polysciences, Inc. The mean square end-to-end distance is estimated to be around 102 nm. The PAL thin films were prepared by spin coating PAL solutions on top of the Au substrates (Spin Coater, P6700, Specialty Coating Systems Inc.). Solution concentration and spinning speed were adjusted to obtain the PAL thin films with desired thicknesses. The PAL films with thicknesses of 9.8 nm, 39 nm, 73 nm, 104 nm, 170 nm, 269 nm, 375 nm, and 559 nm were prepared. All the PAL film samples were annealed in a vacuum oven (Shel lab 1410, Shalton Mfg Inc.) at 60 °C for 24 h. Before further characterization, the samples
were put in a desiccator. The Variable-Angle Spectroscopic Ellipsometer (J.A. Woollam Co., Inc.) was used to measure the thicknesses of the PAL thin films on the Au substrates. The RA-FTIR spectroscopy (RA-FTIR, FTS-6000, Bio-Rad Inc.) were applied to capture the polarized infrared spectra of the PAL thin films on the Au substrates. The polarizer was mounted before the infrared light shot at the sample surface. The s- and p-polarized lights with the grazingangle (85°) incidence were used, as shown in Fig. 1. With this grazing-angle incidence, the transition dipole moments of molecular groups parallel and normal to the Au surface can be separately detected simply by using sand p-polarized lights respectively, as demonstrated by the selection rules for RA-FTIR on the reflected metal surfaces [27–29]. The ZnSe polarizer (Tydex J.S. Co., spectral range from 1.5 lm to 14 lm or 6667 cm 1 to 714 cm 1) was used to obtain the light with >95% polarization. The orientation of a certain PAL molecular group can thus be analyzed based on its polarization-dependent infrared absorption. During the infrared experiment, the sample chamber of the infrared spectrometer was vacuumized for at least 1 h before collecting the background and the sample spectra. Otherwise the interference from humidity and carbon dioxide in the sample chamber would lead to the low signal-to-noise ratio. 3. Results and discussion 3.1. Polarization-dependent spectra and splitting of amide I band The normal infrared spectrum for a PAL bulk sample has been reported previously [30]. The amide I mode and the N–H stretching modes from the amide groups (ACONH2) are the strongest modes. Here as shown in Fig. 2, using RA-FTIR, the collected spectra for a series of PAL thin films spin-coated onto the Au substrates show strong characteristics depending on the light polarization and film thickness. Fig. 2 shows the RA-FTIR spectra of the PAL thin film
Fig. 1. Optics of RA-FTIR with a grazing-angle incidence (left panel). For a certain molecular group, with this grazing angle incidence, the infrared transition dipole moment parallel to the Au substrate surface is detected when the light polarization is adjusted to be ‘‘s’’ and the infrared transition dipole moment normal to the Au substrate surface is detected when the light polarization is adjusted to be ‘‘p’’. The monomer unit of PAL is also shown (right panel).
249
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
0.6
Es
9.8 nm 39 nm 73 nm 104 nm 170 nm 269 nm 375 nm 559 nm
NH2
Absorbance
0.4
A
Ep
0.8
B
9.8 nm 39 nm 73 nm 104 nm 170 nm 269 nm 375 nm 559 nm
0.6
Amide I
0.4
Amide I
NH2
0.2
Thickness increase
Thickness increase
0.2
0.0
0.0 3500
3000
2500
Wavenumber
2000
1500
3500
(cm-1)
3000
2500
Wavenumber
2000
1500
(cm-1)
Fig. 2. The s- (Panel A) and p-polarized (Panel B) RA-FTIR spectra of the PAL thin films with different thicknesses spin-coated on the Au surface with the grazing angle (85°) incidence. From the bottom to top, the film thicknesses are 9.8 nm, 39 nm, 73 nm, 104 nm, 170 nm, 269 nm, 375 nm, and 559 nm, respectively.
samples with eight different thicknesses (from 9.8 nm to 559 nm) collected using the s- (Panel A) and p-polarized (Panel B) lights, respectively. Two remarkable features can be read from those spectra. Firstly, as the film thickness increases, the intensities for the N–H stretching modes (3000–3600 cm 1) of the amino groups using the s-polarized light increase more significantly than those using the p-polarized light. Secondly, the splitting of the amide I mode is clearly observed when the film thickness is above 73 nm and such band splitting appears polarization-dependent. Fig. 3 shows the rescaled spectra in the range of amide
Es
Amide I -1 ~1658 cm
I band. It is pretty clear that the splitting of the amide I band gradually happens as the film thickness increases, with one band at 1690 cm 1 and the other band at 1658 cm 1. Besides, using the s-polarized light, the band splitting is more significant than that using the p-polarized light, as shown in the left panel of Fig. 3. Such spectral characteristics strongly indicate this is an interface-induced phenomenon, i.e. the chain packing state at or near the interface is different from that away from the interface. Previous studies on infrared spectroscopy of protein molecules have proved that different protein secondary
A
Ep
0.8
B Amide I -1 ~1690 cm
0.3 -1
-1
Absorbance
~1690 cm
0.2
~1658 cm
0.6
Amide II
Thickness increases
Thickness increases
0.4
Amide II 0.1 0.2
0.0 1800
0.0 1700
1600
Wavenumber
1500
1800
(cm-1)
Fig. 3. The rescaled RA-FTIR spectra in the amide I range from 1800 cm 73 nm, 104 nm, 170 nm, 269 nm, 375 nm, and 559 nm, respectively.
1700
1600
Wavenumber 1
to 1500 cm
1
1500
(cm-1)
. From the bottom to top, the film thicknesses are 9.8 nm, 39 nm,
250
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
structures result in different absorption positions for amide I bands [31], demonstrating different chain conformation states can be differentiated from the band positions. Although there is not direct correlation between protein amide I band and PAL amide I band, for example, the two resolvable bands of PAL cannot be directly correlated to any specific conformation states of PAL, the two resolvable bands with the polarization dependence for the thick PAL thin films can still suggest the PAL chains in the thick films at least have two different packing states. Meanwhile, in Fig. 3, splitting of the amide I band show strong polarization-dependence, indicating the amide groups in these two packing states have different orientations on the Au substrate. As shown in Fig. 2, the intensities for the N–H stretching modes increases significantly with increasing the thickness when the s-polarized light is used (Panel A) compared to those when the p-polarized light is used (Panel B). As shown in Fig. 3, when the films are thicker than 73 nm, the splitted low-frequency band at 1658 cm 1 also increases more significantly using the s-polarized light (Panel A) compared to that using the p-polarized light (Panel B). These results suggest the side amide groups (including amino and carbonyl groups) show parallel orientation to the Au surface for the PAL molecules in the top layer of the film. Furthermore, these results also suggest that most of the PAL molecules in this packing state show parallel orientation to the Au surface since spin coating can render the expanded conformation for the polymer chains during preparation. The parallel orientation of the side amide groups indicates the polar interaction between the PAL molecules mainly lie along the surface plane instead of being normal to the surface. In Fig. 3, for spectra of the PAL films thinner than 73 nm, no splitting of amide I band can be directly observed whether using the s- or p-
polarized light, suggesting only one homogeneous state exists or the fraction of the new packing state indicated by the splitted band at 1658 cm 1 is relatively small. In the next section, we will see the latter argument is correct from fitting these spectra. Up to now, a qualitative picture for the PAL thin films spin-coated on top of the Au substrates can be drawn. As the PAL film thickness increases, two packing/orientation states for the PAL molecules inside the films can be recognized, the PAL molecules on the top part of the film show parallel orientation to the surface plane while the PAL molecules on the bottom part of the film do not show this parallel orientation, possibly induced by the non-favorable interaction from the Au surface (compared to the strong polar interaction among the PAL molecules themselves). 3.2. Further evidences by spectral fitting and a new experimental scheme Since the splitted amide I bands show strong polarization and thickness dependent (Fig. 3), the orientation of the amide groups related to the film thickness needs to be further analyzed in order to acquire an accurate structural picture at the molecular level. In this study, the collected experimental spectra were normalized with respect to that of the bare Au substrate. Since the splitting of the amide I band was observed, deconvolution of the two amide I bands is needed for the thickness-dependent analysis. A combined Lorentzian–Gaussian deconvolution method was used to fit the spectra and obtain the integrated intensity for each deconvoluted band. The fitted results for the spectra in Fig. 3 were shown in Table 1. Fig. 4 shows the example deconvoluted spectra of the 39-nm and 269-nm thick PAL films (for deconvoluted spectra of other thicknesses, please see the supporting
Table 1 The fitting results for the spectra shown in Fig 3. Sample
Polarization
Amide I (first)
Amide I (second)
Center (cm 1)
FWHM (cm 1)
Area
Center (cm 1)
FWHM (cm 1)
Amide II Area
Center (cm 1)
FWHM (cm 1)
Area
9.8 nm
Es Ep
1686.5 1685.9
39.2 36.6
0.30 3.49
1661.9 1655.8
34.8 34.4
0.10 0.78
1608.7 1613.8
49.8 41.6
0.10 0.96
39 nm
Es Ep
1686.1 1685.2
37.6 36.1
0.87 11.95
1658.7 1657.8
34.9 32.3
0.31 2.18
1611.3 1614.9
43.8 43.3
0.12 2.57
73 nm
Es Ep
1688.0 1688.3
36.6 36.2
1.55 17.30
1658.6 1659.6
34.6 34.9
0.46 3.92
1616.0 1616.1
43.2 42.0
0.55 4.41
104 nm
Es Ep
1689.0 1691.2
35.7 34.6
2.11 25.44
1658.6 1660.0
34.8 33.3
0.97 7.79
1612.3 1614.9
45.9 46.2
0.89 6.85
170 nm
Es Ep
1691.1 1692.7
34.5 39.3
4.45 27.83
1657.5 1656.8
32.9 27.8
3.33 6.03
1612.1 1615.7
51.3 56.2
2.67 9.39
269 nm
Es Ep
1692.8 1693.7
32.3 35.1
4.48 27.97
1657.7 1658.3
36.1 30.2
3.91 7.99
1611.1 1614.5
47.3 46.8
2.62 7.97
375 nm
Es Ep
1692.9 1694.1
34.7 32.9
5.76 24.50
1654.1 1658.8
34.3 35.3
6.06 11.26
1612.3 1612.1
50.3 45.6
6.42 8.36
559 nm
Es Ep
1693.2 1694.0
35.1 33.9
5.78 26.5
1653.9 1657.0
36.2 37.7
9.64 13.7
1608.8 1611.1
49.3 48.6
9.53 10.7
Note: during fitting, three band centers were set as 1690 ± 5 cm was within 10%.
1
, 1658 ± 5 cm
1
, and 1612 ± 5 cm
1
. Error of FWHMs was within 5 cm
1
and error of area
251
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
0.03
Absorbance
39 nm, E p
0.3
39 nm, E s 0.02
0.2
0.01
0.1
0.0
0.00 1750
1700
1650
1600
1550
1750
1700
1650
1600
1550
0.8
269 nm, E s
Absorbance
0.15 0.10
269 nm, E p
0.6 0.4
0.05
0.2
0.00
0.0 1750
1700
1650
Wavenumber
1600
(cm-1)
1550
1750
1700
1650
Wavenumber
1600
1550
(cm-1)
Fig. 4. The deconvoluted spectra of the 39-nm and 269-nm thick PAL films in the amide I range. Original spectra are shown as black thin lines; fitted spectra are shown as black thick lines; deconvoluted bands are shown as gray lines.
information). Therefore, the area ratio of the splitted amide I band (A1658/A1690) is plotted in terms of the film thickness for the s- and p-polarized spectra, respectively. Clearly shown in Fig. 5, as the film thickness increases, A1658/A1690 using the s-polarized light increases more significantly than that using the p-polarized light, evidenced by the two different slopes by linearly fitting the data, i.e. 0.0024 and 0.0006. This again suggests the two packing states indicated by the splitted bands at 1690 cm 1 and 1658 cm 1 show different orientations on the Au substrate surfaces. The carbonyl groups in the PAL chains related to the 1658 cm 1 band preferentially show parallel orientation to the Au surface.
Ratio of splitted amide I band
2.0
A~1658 /A ~1690
Es
1.5
Slope=0.0024
1.0
Ep Slope=0.0006
0.5
0.0 0
100
200
300
400
500
600
Thickness (nm) Fig. 5. Intensity ratio of the splitted amide I bands (A1658/A1690) in terms of the film thickness, for the s- (Es) and p-polarized (Ep) spectra, respectively.
Furthermore, the preferred orientation of the amino groups in term of the film thickness also needs to be evaluated. Similarly, the spectra in the N–H stretching range (3550–3000 cm 1) were deconvoluted for the s- and ppolarized spectra. Based on the previous studies [26,30], there are three discernable bands in this range, i.e. the N–H symmetric stretching (ss) of the amino groups at 3195 cm 1, the N–H anti-symmetric stretching (as) mode of the amino groups at 3340 cm 1, and a hydrogen-bond related vibrational mode at 3450 cm 1. As an example, the deconvoluted spectra of the 39-nm and 269-nm thick PAL thin films are shown in Fig. 6. The fitting results are shown in Table 2. We also plotted the intensity areas of the N–H ss and as modes of amino groups in terms of the film thickness, as shown in Fig. 7. It can be seen, whether for the N–H ss or as modes, the integrated area increases more significantly for the s-polarized spectra than those for the p-polarized spectra, suggesting the amino groups show certain parallel orientation to the surface plane as the film thickness increases. Moreover, a sharp increase of the intensity areas of the N–H ss and as modes is remarkably observed when the film thickness increases from 300 nm to 400 nm, when the s-polarized light is used, suggesting the amino groups show strong parallel orientation to the surface plane when the film thickness is above this length scale. We speculate that the parallel orientation of the amino groups is correlated to the parallel orientation of the carbonyl groups, indicated by the lowfrequency splitted amide I band at 1658 cm 1 in Fig. 3 as the film thickness increases. Therefore, a gradiently varied bi-layered structure in the spin-coated PAL thin films should be responsible for the splitted amide I bands and the parallel orientation of the amino and carbonyl groups discussed above. The layer adjacent to the Au substrate, with most amino and carbonyl groups randomly
252
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
0.08
39 nm, E p
39 nm, Es
0.010
Absorbance
0.06 0.04 0.005
0.02 0.00 3500 3400 3300 3200 3100 3000
3500 3400 3300 3200 3100 3000 0.20
269 nm, E s
Absorbance
0.15
269 nm, E p
0.15
0.10 0.10 0.05
0.05
0.00
0.00 3500 3400 3300 3200 3100 3000
Wavenumber
3500 3400 3300 3200 3100 3000
(cm-1)
Wavenumber (cm-1)
Fig. 6. The deconvoluted spectra of the 39-nm and 269-nm thick PAL films in the amino stretching range. Original spectra are shown as black thin lines; fitted spectra are shown as black thick lines; deconvoluted bands are shown as gray lines.
orientated, is believed to be induced by the non-favorable interaction between the PAL molecules and the Au surface. The other layer on top with most of the PAL molecules orientating parallel to the Au surface (also correlated to the parallel orientation of amide groups), is believed to be induced by spin-coating and stabilized by the polar interaction (hydrogen bonding) between the amino and carbonyl groups. Because of the interplay between the interfacial and spin-coating effects, the chain orientation
induced by spin coating is more remarkable when the film thickness reaches 300 nm. Finally, in order to further confirm the gradiently bi-layered structure, the RA-FTIR spectra using the p-polarized light with different incidence angles were collected for the spin-coated PAL films of 73 nm and 297 nm. The scheme of applying different incidence angles and the ppolarized light for RA-FTIR is shown in Fig. 8. The angles between the E vectors and the Au surface are 10°, 30°,
Table 2 The fitting results for the amino stretching range in Fig. 2. Sample
Polarization
H-bonded
NH2 as
Center (cm 1)
FWHM (cm 1)
Area
NH2 ss
Center (cm 1)
FWHM (cm 1)
Area
Center (cm 1)
FWHM (cm 1)
Area
9.8 nm
Es Ep
3470.3 3449.4
62.6 63.5
0.03 0.34
3347.3 3341.4
179.3 180.0
0.36 3.01
3197.8 3193.1
94.6 91.5
0.20 1.42
39 nm
Es Ep
3449.0 3452.6
72.2 77.2
0.19 2.16
3349.5 3350.0
139.1 150.8
0.77 7.66
3197.8 3201.0
98.8 101.1
0.81 6.40
73 nm
Es Ep
3448.0 3447.6
71.6 90.1
0.34 3.90
3346.6 3347.5
153.9 142.9
2.17 14.35
3196.9 3202.4
90.4 101.8
1.28 8.51
104 nm
Es Ep
3448.4 3448.1
65.5 89.6
0.47 5.03
3343.4 3342.7
153.7 151.7
2.96 16.68
3193.2 3193.8
90.3 96.5
1.75 11.33
170 nm
Es Ep
3449.5 3450.8
86.3 84.2
2.88 5.08
3338.1 3342.5
163.2 159.7
10.42 19.84
3188.2 3191.8
96.2 95.7
7.26 14.06
269 nm
Es Ep
3446.2 3451.2
98.2 82.8
5.78 4.44
3339.6 3342.8
161.3 161.9
18.27 20.24
3191.6 3190.4
92.5 96.1
12.20 13.70
375 nm
Es Ep
3454.5 3446.6
124.1 117.5
37.47 14.91
3336.8 3331.8
177.4 154.1
68.05 27.89
3186.2 3185.9
152.4 110.0
56.25 23.43
559 nm
Es Ep
3454.5 3450.3
120.5 110.4
43.70 19.35
3341.6 3334.6
158.4 157.7
72.66 43.03
3192.8 3186.0
141.9 103.3
69.69 30.18
Note: during fitting, three band centers were set as 3450 ± 5 cm was within 10 cm 1 and error of area was within 10%.
1
(except for the 9.8-nm thick film), 3340 ± 10 cm
1
, and 3195 ± 10 cm
1
. Error of FWHMs
253
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
Area of amino stretching band
100
N-H stretching mode of amino groups
Es, ss
80
Es, as Ep, ss
60
Ep, as
40
20
0 0
100
200
300
400
500
600
Thickness (nm) Fig. 7. The intensities of the amino ss and as modes in terms of the PAL film thickness.
Fig. 8. The schematic of applying different incidence angles and the ppolarized light for RA-FTIR. When the incidence angle changed from 85° to 10°, the E vector direction changes from nearly perpendicular to nearly parallel to the surface.
0.2
A
Absorbance
PAL on Au, 73 nm 85 o p-polarized light o
60 30 o, 10 o
0.1
0.2
PAL on Au, 73 nm p-polarized light
85 o 60 o 30 o, 10 o
0.1
A'
0.0
0.0 3500
3000
2500
2000
1500
60 30 o, 10 o
0.4
1800
B
PAL on Au, 297 nm o p-polarized light 85 o
Absorbance
60°, and 85°, respectively. The directions of the E vectors, as indicated by E10°, E30°, E60°, and E85° in Fig. 8, is changed from approximately parallel to perpendicular to the Au surface in the above sequence. Thus the molecular groups with transition dipole moments from parallel to perpendicular orientation can be detected with respect to the surface plane. Fig. 9 shows the RA-FTIR spectra for the 73-nm and 297-nm thick PAL films on the Au surfaces based on this scheme. It can be seen, for the 73-nm thick PAL film, all the spectra show the similar feature. For the 297-nm thick PAL film, when the incidence angle changes from 85° to 10°, intensities of the N–H ss and as modes of the amino groups increase significantly; in the meantime, the splitting of the amide I band gradually appears with increasing the low-frequency band at 1658 cm 1 and decreasing the high frequency band at 1690 cm 1. Such experimental results further solidify our argument that the spin-coated PAL thin films on top of the Au substrates are of the gradiently bi-layered structure. When the PAL thin film is thick enough, the bi-layered structure is more distinct. Most of the PAL molecules in the bottom layer attached to the Au surface is of random orientation while most of the PAL molecules in the top layer show strong parallel orientation to the surface plane, as evidenced by the increased N–H stretching vibrational modes of the amino groups and the low-frequency band at 1658 cm 1 when the s-polarized is used (Fig. 3) or the E vector is adjusted prone to be parallel to the Au surface (Fig. 9). It should be noted, inside the PAL thin films of the gradiently bi-layered structure, there should be no sharp interface between the bottom random layer and the top orientated layer. On the contrary, there should exist a smooth gradient in between, as suggested by the data in Figs. 5, 9 (the new scheme) and Table 1, where for all the PAL thin films two amide I bands can be deconvoluted, even for the very thin films.
1750
1700
1650
1600
PAL on Au, 297 nm o ~1658 85o p-polarized light
1550
B'
60 30o, 10o
0.4
0.2
0.2
0.0
0.0 3500
3000
2500
2000
Wavenumber (cm-1)
1500
1800
~1690 1750
1700
1650
1600
1550
Wavenumber (cm-1)
Fig. 9. The p-polarized spectra of the 73-nm and 297-nm thick PAL films collected with different incidence angles of 10°, 30°, 60°, and 85°.
254
X. Lu, Y. Mi / European Polymer Journal 63 (2015) 247–254
4. Conclusions Using the polarized RA-FTIR spectroscopy with the grazing angle incidence and the scheme depicted in Fig. 8, we explicitly reveal there exists a gradiently varied bi-layered structure for the spin-coated PAL thin films on top of the Au substrates. The splitting of the amide I band and the polarization-dependent infrared absorption with respect to the PAL film thickness indicate the two packing states are orientationally different. The bottom layer adjacent to the Au substrate is mainly composed of the PAL molecules randomly orientated, induced by the non-favorable interaction from the Au surface. The top layer is mainly composed of the PAL molecules with parallel orientation to the Au surface indicated by the polarization-dependent absorption of amide I band and amino stretching modes, which is induced by spin coating. Splitting of the amide I band is more significant when the film thickness is more than 102 nm, suggesting the interfacial interaction could affect orientation of the PAL molecules over the length scale of at least a molecular size. The observation described in this study is a distinct phenomenon for polymer thin films spin-coated on the solid substrates. The orientated PAL molecules induced by spin coating are supposed to be quasi-stabilized by the strong polar interaction between the PAL chain units. In this respect, the polymer thin films are metastable and the chain evolution could still happen in the long run. Conflicts of interest The authors declare no competing financial interest. Acknowledgments This study was supported by the National Natural Science Foundation of China (51173169), Scientific Research Foundation for the Returned Overseas Chinese Scholars, Ministry of Education, and Human Resources and Social Security Bureau of Zhejiang Province, China. X.L. is also grateful for the Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions, China (PAPD). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.eurpolymj.2014.12.030. References [1] Kheshgi HS, Scriven LE. Dewetting: nucleation and growth of dry regions. Chem Eng Sci 1991;46:519–26. [2] Tsui OKC. In: Tsui OKC, Russell TP, editors. Polymer thin films. Singapore: World Scientific; 2008. p. 267–94 [chapter 11]. [3] Reiter G. Mobility of polymers in films thinner than their unperturbed size. Europhys Lett 1993;23:579–84. [4] Reiter G, de Gennes PG. Spin-cast, thin, glassy polymer films: highly metastable forms of matter. Eur Phys J E 2001;6:25–8.
[5] Richardson H, Carelli C, Keddie L, Sferrazza M. Structural relaxation of spin-cast glassy polymer thin films as a possible factor in dewetting. Eur Phys J E 2003;12:437–41. [6] Reiter G, Hamieh M, Damman P, Cclavons S, Gabriele S, Vilmin T, et al. Residual stresses in thin polymer films cause rupture and dominate early stages of dewetting. Nat Mater 2005;4:754–8. [7] Gu X, Reghavan D, Douglas JF, Karim A. Hole-growth instability in the dewetting of evaporating polymer solution films. J Polym Sci, Part B: Polym Phys 2002;40:2825–32. [8] Green PF, Ganesan V. Dewetting of polymeric films: unresolved issues. Eur Phys J E 2003;12:449–54. [9] Ellison CJ, Torkelson JM. The distribution of glass-transition temperatures in nanoscopically confined glass formers. Nat Mater 2003;2:695–700. [10] Tomczak N, Vallee RAL, van Dijk EMHP, Kuipers L, van Hulst NF, Vancso GJ. Segment dynamics in thin polystyrene films probed by single-molecule optics. J Am Chem Soc 2004;126:4748–9. [11] Priestley RD, Ellison CJ, Broadbelt LJ, Torkelson JM. Structural relaxation of polymer glasses at surfaces, interfaces, and in between. Science 2005;309:456–9. [12] Habuchi S, Oba T, Vacha M. Multi-beam single-molecule defocused fluorescence imaging reveals local anisotropic nature of polymer thin films. Phys Chem Chem Phys 2011;13:7001–7. [13] Keddie JL, Jones RAL, Cory RA. Interface and surface effects on the glass-transition temperature in thin polymer films. Faraday Discuss 1994;98:219–30. [14] Keddie JL, Jones RAL, Cory RA. Europhys Lett 1994;27:59–64. [15] Sharp JS, Forrest JA. Free surfaces cause reductions in the glass transition temperature of thin polystyrene films. Phys Rev Lett 2003;91. 235701-1–4. [16] Forrest JA, Dalnok-Veress K, Dutcher JR. Interface and chain confinement effects on the glass transition temperature of thin polymer films. Phys Rev E 1997;56:5705–16. [17] Forrest JA, Dalnoki-Veress K, Stevens JR, Dutcher JR. Effect of free surfaces on the glass transition temperature of thin polymer films. Phys Rev Lett 1996;77:2002–5. [18] Forrest JA, Mattsson J. Reductions of the glass transition temperature in thin polymer films: probing the length scale of cooperative dynamics. Phys Rev E 2000;61:R53–6. [19] Wallace WE, van Zanten JH, Wu WL. Influence of an impenetrable interface on a polymer glass-transition temperature. Phys Rev E 1995;52:R3329–32. [20] van Zanten JH, Wallace WE, Wu WL. Effect of strongly favorable substrate interactions on the thermal properties of ultrathin polymer films. Phys Rev E 1995;53:R2053–6. [21] Gracias DH, Zhang D, Lianos L, Ibach W, Shen YR, Somorjai GA. A study of the glass transition of polypropylene surfaces by sumfrequency vibrational spectroscopy and scanning force microscopy. Chem Phys 1999;245:277–84. [22] Zhang C, Hong SC, Ji N, Wang YP, Wei KH, Shen YR. Sum-frequency vibrational spectroscopic study of surface glass transition of poly(vinyl alcohol). Macromolecules 2003;36:3303–6. [23] Harp GP, Rangwalla H, Yeganeh MS, Dhinojwala A. Infrared-visible sum frequency generation spectroscopic study of molecular orientation at polystyrene/comb-polymer interfaces. J Am Chem Soc 2003;125:11283–90. [24] Li Q, Hua R, Cheah IJ, Chou KC. Surface structure relaxation of poly(methyl methacrylate). J Phys Chem B 2008;112:694–7. [25] Li B, Zhou J, Xu X, Yu J, Shao W, Fang Y, et al. Solvent quality affects chain conformational order at the polymer surface revealed by sum frequency generation vibrational spectroscopy. Polymer 2013;54:1853–9. [26] Lu X, Cheng I, Mi Y. Discovery of the bi-layered structure in spincoated polyacrylamide films on the gold surface. Polymer 2007;48:682–6. [27] Greenler RG. Infrared study of adsorbed molecules on metal surfaces by reflection techniques. J Chem Phys 1966;44:310–5. [28] Allara DL, Baca A, Pryde CA. Distortions of band shapes in external reflection infrared spectra of thin polymer films on metal substrates. Macromolecules 1978;11:1215–20. [29] Buffeteau T, Desbat B, Turlet JM. Polarization modulation FT-IR spectroscopy of surfaces and ultra-thin films: experimental procedure and quantitative analysis. Appl Spectrosc 1991;45:380–9. [30] Lu X, Mi Y. Characterization of the interfacial interaction between polyacrylamide and silicon substrate by Fourier transform infrared spectroscopy. Macromolecules 2005;38:839–43. [31] Tamm LK, Tatulian SA. Infrared spectroscopy of proteins and peptides in lipid bilayers. Q Rev Biophys 1997;30:365–429.