Porphyrin-containing polyimide films deposited by high vacuum co-evaporation

European Polymer Journal 44 (2008) 3628–3639

Contents lists available at ScienceDirect

European Polymer Journal journal homepage: www.elsevier.com/locate/europolj

Porphyrin-containing polyimide films deposited by high vacuum co-evaporation G. Maggioni a,b,*, S. Carturan a,b, M. Tonezzer a, M. Buffa a,c, A. Quaranta a,c, E. Negro a,c,1, G. Della Mea a,c a

INFN Legnaro National Laboratories, University of Padua at INFN-LNL, Viale dell’Università 2, 35020 Legnaro (Pd), Italy University of Padua at INFN-LNL, Viale dell’Università 2, 35020 Legnaro (PD), Italy c University of Trento, Department of Materials Engineering and Industrial Technologies, Via Mesiano 77, 38050 Povo (TN), Italy b

a r t i c l e

i n f o

Article history: Received 8 January 2008 Received in revised form 21 July 2008 Accepted 25 August 2008 Available online 12 September 2008

Keywords: Polyimide Porphyrin Luminescent material

a b s t r a c t Thin films of porphyrin-containing polyimide were produced by high vacuum co-evaporation of 4,40 -hexafluoroisopropylidene diphthalic anhydride (6FDA), 3,30 -diaminodiphenyl sulfone (DDS) and 5,10,15,20 meso-tetraphenyl porphyrin (TPP). The films were characterized by FT-IR analysis, optical absorption and emission spectroscopy. FT-IR analysis shows that the film matrix is comprised of only unreacted monomers. The conversion of monomers to polyamic acid and the following condensation to polyimide were studied by curing the samples at temperatures up to 240 °C. The amount of polyamic acid increases from room temperature to 120 °C, while at higher temperature it starts to condense to polyimide. Optical analysis shows that TPP is incorporated in the film matrix and its chemical state is determined by the interaction with the monomers, polyamic acid and polyimide. After curing the TPP molecules are finely dispersed in the polyimide matrix and their absorption and fluorescence properties are wholly preserved. Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction Porphyrins are macrocyclic compounds which have been arousing increasing interest in the last decade due to their peculiar properties such as chemical stability, thermal resistance, synthetic versatility and biocompatibility. Particularly interesting are the optical properties of these macrocycles, which can be tailored by modifying their basic molecular structure, such as the intense absorption of visible light, the chemical stability upon illumination, the high fluorescence quantum yield in the visible region and the large Stokes’ shift.

* Corresponding author. Address: INFN Legnaro National Laboratories, University of Padua at INFN-LNL, Viale dell’Università 2, 35020 Legnaro (Pd), Italy. Tel./fax: +390498068476. E-mail address: [email protected] (G. Maggioni). 1 Present address: University of Padua, Department of Chemical Sciences, Via Marzolo 1, 35131 Padua, Italy. 0014-3057/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.eurpolymj.2008.08.045

In order to exploit the properties of porphyrins, they can be deposited as thin films by using several techniques such as solvent casting, Langmuir–Blodgett [1], spin coating, high vacuum evaporation [2] and glow discharge induced sublimation [3]. However, the aggregation of porphyrin molecules in the solid film can jeopardize some of their properties, particularly the optical emission is strongly quenched, because the aggregates act as a deep sink of excitation energies on porphyrin monomers and prevent successive photophysical processes, such as excitation energy transfer and photoinduced electron-transfer processes. To overcome this drawback, porphyrins have been dispersed in optically transparent matrices such as polymers [4–7] and glasses [8,9]. However, the optical transparency is not the only important property of the matrix, because the interaction between porphyrin and matrix can change the emission features of the former: for instance Kumar et al. [8] found that when 5,10,15,20 meso-tetraphenyl porphyrin (TPP) is dispersed in borate glasses, its fluorescence is completely lost. Moreover, when

3629

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639

porphyrin-containing materials have to be used as active elements of such devices as optical gas and chemical sensors, solar cells and radiation detectors, the matrix must have specific properties such as chemical and thermal stability, which are not common to the polymer and glass matrices used so far. In this work a fluorinated polyimide has been used as a matrix for a porphyrin compound. In spite of the outstanding properties of polyimides, very little attention has been paid to the use of these polymers as matrices for porphyrins, although their chemical structure enables one to tailor their physico–chemical properties by changing diamines and/or dianhydrides, while keeping the thermal, mechanical and chemical stability. Zhu et al. [10] tried to disperse TPP molecules into a porphyrin-containing copolyimide in order to enhance the photoconductivity of the copolyimide, but the results were poor due to the phase separation of the resulting polymer. A possible reason of this lack is that for many applications the optical absorption and the fluorescence of the porphyrins are the most important properties, while the common polyimides are not optically transparent in the wavelength range corresponding to the most intense absorption band of porphyrins (Soret band) so that they can not be used as a matrix. On the other hand, when fluorinated monomers are introduced in the polyimide chain, these polymers can become very optically transparent without losing their thermal and mechanical stability [11–13]. Although the polyimides have not been widely used as a matrix for porphyrins, several papers describe the synthesis of copolyimides containing custom-synthesized, aminophenyl porphyrins which are inserted into the polyimide chain as monomer units [14–18]. The synthesized copolyimides are then deposited as thin films by Langmuir–Blodgett or spin coating. Nevertheless, although the optical absorption and fluorescence of porphyrins incorporated in the polyimide chain can be preserved by using this synthetic route, the complexity of the production process of the films and the resulting costs are enormously increased, thus jeopardizing the large-scale application of these materials. In this work porphyrin-containing polyimide films have been produced by using commercially available chemicals: 4,40 -hexafluoroisopropylidene diphthalic anhydride (6FDA) and 3,30 -diaminodiphenyl sulfone (DDS) were chosen as the precursor monomers in order to obtain a highly optically transparent polyimide [19] and TPP was selected as the porphyrin owing to its high fluorescence yield. The porphyrin-containing polyimide films were produced by high vacuum co-evaporation of 6FDA, DDS and TPP. This deposition technique allows one to produce organic films without using any solvent and assures greater reproducibility, higher uniformity and stricter control of the film thickness in comparison with standard chemical techniques [2]. In the past the same deposition technique has been also used to produce phthalocyanine-containing polyimide films [20]. 6FDA, DDS and TPP were co-evaporated and after the deposition the films were cured at temperatures ranging from 120 to 240 °C. Since there are no previous works on the co-evaporation of 6FDA and DDS, to the best knowledge of the authors, a study of the film

matrix was mandatory and was performed by using FT-IR analysis. The optical properties of the films before and after curing were studied by measuring their light absorption and emission. The light emission of the TPP-containing films was stimulated both by directly exciting TPP molecules at wavelengths around 420 nm (Soret band) and by exciting the film matrix in the near-UV range, in order to study the chemical state of TPP molecules and the occurrence of an energy transfer between matrix and TPP. The study of the energy transfer process under excitation in the near-UV range is particularly important in view of applying the porphyrin-containing films to enlarge the spectral sensitivity of semiconductor-based radiation detectors [21] and solar cells [22]. 2. Experimental The chemical structures of monomers and porphyrin are shown in Fig. 1. 4,40 -hexafluoroisopropylidene diphthalic anhydride (6FDA) and 3,30 -diaminodiphenyl sulfone (DDS) were obtained from Lancaster at 98% purity; 5,10,15,20 meso-tetraphenyl porphyrin (TPP) was obtained from Sigma–Aldrich at 99.5% purity. All the chemicals were used without further purification. The experimental equipment used for the evaporation of the organic molecules is comprised of a stainless steel vacuum chamber evacuated by a turbomolecular pump to a base pressure of 1  104 Pa. The chamber is equipped

O

O CF3

O

O CF3

O

H2N

O

SO 2

NH2

Fig. 1. Chemical structure of 5,10,15,20 meso-tetraphenyl porphyrin (TPP), 4,40 -hexafluoroisopropylidene diphthalic anhydride (6FDA) and 3,30 -diaminodiphenyl sulfone (DDS).

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639

3. Results and discussion 3.1. Film deposition Fig. 2 shows the plots of the deposition rates of the 6FDA, DDS and TPP compounds vs the reciprocal crucible temperature. The deposition rates were evaluated as number of moles deposited in the unit of surface and time (mol cm2 s1) by normalizing the readings of the thickness sensor to the molecular weight of each compound. Each curve can be fit by the following function [23]:

log D ¼ A  B=T; where D is the deposition rate, T the absolute temperature and A and B are constants determined from the fits. A and B constants for the three compounds are reported in Table 1.

1E-9

TPP 6FDA DDS

-1 -2

with three copper crucibles that are wrapped with a heating wire. The crucibles were filled with a single compound and heated separately each other. The substrates were placed on a fixed sample holder placed 11 cm above the crucible and kept at room temperature. The deposition rate and film thickness were measured by a quartz crystal microbalance thickness sensor (Sycon) put in the centre of the sample holder. A movable shutter was put between the crucibles and the sample holder: before and after the deposition, when the crucibles were either heated up to the deposition temperature or cooled down to the room temperature, the shutter was placed between crucibles and substrates in order to prevent the uncontrolled deposition of organic molecules onto the substrates. When the deposition temperature of the three crucibles was achieved, the shutter was removed and the deposition started. The films were deposited on two different substrates: Pdoped (100) silicon wafers lapped on both faces (Bayville Chemical Co.) for FT-IR analyses and fluorescence measurements and UV-grade quartz slides (Heraeus Spa) for optical absorption measurements. The thickness of all the films studied in this work ranged from 500 to 1000 nm, as measured by a stylus profilometer. After the deposition some of the deposited films underwent heat treatments in nitrogen for 2 or 4 h at temperatures of 120, 150, 180, 210 and 240 °C. FT-IR spectra of the samples were recorded in the 4000– 400 cm1 range using a Jasco FTIR 660 Plus spectrometer with a resolution of 4 cm1. During the measurements, the sample cell and the interferometer were evacuated so as to remove from the spectra the absorption peaks of water and atmospheric gases. UV–visible absorption measurements were performed in the 250–800 nm range using a Jasco V-570 dual-beam spectrophotometer. The spectra were recorded with a resolution of 2 nm. Emission and excitation spectra at room temperature were collected by a Jasco FP-770 spectrofluorometer equipped with a 150-W xenon lamp. The spectral bandwidth was 5 nm for all the spectra. All the spectra were corrected by taking into account the spectral response of the overall detection system (emission monochromator and detection photomultiplier).

Dep. Rate (mol cm s )

3630

1E-10

1E-11

1E-12 1.5

1.6

1.7

1.8

1.9

2.0

2.1

2.2

2.3

2.4

2.5

1/T (1000/K) Fig. 2. Deposition rates of TPP, 6FDA and DDS vs reciprocal temperature.

Table 1 Fit parameters (A and B), molar ratio, deposition rate and deposition temperature of TPP, 6FDA and DDS compounds

TPP 6FDA DDS

A

B (K)

Molar ratio with respect to TPP

Deposition rate (mol cm2 s1)

Crucible temperature (K)

11 6.4 5.0

1.41  104 7.4  103 6.1  103

1 100 100

3.0  1012 3.0  1010 3.0  1010

626 464 422

The error of the fit parameters is ±10% (A) and ±5% (B).

In order to have a 1:1 stoichiometric ratio between 6FDA and DDS molecules and a 1:100 ratio between TPP molecules and the repeating units of the final polyimide, the deposition rates of TPP, 6FDA and DDS (measured in mol cm2 s1) must be in the ratio 1:100:100, respectively. Taking into account the values of the deposition rates of each compound, the deposition rate chosen for TPP was 3.0  1012 mol cm2 s1, while for each monomer the deposition rate was 3.0  1010 mol cm2 s1. The corresponding crucible temperatures were determined from the reported curves and shown in Table 1. 3.2. FT-IR analysis Fig. 3 shows the FT-IR spectra of the evaporated 6FDA and DDS monomers (curves a and b) and of the film obtained by co-evaporation of the two monomers and TPP (curve c). The positions of the most pronounced characteristic peaks in the three spectra are reported in Table 2 together with the corresponding assignments. In the spectra of the two monomers all the typical peaks of these two compounds appear thus confirming that the deposition process did not change their molecular features, in agreement with [24]. As can be seen, all the main peaks of the two monomers also appear in the spectrum of the asdeposited film, showing that unreacted monomers are incorporated in the film. The lack of any extraneous peak in the as-deposited film, especially the peaks of the polyamic acid, indicates that the reaction between the two monomers did not occur during the condensation of the sublimated molecules on the substrate, which was at room

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639

3631

nyl tetracarboxylic dianhydride (BPDA) and 6FDA, are in the following order:

c

Transmittance (a.u.)

PMDA > BTDA > BPDA  6FDA

b

a

3600

3200

2000 1800 1600 1400 1200 1000 800

600

400

-1

Wavenumber (cm ) Fig. 3. FT-IR spectra of evaporated DDS (a) and 6FDA (b) films and of coevaporated 6FDA–DDS–TPP film (c) as deposited. Wavenumber range: 3800–2900 cm1; 2000–400 cm1.

temperature. The rate of formation of polyamic acid in the as-deposited film is strictly related to the reactivity of the two monomers, i.e. the electron affinity of the dianhydride and basicity of the dianiline. From literature data [25], it is known that the electron affinities of selected dianhydrides, i.e. pyromellitic dianhydride (PMDA), 3,30 ,4,40 -benzophenone tetracarboxylic dianhydride (BTDA), 3,30 -4,40 -biphe-

In fact, in the co-evaporation of the more reactive dianhydrides PMDA and BTDA with different diamines, PMDA with 4,40 -diaminodiphenyl ether (ODA) [26] and BTDA with ODA [27], a considerable fraction of polyamic acid (50–70% for PMDA–ODA) is formed already at the end of the film deposition; on the other hand, when the less reactive BPDA dianhydride and o-tolidine (OTD) are co-evaporated, there is no evidence of reaction between the two monomers and the film is comprised of only unreacted monomers [28]. As to the TPP evaporation, the presence of TPP molecules in the as-deposited film is not highlighted by FTIR analysis. In fact, taking into account that the amount of TPP is much lower than that of the two monomers, it is expected that the typical peaks of TPP are too weak to be detected. Moreover, most TPP peaks overlap with those of the two monomers and then they could be masked. However, the integrity of the TPP molecules deposited by evaporation and their incorporation in the as-deposited film will be shown by the optical measurements (see below).

Table 2 FT-IR peaks and assignments of deposited samples: DDS and 6FDA films, 6FDA–DDS–TPP film as deposited and after curing at 120 and 240 °C for 4 h Sample DDS

Assignment 6FDA

6FDA–DDS–TPP

120 °C

240 °C

3466 3367

3473 3379

3475 3374 3330

3484 3385

3225 3067

3240 3070

3075

3078

1857 1783

1856 1784

1629 1600

1627 1600

1484 1455

1485 1454

1292

1148

1091 990 909 780 717 687 616 526

1256 1211 1194 1159 1150 1117

902 735 718 689

1301 1257 1212 1193

1858 1785 1729 1685 1628 1598 1539 1482 1452 1426 1374 1301 1257 1211

1787 1729 1626 1598 1482 1455 1436 1371 1302 1256 1210 1193 1159 1149

Asymmetric NAH stretching (DDS) Symmetric NAH stretching (DDS) NAH stretching (polyamic acid) Overtone of NAH bending band, intensified by Fermi resonance Aromatic CAH stretching (DDS) Asymmetric C@O stretching (6FDA) Symmetric C@O strecthing (6FDA); asymmetric C@O stretching (Imide I) C@O stretching (COOH group); symmetric C@O stretching (Imide I) C@O stretching (ACONH group) (Amide I) NH bending (primary amine) Aromatic CC stretching NH bending (Amide II) Aromatic CC stretching Aromatic CC stretching

CN stretching (Imide II) Asymmetric stretching SO2; CN stretching (primary amine)

1150 1117

1149

1106

CN stretching (Imide III)

1094 990 905 790 735 718 689 615 526

1099 991 903 795

986 899 796

Out-of-plane aromatic CH bending Aromatic CH bending

723 687 612 526

721 687 609 526

C@O bending; Imide IV

3632

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639

As previously observed by several authors, the structure of the as-deposited film can change during its storage in ambient air and the change rate is related to the chemistry of the reaction between the two monomers: in the case of PMDA–ODA [29,30] and BPDA–OTD [28] films, FT-IR analysis shows that the carbonyl absorptions of the dianhydride moiety at 1780 and 1850 cm1 decrease with time, while the amide absorption at 1650 cm1, which does not appear in the spectrum of the as-deposited BPDA– OTD films immediately after the deposition, increases with time. This indicates that the reaction between the monomers to form polyamic acid is proceeding at room temperature after the deposition. On the other hand, in the case of molecular beam deposited PMDA–ODA films on substrates cooled at 50 °C, the change of the structure also depends on the storage atmosphere: when the film is kept in N2 atmosphere, no change is seen after 22 days; but, if the film is stored in air, the reaction between PMDA and ODA does not proceed and hydrolysis of unreacted anhydride moieties by ambient humidity takes place instead [31]. In the case of the 6FDA–DDS system the change of the structure of as-deposited films has not been studied previously, to the best knowledge of the authors. Since this change is important for the final properties of the films, the FT-IR spectrum of the as-deposited film after 0, 2 and 8 days storage in air has been collected and reported in Fig. 4. As can be seen the storage produces a strong change of the film structure: this change consists in the reaction between 6FDA and DDS to form polyamic acid, while hydrolysis of 6FDA moieties is not significant. In fact, the carbonyl peaks of 6FDA monomer at 1857 and 1783 cm1 decrease after the storage and, in the long wavenumber region, a decrease of the absorptions of NH stretching of NH2 group of DDS monomer at 3466 and 3367 cm1 is also observed. Moreover, new peaks, which can not be ascribed to the monomers, appear in the spectrum of the as-deposited film, especially the wide peaks at 1728, 1681, 1537 e 1424 cm1 and a shoulder at 3330 cm1. These new peaks have been assigned as follows: C@O stretching of the COOH group (1728 cm1); C@O stretching of the CONH group (1681 cm1, Amide I); NH bending of the CNH group

Transmittance (a.u.)

c

b

a

3600

3200

2000 1800 1600 1400 1200 1000

800

600

-1

Wavenumber (cm ) Fig. 4. FT-IR spectra of co-evaporated 6FDA–DDS–TPP film: (a) as deposited; (b) aged 2 days; (c) aged 8 days. Wavenumber range: 3800– 2900 cm1; 2000–500 cm1.

(1537 cm1, Amide II); NH stretching of the amid group (3330 cm1). The assignment of the peak at 1424 cm1 is more difficult: in the BTDA–DMDA polyimide Ishida et al. [32] assigned it to a vibration mode of a tri-substituted phenyl ring (mode 19a of 1,2,4-C6H3); however this mode is accompanied by the mode 19 b which should appear around 1480 cm1. In the spectra of Fig. 4 the peak corresponding to the mode 19b is not evident, because it is partially masked by the CC stretching of the phenyl ring of the DDS monomer (1485 cm1). Moreover, one should expect that these peaks appear also in the spectrum of 6FDA monomer, but in this case they are too weak to be revealed. On the other hand, according to Silverstein [33], the CAN stretching of primary amides occurs near 1400 cm1: therefore the peak at 1424 cm1 could be also assigned to this vibration. Hydrolysis of 6FDA moieties by ambient humidity is not significant, because the only band of hydrated 6FDA, which can be picked out between those of the polyamic acid, i.e. the band at 1404 cm1, arising from the overlap of CO stretching and OH bending of COOH group [34], is not present. To confirm this result, the IR spectrum of the evaporated 6FDA film has been also collected after 9 days storage in air and no change was observed, suggesting that the hydrolysis rate is negligible on this time scale. In order to study the evolution of the structure of the deposited films at increasing temperature, especially the formation of polyamic acid and its condensation to polyimide, the films underwent thermal treatments at temperatures from 120 to 240 °C in nitrogen atmosphere. Fig. 5 shows the infrared spectra of heated films in the regions from 3600 to 2900 cm1, from 1910 to 1500 cm1 and from 1470 to 1040 cm1. The most intense peaks of the samples cured at 120 °C and at 240 °C for 4 h are also reported in Table 2. At the higher wavenumbers, a broad band around 3330 cm1 appears after 2 h at 120 °C giving rise to a shoulder of the peak of NH stretching of DDS (3367 cm1) and its intensity decreases at 150 °C and becomes negligible at 180 °C; since this band is the absorption of the NH stretching of the polyamic acid, this trend indicates an increase of the amount of polyamic acid in the film from room temperature to 120 °C, followed by a decrease at higher temperature. This is confirmed by the similar trend of the peaks at 1681 cm1 (Amide I) and at 1537 cm1 (Amide II), which reach the maximum intensity at 120 °C and then decrease down to disappear starting from 180 °C. With the increase of the fraction of polyamic acid, a continuous decrease of the amount of unreacted monomers in the film takes place as shown by the decrease of carbonyl absorptions of 6FDA and NH stretching absorptions of DDS. As to the condensation of polyamic acid to polyimide, the most significant absorptions of the imide group are the peaks at 1787 and 1728 cm1 (Imide I [35]), 1371 cm1 (Imide II), 1159 and 1106 cm1 (Imide III) and 721 cm1 (Imide IV). At increasing temperature, these peaks exhibit different trends: the peak at 1787 cm1 decreases at 120 °C and afterwards increases at higher temperature; the peaks at 1728 and 1371 cm1 increase at increasing temperature; the peak at 1159 cm1 appears as a shoulder at 180 °C; the peaks at 1106 and 721 cm1

3633

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639

Transmittance (a.u.)

asdep

120C 2h 120C 4h 150C 4h

180C 4h

b asdep

Transmittance (a.u.)

a

120C 2h 120C 4h 150C 4h 180C 4h 210C 4h 240C 4h

210C 4h 240C 4h

3600

3400

3200

1900

3000

1850

1800

1750

1700

1650

1600

1550

1500

-1

-1

Wavenumber (cm )

Wavenumber (cm )

c

asdep 120C 2h

Transmittance (a.u.)

120C 4h 150C 4h 180C 4h 210C 4h 240C 4h

1450

1400

1350

1300

1250

1200

1150

1100

1050

-1

Wavenumber (cm ) Fig. 5. FT-IR spectra of co-evaporated 6FDA–DDS–TPP films as deposited and cured at 120 °C (2 and 4 h), 150 °C (2 h), 180 °C (4 h), 210 °C (4 h) e 240 °C (4 h).Wavenumber range: (a) 3600–2900 cm1; (b) 1910–1500 cm1; (c) 1470–1040 cm1.

appear at 150 °C and increase at higher temperature. In order to have a clear understanding of the trend of the formation of polyimide, the peak at 1371 cm1 has been preferred for the following reasons: (i) the peaks at 1787 and 721 cm1 arise in addition to that of the carbonyl stretching (1783 cm1) and C@O bending (717 cm1) of 6FDA monomer, respectively; therefore their starting intensity decrease corresponds to the decrease of dianhydride moiety of 6FDA owing to the dianhydride ring opening and formation of polyamic acid; (ii) the peak at 1728 cm1 overlaps with that of C@O stretching of the COOH group of the amide, so that its starting increase is also partly due to an increase of polyamic acid; (iii) the peaks at 1159 and 1106 cm1 are relatively weak and overlap with medium and strong absorptions of the monomers; for this reason they are not useful at the lower temperatures when the amount of polyimide is low; (iv) the peak at 1371 cm1 does not appear in the spectra of the two monomers and of the co-evaporated as-deposited film, confirming that this absorption can be only assigned to the imide group. The experimental spectra show that at increasing temperature the intensity of this peak increases more and more. The appearance of this peak in the film heated at 120 °C indicates that the condensation of polya-

mic acid to polyimide is proceeding ever since this temperature. Therefore, to sum up, the reaction between 6FDA and DDS monomers in the film to form polyamic acid proceeds until 120 °C, where the film is mainly comprised of polyamic acid and only few unreacted monomers remain; afterwards the polyamic acid turns itself to polyimide. So far the description of the change of the structure of the heated films has been only qualitative. In order to have a semiquantitative measurement of the imidization level of the heated films and of the polyamic acid evolution, the integral areas of the imide peak at 1371 cm1 (A1371norm) and of the amide peak at 1537 cm1 (A1537norm) have been evaluated: in Fig. 6 the areas are reported versus heating temperature. The areas have been normalized to the area of the sample heated at 240 °C (imide peak) and at 120 °C for 4 h (amide peak), respectively. The graph shows two different stages of the imidization reaction: first the increase of the imidization level is fast between 120 and 180 °C and then the imidization rate decreases considerably as the reaction proceeds. Taking into account that the sample area probed by the infrared beam during these measurements was not always the same, since the sample should be removed and put in the oven, the adopted procedure could suffer from some error in the integral area

3634

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639

3.3. UV–visible analysis

1.2 norm

A1537 Normalized Peak Area

A1371

norm

norm

norm

norm

A1537 /A 795 0.8

A1371 /A 795

0.6

0.4

0.2

0.0 0

50

100

150

200

250

T (°C) Fig. 6. Trend of the normalized integral areas of the peaks of polyamic acid (A1537norm) and polyimide (A1371norm) in the co-evaporated 6FDA– DDS–TPP film vs curing temperature. The same areas have been also divided by the area of the internal reference peak at 795 cm1 (A795norm).

evaluation due to the different sample position. In order to remove this uncertainty, the area of the selected peaks has been also divided by that of an internal reference peak, which should not change during heat treatments. A survey of the experimental spectra shows that the best choice is the peak at 795 cm1, assigned to the CAH out-of-plane bending vibration of the aromatic ring of DDS moiety: in fact this peak does not depend on the substitution of the amine group after polyamic acid formation and following imidization and it does not overlap with different peaks. This choice is also strengthened by a previous work [36], where the same peak was used for the measurement of the imidization level of a different polyimide. The remaining peaks, which should not change after heat treatment, either overlap with different peaks or are placed in the low-signal region of the infrared spectrometer (peak at 526 cm1); for these reasons they have been ruled out. Fig. 6 shows a substantial agreement in the trend of the curves calculated in the two different ways. A final remark concerns the possible appearance of isoimide in the heated films: according to Karamancheva et al [37] an excess of diamine in the co-evaporation of PMDA– ODA films causes formation of isoimide groups. The presence of isoimide structure is readily identified by FT-IR analysis: in fact isoimides show strong absorptions in the region 1795–1820 cm1 and 921–934 cm1 and a medium absorption around 1700 cm1 [38]. In the case of PMDA– ODA polyimide, Salem et al [26] found that all these IR peaks were quite small so that the concentration of isoimide in the film was expected to be no more than few percents. In our spectra the peaks between 1795 and 1820 cm1 and around 1700 cm1 overlap with those of the carbonyl stretching of 6FDA, polyamic acid and polyimide so that their presence can not be determined for a certainty. On the other hand, in the region 921–934 cm1 no peak appears in the experimental spectra and after the treatment at 180 °C no peak is visible around 1700 cm1: it is then inferred that the formation of isoimide groups in our films can be ruled out.

3.3.1. Absorption spectra of as deposited films Fig. 7 shows the UV–visible absorption spectra of TPP dissolved in CHCl3 (106 M) and of the as-deposited films of 6FDA, DDS, TPP, co-evaporated 6FDA–DDS, co-evaporated 6FDA–DDS–TPP zero and eight days after the deposition, respectively. The absorption spectra of both evaporated monomers are very similar to those of the monomers dissolved in tetrahydrofuran (not reported), confirming that the deposition process did not change their optical features. For both monomers the absorption is limited to the UV region: the absorption peaks of the dissolved monomers are at 250 and 300 nm (6FDA) and at 250 and 317 nm (DDS). In the evaporated films the peaks at lower wavelengths can not be distinguished due to saturation effects, while the peak at 317 nm can be identified in the DDS film. When the monomers are co-evaporated, the absorption does not change: it drops steeply around 300 nm, with a small tail up to 350 nm, and becomes negligible at wavelengths higher than 350 nm. The almost complete transparency of the co-evaporated monomers in the visible region is particularly important for the TPPcontaining film, because it makes the study of the optical absorption and emission of TPP molecules embedded in the film easier. The absorption spectrum of TPP compound consists of two main transitions: a strong transition to the

a

4

b

2

c 0 4

Absorbance (a.u.)

norm

1.0

b

2

a 0 4

b a 2

0 200

300

400

500

600

700

800

Wavelength (nm) Fig. 7. Optical absorption spectra of films and solutions: Top: (a) evaporated 6FDA film; (b) evaporated DDS film; (c) co-evaporated 6FDA–DDS film. Middle: (a) evaporated TPP film; (b) TPP dissolved in CHCl3 (106 M). Bottom: co-evaporated 6FDA–DDS–TPP film: (a) as deposited (no aging); (b) aged 8 days.

3635

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639

In the case of TPP, the acid dication (TPPH22+) exhibits a strong B band around 440 nm, a medium Q band around 655 nm and a weak Q band at 605 nm [45,47,48,41]. In the figure the spectrum of the film after 8 days in air has been also reported for comparison. As can be seen, the spectrum has strongly changed: the B band red-shifted to 442 nm and the four Q band components are partially lost, because only an intense peak at 657 nm and a weak band at 604 nm appear. Taking into account these features, the close similarity between this spectrum and that of the TPPH22+ dication indicates that most TPP molecules in the film have been protonated and appear as dications. This conclusion is also supported by the fluorescence spectra reported below. The protonation of the TPP molecule occurs by means of the polyamic acid which is formed in the film after the deposition as highlighted by the infrared spectra of aged films (Fig. 4). 3.3.2. Absorption spectra of cured films Fig. 8 shows the absorption spectra of 6FDA–DDS and 6FDA–DDS–TPP films cured at 180 °C for 4 h. For both films curing treatment was performed some days after the deposition, when the films were comprised of a mixture of unreacted monomers and polyamic acid. As can be seen, the film without TPP is completely transparent in all the visible range, as could be expected. The spectrum of 6FDA–DDS film is also useful to study the possible presence of isoimide in the cured films, because this group has a strong absorption around 350–400 nm [38]: since the absorption does not appear in the spectrum, the lack of isoimide as shown by infrared analysis is confirmed. In the TPP-containing film the absorption peaks fall at 422 (B band) and 518, 553, 595 and 648 nm (Q band). The appearance of this spectrum highlights that the four Qbands structure has been recovered after the transformation of polyamic acid into polyimide; moreover, the position of the B and Q bands is very close to that of TPP dissolved in CHCl3. This finding indicates that the conversion of polyamic acid to polyimide caused the reversion of the dication to the neutral monomeric form. Moreover

Absorbance (a.u.)

second excited state (S0 ? S2) in the wavelength range 400–500 nm (Soret or B band) and a weak transition to the first excited state (S0 ? S1) in the range 500–700 nm (Q band). The B and Q bands arise from p–p* transitions and can be explained by means of the ‘‘Gouterman four orbital model” [39]. The Q band is comprised of four components (Qy(1, 0), Qy(0, 0), Qx(1, 0) and Qx(0, 0)) arising from the low symmetry of the TPP molecule (D2h). According to the literature, the TPP dissolved in chloroform exhibits the B band at 418 nm, while the Q band components are at 515 nm (Qy(1, 0)), 550 nm (Qy(0, 0)), 590 nm (Qx(1, 0)) and 646 nm (Qx(0, 0)). In the near-UV range the bands N and L [40] are too weak to be clearly distinguished. In the spectrum of the as-deposited 6FDA–DDS–TPP film measured immediately after the deposition the incorporation of TPP molecules in the film is shown by the absorption in the region 400–500 nm (B band) and by the weaker absorption between 500 and 700 nm (Q band). The B band falls at 422 nm (B band) and the four components of the Q band at 518, 553, 595 and 648 nm. The position and the relative intensity of the B and Q bands are very close to those of TPP dissolved in CHCl3. Since it is well known that the absorption spectra of the porphyrins are strictly related to the aggregation of the molecules in the solid phase and to the interactions with the surrounding environment, this finding indicates that: (i) the TPP molecules are finely dispersed in the monomer matrix; (ii) the interaction with the matrix does not change the chemical structure of the molecules; (iii) the slight red shift (3– 5 nm) of the B and Q bands of TPP in the film with respect to TPP in solution is due to a matrix effect. In fact aggregation effects can be excluded for two reasons: (i) the amount of TPP molecules in the film is very low as compared to that of the two monomers; therefore, since the TPP molecules can not move to give rise to aggregates, it is expected that they are finely dispersed in the film; (ii) the absorption peaks of TPP aggregates are shifted with respect to those found in the current spectrum: for J-type aggregates the B band (475 nm) and the Q band (725 nm) are much more red-shifted [41], while a blue shift is generally found in H-type aggregates of complex organic molecules [42,43]. When the aggregation is particularly pronounced, like in the evaporated film of only TPP, a red shift and a broadening of all the absorption bands are also observed: the shift is about 15 nm, while the full width at half maximum (FWHM) of the B band increases from 12 nm for the TPP dissolved in CHCl3 to 69 nm for the evaporated TPP film. On the other hand, it is also well known that the interaction of porphyrin molecules with acids and bases can strongly change their chemical structure and hence their absorption features [44]. When freebase porphyrins are dissolved in an acidic solution, the spectral pattern changes from the four Q-band spectrum, indicating D2h symmetry for free-base porphyrin, to a two Q-band spectrum, indicating D4h symmetry, characteristic of porphyrin coordinated to a metal ion through the four N-heteronuclei. In addition, the B band is red-shifted [45]. These changes in absorption spectra are generally ascribed to the attachment of protons to the two imino nitrogen atoms of the pyrrolenine-like ring in the free-base porphyrin, with the formation of the acid dication [46].

2

b

a 0 300

400

500

600

700

Wavelength (nm) Fig. 8. Optical absorption spectra of cured films: (a) 6FDA–DDS; (b) 6FDA–DDS–TPP.

3636

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639

the close resemblance between the spectra of TPP-containing film and TPP dissolved in CHCl3 shows that the thermal treatments do not affect the optical properties of TPP molecules and do not induce their aggregation. The lack of aggregation in the film is also well highlighted by the narrowness of the B band: the FWHM is only 20 nm, which is much lower than that of the evaporated film of only TPP (FWHM = 69 nm). The strong absorption of these films in a relatively narrow wavelength range make them interesting for such applications as the production of wavelengthselective filters and optical filter coatings of LCD displays [49]. 3.3.3. Excitation and emission spectra of as deposited films In order to study the emission features of the TPP-containing films it should be taken into account that the two monomers, the polyamic acid and the final polyimide are not an optically passive matrix for the TPP but, as shown in previous works [24,19], they exhibit characteristic fluorescence spectra. For this reason the resulting fluorescence of the TPP-containing film can be the sum of several contributions. In order to try to distinguish the different contributions, one has to consider that the two monomers, the polyamic acid and the polyimide absorb only below 400 nm, while for TPP the main absorption bands fall above this wavelength. Therefore, to insulate the response of TPP, the excitation wavelength must be higher than 400 nm. On the other hand, working below 400 nm means to excite both TPP and the matrix. In the case of TPP, the absorption bands below 400 nm (N, L and M) are very weak as compared to the B band [40] and, taking into account the low amount of TPP molecules in the film with respect to that of the two monomers, the TPP contribution to the total absorption in this region is expected to be small. To have a rough estimate of the TPP contribution to the total absorption, it is useful to compare the absorption spectra of TPP and 6FDA–DDS polyimide. Fig. 9 shows the spectra of TPP and 6FDA–DDS polyimide separately dissolved in dimethylformamide (DMF) together with the spectrum of pure DMF. The molar concentrations of the two solutions were 105 M (TPP) and 103 M (6FDA–

b

3

Absorbance (a.u.)

c

2

1

0

a 300

400

500

600

700

Wavelength (nm) Fig. 9. Optical absorption spectra of: (a) pure dimethylformamide (DMF); (b) 6FDA–DDS dissolved in DMF (103 M); (c) TPP dissolved in DMF (105 M).

DDS); these values were chosen taking into account that the molar ratio between TPP and 6FDA–DDS in the TPPcontaining film is about 1:100. Since, the volume of solution probed by the optical beam was the same in the three cases, Fig. 9 shows that the TPP absorption is lower than that of PI between 340 and 260 nm; below 260 nm the contribution of DMF becomes predominant. Therefore, the excitation wavelength should be chosen in this range: in order to minimize the absorption of TPP with respect to PI, a good choice is 300 nm because at this wavelength the contribution of TPP to the total absorption is less than 10%. This value is close to the wavelength chosen by Xu et al. [16], which aimed at investigating the energy transfer from polyimide chain to porphyrin in a TPP-containing copolyimide film. Taking into account these considerations, in this work the TPP-containing films have been always excited at 300 nm and at the wavelength corresponding to the Soret band of TPP (from 420 to 440 nm, depending on the sample). First we have to consider the emission of monomers/PA/PI: a previous work [24] showed that DDS compound dissolved in THF, excited between 280 and 350 nm, exhibits an emission spectrum with a peak centred at 385 nm, which does not depend on the solution concentration and which was ascribed to the monomer. The same peak, but blue-shifted, was also found in the evaporated DDS film. As to 6FDA, the emission spectra are more complex, because it forms aggregates which have different emission features with respect to the monomer: by exciting at 270 and 330 nm, one notices an emission peak at 312 nm (monomer) and some peaks at 380 and 415 nm (aggregates). In the evaporated 6FDA film, the emission spectrum exhibits a wide band between 400 and 570 nm, ascribed to molecular aggregates. Coming to the TPP, this compound has a strong emission in the visible region. The peak position and intensity depend on several factors such as the aggregation state, the chemical state and the chemical environment. TPP dissolved in CHCl3 (kex = 420 nm) exhibits two peaks at 650 and 720 nm [50]. These peaks correspond to the transition from the first excited singlet state S1 to the ground state S0. By exciting the TPP at 400 nm, a very weak peak is also found at 432 nm [51], assigned to the transition from the second excited singlet state S2 to S0, corresponding to the Soret band of the absorption spectrum. In the as-deposited 6FDA–DDS–TPP film measured immediately after the deposition (Fig. 10) the emission spectrum shows a strong peak at 655 nm with a shoulder at 714 nm, for both excitation wavelengths (300 and 426 nm). No fluorescence peaks arising from the monomers appear, maybe due to their too low intensity. As highlighted by the excitation spectrum the TPP emission is especially high when the excitation wavelength corresponding to the Soret band is used. This is a feature common to all the TPP-containing films studied in this work, both before and after aging and/or curing. Therefore, in order to obtain the most intense emission from the deposited films, this excitation wavelength should be chosen. The partial conversion of the monomers in polyamic acid occurring after the film deposition causes a strong change of the emission spectra (Fig. 11): 13 days after

3637

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639

of the 6FDA–DDS film without TPP (Fig. 11, bottom), confirming that it is not a TPP emission. The protonation of TPP molecules by means of the polyamic acid, already shown by the absorption spectra, is confirmed by the appearance of the band at 680 nm, in intermediate position between the emission peaks of TPP neutral form. This band is typical of the TPPH22+ dication [45].

Excitation/Emission Yield (a.u.)

a b

6

4

2

c

0 300

400

500

600

700

Wavelength (nm) Fig. 10. (a) Excitation spectra of as deposited 6FDA–DDS–TPP film (kem = 660 nm); Emission spectra of as deposited 6FDA–DDS–TPP film: (b) kex = 426 nm; (c) kex = 300 nm (the ordinate of this spectrum is magnified five times (5)).

4

3

b

3.3.4. Excitation and emission spectra of cured films Excitation and emission spectra of cured films are shown in Fig. 12. For the excitation spectra the emission wavelength was kem = 660 nm and for the emission spectra the excitation wavelengths were kex = 300 and kex = 420 nm. For the pure 6FDA–DDS film the excitation spectrum was also collected at kem=440 nm in order to study the emission of the polyimide matrix. The latter spectrum shows an intense band peaked at about 300 nm. The 6FDA–DDS film excited at 300 nm exhibits a broad band between 400 and 600 nm, which is typical for this polyimide [19], with two maxima at 447 and 517 nm. On the other hand, by exciting at 420 nm only a weak band centred at 520 nm appears and no emission is found above 600 nm. In the 6FDA–DDS–TPP film the emission spectrum excited at 420 nm shows the two typical peaks of the TPP molecule. The peaks are placed at 655 and 719 nm. The po-

a c

c

16

1 12 0 8

a Excitation/Emission Yield (a.u.)

Excitation/Emission Yield (a.u.)

2

c

6

4

2

4

d

a b

0

a 6

c

b 0

4 300

400

500

600

700

Wavelength (nm) Fig. 11. Top: (a) Excitation spectrum of aged 6FDA–DDS–TPP film (kem = 660 nm); Emission spectra of aged 6FDA–DDS–TPP film: (b) kex = 440 nm; (c) kex = 300 nm. Bottom: (a) Excitation spectrum of aged 6FDA–DDS film (kem = 480 nm); Emission spectra of aged 6FDA–DDS film: (b) kex = 420 nm; (c) kex–300 nm.

2

b 0 300

400

500

600

700

Wavelength (nm)

the deposition the emission spectrum (kex = 300 nm) shows a very wide band between 350 and 600 nm. This band is the emission of the film matrix: in fact the polyamic acid has a much stronger emission than the unreacted monomers due to the formation of charge transfer complexes (CTCs) [52]. The same band appears in the spectrum

Fig. 12. Top: Excitation spectra of: (a) cured 6FDA–DDS–TPP film (kem = 660 nm); (b) TPP dissolved in CHCl3 (106 M) (kem = 660 nm); Emission spectra of cured 6FDA–DDS–TPP film: (c) kex = 420 nm; (d) kex = 300 nm (the ordinate of this spectrum is magnified four times (4)). Bottom: (a) Excitation spectrum of cured 6FDA–DDS film (kem = 440 nm); Emission spectra of cured 6FDA–DDS film: (b) kex = 420 nm; (c) kex = 300 nm.

3638

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639

sition and the relative intensity of these peaks show that the TPP molecules are in the neutral, monomeric form. This finding confirms that after curing the dication changed to the neutral molecule and that no aggregation of TPP molecules occurred. It is interesting to observe the emission spectrum excited at 300 nm: the emission of the polyimide matrix is shown by the appearance of the broad band between 350 and 550 nm. On the other hand, the two emission peaks of TPP also appear, although their intensity is lower as compared to the intensity of the same peaks when the excitation wavelength is 420 nm. In order to study in more detail the occurrence of the energy transfer process between polyimide and TPP when the film is excited at 300 nm, the excitation spectrum of TPP dissolved in CHCl3 (106 M, kem = 660 nm) has been also reported in Fig. 12. As can be seen, in the latter spectrum the yield at 300 nm is very low as compared to that of the Soret band, i.e. I300/I417 = 0.05. On the other hand, this ratio becomes almost five times higher in the 6FDA–DDS–TPP film (I300/ I426 = 0.24). Taking into account that the excitation spectrum of 6FDA–DDS film (kem = 440 nm) exhibits an intense band at about 300 nm, this shows that the energy transfer between polyimide and TPP occurred. The energy transfer process is still less effective than the direct excitation of the Soret band of TPP, as shown by the low value of the ratio I300/I426 (less than unit), maybe due to non-radiative de-excitation processes of the polyimide chains. 4. Conclusions Porphyrin-containing polyimide films have been produced by co-evaporating 4,40 -hexafluoroisopropylidene diphthalic anhydride (6FDA), 3,30 -diaminodiphenyl sulfone (DDS) and 5,10,15,20 meso-tetraphenyl porphyrin (TPP) and by curing the as-deposited samples to complete the polyimide condensation reaction. The as-deposited films are comprised of only unreacted monomers and finely dispersed TPP molecules as shown by FT-IR and UV–visible analyses. The spontaneous reaction of 6FDA and DDS monomers to form polyamic acid occurs at room temperature in ambient air immediately after the deposition and proceeds until completion. The polyamic acid formed in the film interacts with the TPP molecules, which are originally in the neutral form, giving rise to their protonation and to the formation of TPPH22+ dications. The conversion of monomers to polyamic acid and the condensation of polyamic acid to polyimide have been also studied by curing the samples at temperatures from 120 to 240 °C. At increasing temperature, the amount of polyamic acid increases from room temperature up to 120 °C and afterwards it decreases and starts to condense to polyimide. The curing procedure in the temperature range from 120 to 240 °C gives rise to the conversion of polyamic acid to polyimide, which causes the TPPH22+ dications to reverse to the neutral form as shown by the absorption and fluorescence spectra. Optical analysis highlights that curing does not induce any damage in the TPP-containing films and that the aggregation of TPP molecules in the cured films does not occur. This last feature is an important goal as it allows TPP molecules to preserve their optical

properties, in fact: (i) self quenching of the fluorescence intensity due to aggregation is prevented; (ii) the broadening of the absorption peaks, particularly the Soret band, does not occur. The occurrence of an energy transfer process between polyimide matrix and TPP has been pointed out by exciting the films in the near-UV range. The energy transfer is less effective than the direct excitation of TPP; therefore, when the emitted light has to be maximized, the excitation of the Soret band must be preferred. For instance, in optical gas and chemical sensors where LEDs are used as light sources, the excitation of the Soret band allows to enhance the signal/noise ratio and then to improve the detection limit of the sensing device. On the other hand, in applications like radiation detection and solar cell power generation, the collection of light from the near-UV range becomes particularly important. As a final remark, without forgetting the preliminary nature of the results of this work, it is envisaged that the remarkable thermal stability and chemical resistance of both porphyrin and polyimide make these films interesting candidates for applications where harsh conditions are used like in optical gas and chemical sensors, radiation detectors and solar cells. Acknowledgment This research was financially supported by the Fifth Commission of Istituto Nazionale di Fisica Nucleare (DEGIMON project). References [1] D’Amico A, Di Natale C, Paolesse R, Macagnano A, Mantini A. Sens. Actuators B 2000;65:209. [2] Tonezzer M, Quaranta A, Maggioni G, Carturan S, Della Mea G. Sens. Actuators B 2007;122:620. [3] Tonezzer M, Maggioni G, Quaranta A, Carturan S, Della Mea G. Sens. Actuators B 2007;122:613. [4] Sayo K, Deki S, Noguchi T, Goto K. Thin Solid Films 1999;349:276. [5] Yang R-H, Wang K-M, Xiao D, Yang X-H, Li H-M. Anal. Chim. Acta 2000;404:205. [6] Miura A, Yanagawa Y, Tamai N. J. Microsc. 2001;202:401. [7] Itagaki Y, Deki K, Nakashima S-I, Sadaoka Y. Sens. Actuators B 2005;108:393. [8] Kumar GA, Thomas V, Jose G, Unnikrishnan NV, Nampoori VPN. Mater. Chem. Phys. 2002;73:206. [9] Guo L, Zhang W, Xie Z, Lin X, Chen G. Sens. Actuators B 2006;119:209. [10] Zhu B-K, Xu Z-K, Xu Y-Y. Eur. Polym. J. 1999;35:77. [11] Quaranta A, Carturan S, Maggioni G, Della Mea G, Ischia M, Campostrini R. Appl. Phys. A 2001;72:671. [12] Li H, Liu J, Wang K, Fan L, Yang S. Polymers 2006;47:1443. [13] Jang W, Shin D, Choi S, Park S, Han H. Polymers 2007;48:2130. [14] Iwamoto M, Xu X. Thin Solid Films 1996;284–285:936. [15] Ogi T, Kinoshita R, Ito S. J. Coll. Interf. Sci. 2005;286:280. [16] Xu Z-K, Zhu B-K, Xu Y-Y. Chem. Mater. 1998;10:1350–4. [17] Ohkita H, Ogi T, Kinoshita R, Ito S, Yamamoto M. Polymers 2002;43:3571. [18] Anannarukan W, Tantayanon S, Zhang D, Aleman EA, Modarelli DA, Harris FW. Polymers 2006;47:4936. [19] S. Carturan, Ph.D. Thesis, University of Trento, Italy, 2003. [20] Sakakibara Y, Matsuhata H, Tani T. Jpn. J. Appl. Phys. 1993;32:L1688. [21] Hayashida M, Mirzoyan R, Teshima M. Nucl. Instrum. Meth. Phys. Res. A 2006;567:180. [22] Strümpel C, McCann M, Beaucarne G, Arkhipov V, Slaoui A, Svrcek V, et al. Solar Energy Mater. Solar Cells 2007;91:238. [23] Pethe RG, Carlin CM, Patterson HH, Unertl WN. J. Mater. Res. 1993;8:3218. [24] Maggioni G, Quaranta A, Negro E, Carturan S, Della Mea G. Chem. Mater. 2004;16:2394.

G. Maggioni et al. / European Polymer Journal 44 (2008) 3628–3639 [25] Okamoto K-I, Tanaka K, Katsube M, Suboka O, Ito Y. Radiat. Phys. Chem. 1993;41:497. [26] Salem JR, Sequeda FO, Duran J, Lee WY, Yang RM. J. Vac. Sci. Technol. A 1986;4:369. [27] Yanagisita H, Kitamoto D, Haraya K, Nakane T, Tsuchiya T, Koura N. J. Membr. Sci. 1997;136:121. [28] Ukishima S, Iijima M, Sato M, Takahashi Y, Fukada E. Thin Solid Films 1997;308–309:475. [29] Takahashi Y, Iijima M, Inagawa K, Itoh A. J. Vac. Sci. Technol. A 1987;5:2253. [30] Iijima M, Takahashi Y. Macromolecules 1989;22:2944. [31] Dimitrakopoulos CD, Machlin ES, Kowalczyk SP. Macromolecules 1996;29:5818. [32] Ishida H, Huang MT. Spectrochim. Acta 1995;51A:319. [33] Silverstein RM, Bassler GC, Morrill TC. Spectrometric indentification of organic compounds. 4th ed. New York: J. Wiley & Sons; 1981. p. 126. [34] Silverstein RM, Bassler GC, Morrill TC. ibid., p. 121. [35] Ishida H, Wellinghoff ST, Baer E, Koenig JL. Macromolecules 1980;13:826. [36] Navarre M. In: Mittal KL, editor. Polyimides: Synthesis characterization and applications. New York: Plenum Press; 1984. p. 438.

3639

[37] Karamancheva I, Stefov V, Soptrajanov B, Danev G, Spasova E, Assa J. Vibrat. Spectr. 1999;19:369. [38] Takekoshi T. In: Ghosh MK, Mittal KL, editors. Polyimides: Fundamentals and applications. Marcel Dekker, Inc.; 1996. p. 21. [39] Anderson HL. Chem. Commun. 1999:2323. [40] Gouterman M. In: Dolphin D, editor. The Porphyrins, vol. 3. New York: Academic Press; 1978. p. 12. [41] Okada S, Segawa H. J. Am. Chem. Soc. 2003;125:2792. [42] Maiti NC, Ravikanth M, Mazumdar S, Periasamy N. J. Phys. Chem. 1995;99:17192. [43] Valdes-Aguilera O, Neckers DC. Acc. Chem. Res. 1989;22:171. [44] Erdman JG, Corwin AH. J. Am. Chem. Soc. 1946;68:1885. [45] Akins DL, Zhu H-R, Guo C. J. Phys. Chem. 1996;100:5420. [46] Gouterman M. ibid., p. 17. [47] Meot-Ner M, Adler AD. J. Am. Chem. Soc. 1975;97:5107. [48] Sun X, Zhang J, He B. J. Photochem. Photobiol. A: Chem. 2005;172:283. [49] Ohishi T, Non-Cryst J. Solids 2003;332:80. [50] Seybold PG, Gouterman M. J. Mol. Spectrosc. 1969;31:1. [51] Shi Y, Zheng W, Li Z, Wang X, Wang D, Qiu S, et al. Opt. Mater. 2006;28:1178. [52] Hasegawa M, Horie K. Prog. Polym. Sci. 2001;26:259.