Matthias
Edler
ab,
Stefan
Mayrbrugger
a,
Alexander
Fian
c,
Gregor
Trimmel
d,
Simone
Radl
e,
Wolfgang
Kern
ae and
Thomas
Griesser
*abe
aChair of Chemistry of Polymeric Materials, University of Leoben, Otto-Glöckel-Strasse 2, A-8700 Leoben, Austria. E-mail: thomas.griesser@unileoben.ac.at; Tel: +43-3842-402-2358
bChristian Doppler Laboratory for Functional and Polymer based Ink-Jet Inks, University of Leoben, Otto-Glöckel-Strasse 2, A-8700 Leoben, Austria
cInstitute for Surface Technologies and Photonics, Joanneum Research GmbH Weiz, Franz-Pichler-Strasse 30, A-8160 Weiz, Austria
dInstitute for Chemistry and Technology of Materials, Christian Doppler Laboratory for Nanocomposite-Solar Cells, Graz University of Technology, Stremayrgasse 9, A-8010 Graz, Austria
ePolymer Competence Center Leoben GmbH, Roseggerstrasse 12, A-8700 Leoben, Austria
First published on 26th April 2013
Patterning and tuning of the refractive index in polymers by means of UV-light is of great interest for optical applications such as polymeric waveguides or optical data storage devices. In this contribution, we report on the synthesis of a polynorbornene based copolymer bearing ortho-nitrobenzyl and phenyl ester groups capable of undergoing the photo-induced cleavage reaction and in a subsequent step optionally the photo-Fries rearrangement upon irradiation with UV-light. The photoreaction of these aromatic ester groups was investigated by means of FTIR, UV-VIS and spectroscopic ellipsometry. Due to the difference in UV absorption of the photoreactive units, the o-nitrobenzyl ester can be selectively excited by UV-light above 300 nm, while a subsequent illumination with 254 nm induces the photo-Fries rearrangement of the remaining phenyl ester groups. The structural changes in the chemical composition upon UV illumination lead to significant changes in the refractive index of the polymeric materials. Whilst the photo-Fries rearrangement of the phenyl ester groups causes a remarkable increase in the refractive index, it was found that the photo-induced cleavage reaction of ortho-nitrobenzyl ester moieties causes a decrease of the refractive index. This fact enables a selective tailoring of the refractive index by the choice of the applied wavelength. In addition, a two-step irradiation procedure using a sequence of different wavelengths provides the possibility of erasing and even inverting the index contrast.
Such photosensitive groups are capable of undergoing defined photoreactions inducing structural changes in the chemical composition. As a consequence a shift in the refractive index is caused. Besides these photoreactive materials, other approaches are based on photo-polymerizable groups such as acrylates or methacrylates.4
In our previous work we reported on photo-induced refractive index changes in polymers containing aromatic ester groups that undergo the irreversible photo-Fries rearrangement upon irradiation with UV-light. Refractive index changes in the range of Δn = +0.03 up to Δn = +0.08 were obtained in these polymeric materials.3c,5 In polymers bearing aryl ester groups the amount of the photoproduct directly determines the value of the refractive index change enabling a selective control by means of the applied UV dose. In this context it was demonstrated that in a copolymer bearing both naphthyl and phenyl ester groups the refractive index can be modulated using UV-light of different wavelengths.5 Due to the different UV absorption behavior of both chromophores, the naphthyl ester group can be selectively excited with UV-light of 313 nm causing an increase in the refractive index by Δn = +0.010, whilst an illumination with 254 nm leads to a photoreaction in both esters (Δn = +0.036). Interestingly, when already irradiated films of this copolymer (with UV-light of 313 nm) are subjected to a second illumination step with 254 nm, a significant decrease of the refractive index is observed. The reason for this behavior seems to arise from the formation of carboxylic acid groups, which are known as side products in the photo-Fries rearrangement.
In the present study we combine this photochemistry with that of photosensitive ortho-nitrobenzyl ester groups which undergo a defined photoreaction to yield carboxylic acid groups upon irradiation with UV-light, thus influencing the refractive index of the polymeric material. In the past years such photo-sensitive ortho-nitrobenzyl ester derivatives have gained tremendous attention in the area of polymer chemistry and beyond.6 Some recent developments deal with cross-linkers for photodegradable hydrogels,7 photo-cleavable block copolymers,8 stimuli responsible micelles,9 photopatterning of polymeric10 and inorganic surfaces,11 and photosensitive interfacial layer in organic electronics.12
The generally accepted photoisomerization mechanism of o-nitrobenzyl esters into the corresponding carboxylic acid group proceeds via a radical mechanism (see Scheme 1) based on a Norrish-type II reaction. Upon irradiation with UV-light an n–π transition occurs. The excited singlet state is transformed into a triplet state. An oxygen of the nitro group abstracts a proton from the methylene carbon in the γ-H position and an aci-nitro intermediate is formed. The subsequent rearrangement and cleavage of the resonance stabilized five-membered ring rapidly regenerates the parent carboxylic acid and affords nitroso aldehyde as an additional product.13
Scheme 1 Reaction mechanism of ortho-nitrobenzyl esters upon irradiation with UV-light (ref. 13). |
In the present contribution, a copolymer bearing ortho-nitrobenzyl ester and phenyl ester groups, capable of generating carboxylic acids (via the photocleavage of nitrobenzyl ester groups) and hydroxy ketones (via the photo-Fries rearrangement), is investigated with respect to a wavelength selective refractive index modulation. The difference in UV absorption enables a selective excitation of the ortho-nitrobenzyl ester groups by using UV-light >300 nm, while a subsequent illumination with 254 nm induces a photo-Fries rearrangement of the remaining phenyl ester groups.
Endo,exo-di(2-nitrobenzyl)bicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate (mono- 1 ): (2-nitrophenyl)methanol (2.42 g; 15.8 mmol) was added to a stirred solution of bicyclo[2.2.1]hept-5-ene-2,3-dicarbonyl dichloride (1.56 g; 7.18 mmol) and pyridine (1.24 mL; 15.8 mmol) in dichloromethane. The reaction was stirred for 48 h at ambient temperature. The reaction mixture was filtered to remove the pyridinium salt and extracted with dichloromethane. The organic layer was extracted with 3 × 20 mL of 5% hydrochloric acid solution to remove excess pyridine, and with 20 mL of saturated sodium bicarbonate solution, and then dried over sodium sulphate. The solvent was removed in a vacuum and subsequently a column separation using cyclohexane/ethyl acetate (10:1) for product purification was performed. Yield: 2.69 g of a white solid (83% of theoretical yield). 1H-NMR: (δ, 400 MHz, 20 °C, CDCl3): 8.09 (m, 2H, ph3); 7.63 (m, 2H, ph4); 7.58 (m, 2H, ph5); 7.49 (d, 2H, ph6); 6.30, 6.11 (m, 2H, nb5,6); 5.55, 5.47 (m, 4H, O–CH2-ph); 3.51 (m, 1H, nb3); 3.36 (s, 1H, nb4); 3.21 (s, 1H, nb1); 2.83 (m,1H, nb2); 1.67, 1.54 (d, 2H, nb7) ppm. 13C-NMR: (δ, 125 MHz, 20 °C, CDCl3): 173.6, 172.4 (2C, COO); 137.71 (2C, ph2-NO2); 137.62, 135.16 (2C, nb5,6); 133.74, 133.63 (2C, ph5); 131.98, 131.67 (2C, ph1); 129.26, 129.15 ( 2C, ph4); 128.87 (2C, ph6); 125.87; 125.02 (2C, ph3); 63.31, 63.54 (2C, O–CH2-ph); 48.67 (1C, nb1); 47.60 (1C, nb4); 47.57 (1C, nb2); 47.20 (1C, nb3); 45.78 (1C, nb7) ppm. IR-data (CaF2, cm−1): 1733, 1613, 1578, 1525, 1447, 1342, 1306, 1163.
Polymerization of mono- 1 to give poly(endo,exo-di(2-nitrobenzyl)bicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate) (poly- 1 ): a solution of mono-1 (500 mg, 1.105 mmol) in 3 mL of dichloromethane was added to a solution of (H2IMes)(py)2(Cl)2RuCHPh (6.7 mg, 9.2 μmol) in 2 mL of CH2Cl2 and the mixture was stirred at ambient temperature for 3 h. After complete conversion (TLC), the reaction was stopped by adding 0.15 mL of ethyl vinyl ether. The polymer was precipitated by dropping the solution into a tenfold excess of cold methanol. The precipitate was dried in vacuo. Yield: 420 mg of a white solid (84%). 1H-NMR: (δ, 400 MHz, 20 °C, CDCl3): 8.01–7.23 (8H, ph); 5.6–5.1 (2H, (CHCH) + 4H, (O–CH2-ph)); 3.74–3.03 (4H, nb1,2,3,5); 1.91–1.55 (2H, nb4) ppm. FTIR (film on CaF2, cm−1): 2963–2850, 1736, 1611, 1576, 1518, 1448, 1336, 1164, 1002. SEC: (CHCl3): Mn = 45600 g mol−1; Mw = 47900 g mol−1; PDI = 1.05. Tg = 57.9 °C.
Copolymerization of mono- 1 and mono- 2 to give poly[(endo,exo-di(2-nitrobenzyl)bicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate)-co-(endo,exo-diphenyl bicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate)] (poly( 1 -co- 2 ): a solution of H2IMes)(py)2(Cl)2RuCHPh (2.7 mg, 3.7 μmol) in 2 mL of CH2Cl2 was added to a mixture of mono-1 (50.2 mg, 0.11 mmol) and mono-2 (110.5 mg, 0.33 mmol) in 3 mL of dichloromethane and the mixture was stirred at RT for 24 h. After complete conversion (TLC), the reaction was stopped by adding 0.15 mL of ethyl vinyl ether. The polymer was precipitated by dropping the solution into a tenfold excess of cold methanol. The precipitate was dried in vacuo. Yield: 0.132 g (82%) of a white solid. SEC (CHCl3): Mn = 37000 g mol−1; Mw = 39900 g mol−1; PDI = 1.08; Tg = 79.9 °C. (δ, 400 MHz, 20 °C, CDCl3): 8.01–7.37 (18H, ph); 5.6–5.1 (4H, CHCH) (4H, O–CH2-ph); 3.74–3.03 (8H, nb1,2,3,5); 1.91–1.55 (4H, nb4) ppm. Composition: 28 mol% mono-1 and 72 mol% mono-2 as estimated from elemental analysis measurements; FTIR (film on CaF2, cm−1): 3125–2800, 1747, 1591, 1528, 1491, 1456, 1380, 1340, 1198, 1160, 1142.
For UV-VIS and FTIR measurements a 10 mg mL−1 solution of the corresponding polymer dissolved in tetrahydrofuran (THF) was spincast on CaF2 discs. The film thickness was in the range of 20–25 nm for ellipsometry. For UV-VIS and FTIR-spectroscopy measurements 100 nm thick polymer films have been applied.
UV-VIS spectra were recorded with a Varian Cary 50 UV-VIS spectrophotometer. All UV-VIS spectra were taken in the absorbance mode. Differential scanning calorimetry (DSC) measurements were carried out on a Perkin Elmer “Pyris Diamond” under a nitrogen flow of 20 mL min−1 and a heating rate of 10 °C min−1. Glass transition temperatures (Tg) from the second heating run were read as the midpoint of change in heat capacity.
For ellipsometric measurements a 10 mg mL−1 solution of the corresponding polymer dissolved in anisole was spincast on silicon wafers. Ellipsometric measurements were carried out with a Woollam VASE spectroscopic ellipsometer (Xenon short arc lam, wavelength range 240–1100 nm, and spectral bandwidth 4 nm). The implemented software uses the Levenberg–Marquardt fit algorithm. FTIR spectra were recorded with a Perkin Elmer “Spectrum One” instrument (spectral range between 4000 and 450 cm−1). All FTIR spectra were recorded in transmission mode.
Size exclusion chromatography: the weight and number average molecular weights (Mw and Mn) as well as the polydispersity index (PDI) were determined by gel permeation chromatography with tetrahydrofuran (THF) as the solvent using the following arrangement: Merck Hitachi L6000 pump, separation columns of Polymer Standards Service, 8 × 300 mm STV 5 μm grade size (106 Å, 104 Å, and 103 Å), combined refractive index–viscosity detector from Viscotec, Viscotec 200. Polystyrene standards purchased from Polymer Standards Service (PSS) were used for calibration.
Scheme 2 Synthesis of poly-1. |
The polymers were prepared by ring opening metathesis polymerization (ROMP) using (H2IMes)(py)2(Cl)2RuCHPh (H2IMes = N,N-di(mesityl)-4,5-di-hydroimidazolin-2-ylidene, py = pyridine) as an initiator for ROMP. The polymerization reactions were performed in dichloromethane under an inert atmosphere at room temperature in a glove box. A homopolymer, poly-1, with a theoretical polymerization degree Xn = 120 as well as a statistical copolymer poly(1-co-2) (Scheme 3, Xn = 120, mono-1:mono-2 = 30:90) were prepared by this method.
Scheme 3 Structure of poly(1-co-2). |
Both polymers were obtained in high yields of 84% and 82%, respectively. Due to overlapping NMR signals, elemental analysis was performed to determine the content of mono-1 and mono-2 in poly(1-co-2). The found value for the molar ratio of the incorporated monomers mono-1 and mono-2 amounts to be 1:2.5. The physico-chemical data of the investigated polymers are compiled in Table 1.
For poly(1-co-2) only one glass transition is observed (Tg = 79.9 °C), which is between the Tg of poly-1 (57.9 °C) and poly(endo,exo-diphenyl bicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate) (96.5 °C, ref. 3c) which is expected for a statistical copolymer. Both polymers show excellent film forming properties when spin-cast from CH2Cl2 and anisole solutions and fully transparent films with optical quality were obtained.
Fig. 1 shows the UV absorption spectra of the new polymers poly-1 and poly(1-co-2) compared to the absorption spectrum of poly(endo,exo-diphenyl bicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate). While the phenyl ester groups in poly(endo,exo-diphenyl bicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate) absorb light up to a wavelength λ ∼ 280 nm which is typical of a π–π* transition, the UV absorption spectra of the new polymers poly-1 and poly(1-co-2) exhibit a maximum at around 265 nm which extends the absorption to >300 nm as expected for ortho-nitrobenzyl ester units.
Fig. 1 UV spectra of films of poly-1 (solid line), poly(1-co-2) (dotted line) and poly(endo,exo-diphenyl bicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate) (dashed line) on CaF2. Film thicknesses are in the range of 100 nm. |
Scheme 4 Photo-reaction in poly-1 containing o-nitrobenzyl ester units and post-reaction of o-nitrosobenzaldehyde. |
In Fig. 2A the UV-VIS spectra of a film of poly-1 prior to and after UV illumination (energy density E = 19.8 J cm−2, λ > 300 nm) under a nitrogen atmosphere are shown. The film thickness was in the range of 100 nm. Upon illumination a significant increase in the UV absorption can be observed. This behavior can be attributed to the formation of 2-nitrosobenzaldehyde and the subsequent formation of the corresponding azobenzene moieties.16 Additionally, these photo-generated azobenzene groups lead to a discoloration of the polymeric film from transparent to a yellow shade in the illuminated areas. Regarding optical applications the increase of the UV-VIS absorption between 210 and 550 nm has to be taken into account. However, higher wavelengths typically used for optical data communication (λ > 800 nm) are not affected (see Fig. 2A).
Fig. 2 (A) UV spectra of a film of poly-1 on CaF2. Solid line: prior to irradiation; dotted line: after UV irradiation (E = 19.8 J cm−2, λ > 300 nm). (B) FTIR spectra of poly-1 on CaF2 prior to illumination (solid line) and after illumination (dotted line) with UV-light of >300 nm (energy density E = 20 J cm−2). |
Fig. 2B shows detailed FTIR spectra of a film of poly-1 before (solid line) and after illumination (dotted line) with UV-light (E = 19.8 J cm−2, λ > 300 nm). The film thickness was in the range of 100 nm. In the spectrum of the non-irradiated film the signals at 1728 cm−1 (CO stretch) and at 1165 cm−1 (asym. C–O–C stretch) are typical for the ester units. Additionally, strong signals at 1528 cm−1 (asym. NO2 stretch) and 1342 cm−1 (sym. NO2 stretch) are observed, which can be attributed to the nitro group.
After UV irradiation, the vibrational bands of the ester (at 1728 cm−1 and 1165 cm−1) as well as the nitro group (at 1528 cm−1 and 1342 cm−1) decreased considerably. Instead, a new broad band at 3400 cm−1 that can be assigned to the O–H stretching vibration of the carboxylic acid group is observed. Additionally, a broad shoulder at approximately 1705 cm−1 is clearly discernible. This signal indicates the formation of the expected carboxylic acid group (CO stretching vibration), partially overlapping with the carbonyl vibration of the remaining ester.
A quantification of the yield was not possible from FTIR data as it is difficult to distinguish between the formed carboxylic acid groups, the remaining ester groups and the side products (i.e. the azobenzene derivative) of the photoreaction.
Due to the overlap of the C–O signal of the photo-product with the ester band, the NO2 stretching vibration at 1528 cm−1 was used for kinetic investigations in poly-1 as well as poly(1-co-2).
Fig. 3 shows the decrease of the NO2 groups during UV illumination (λ > 300 nm) of poly-1. A quantitative conversion of the nitro-groups can be observed upon irradiation with E = 20 J cm−2.
Fig. 3 Monitoring the photo-reaction in poly-1 by FTIR spectroscopy: depletion of the NO2 band at 1528 cm−1 during UV illumination (power density P = 13.2 mW cm−2). |
The refractive index change in a thin film of poly-1 was investigated by means of spectroscopic ellipsometry. Fig. 4A displays the Cauchy fits of the dispersion of the refractive index for the pristine and the UV illuminated films of the polymers. In these experiments, the duration of UV irradiation was adjusted to obtain a maximum conversion of the ester groups.
Fig. 4 (A) Cauchy fits of the dispersion of the refractive index of poly-1 before (solid line) and after (dotted line) illumination with UV-light >300 nm (E = 15.8 J cm−2). (B) Phase contrast image of a film of poly-1 after UV patterning with a mask aligner, (MJB4 from SUSS) using a 500 W HgXe lamp equipped with a filter for the range 270–353 nm. |
The refractive index of poly-1 decreased by Δn = −0.047 at 589 nm, which is remarkably high and more than sufficient for optical applications. This expected result can be explained by the formation of carboxylic acid groups. In general, carboxylic acids possess comparably low refractive indices, which explains the observed decrease after the photo-induced acid formation.5
Further support for this assumption is provided by a comparison of known refractive indices of low-molecular-weight model compounds. o-Nitrobenzyl acetate – the model compound for the pristine polymer – has a refractive index of Δn589 = 1.545, whilst acetic acid, the analogue for the formed carboxylic acid group, has a very low refractive index of Δn589 = 1.371.
Although the side product of the photoreaction, see Scheme 4, also influences the refractive index, it can be assumed that the additional carboxylic acid groups of this azobenzene derivative also contribute to the refractive index decrease.
The observed decrease in the refractive index can also be followed by optical microscopy using a phase contrast set-up for visualization. In Fig. 4B, a phase contrast image of a patterned sample is shown. Areas with higher refractive indices appear dark (non-illuminated), while areas with lower refractive indices appear bright (illuminated). To create the pattern, a thin film of poly-1 was illuminated by using a mask aligner system and a suitable quartz-chromium mask (contact lithography). Without any optimization, a resolution of 5 μm was achieved in this experiment.
Fig. 5 Two step irradiation procedure: (i) irradiation using UV-light of λ = >300 nm (E = 20 J cm−2 and λ > 300 nm) and (ii) irradiation using UV-light of λ = 254 nm (E = 0.85 J cm−2 and λ = 254 nm). |
Fig. 6 shows a comparison of the FTIR spectra prior to (A) and after illumination with UV-light >300 nm (B) and after a subsequent illumination with 254 nm (C), respectively. The first illumination step (>300 nm) causes the formation of carboxylic acid groups as indicated by the formation of the shoulder at 1705 cm−1 (CO stretch) and also a decrease of the ester vibration (1750 cm−1). This FTIR spectrum also reveals that no photo-product of the photo-Fries rearrangement (i.e. hydroxy ketone, which would appear at 1630 cm−1) is formed. This fact is expected as only the ortho-nitrobenzyl ester absorbs light above 300 nm. Also a significant decrease of the corresponding NO2 vibrations at 1528 cm−1 (asym. stretching vibration) and 1342 cm−1(sym. stretching vibration) is observed. After irradiation of poly(1-co-2) with UV-light of >300 nm a subsequent illumination with 254 nm light was performed. Hence, the formation of hydroxy ketone moieties (1631 cm−1, CO stretch) and also a further decrease of the ester signal at 1750 cm−1, attributed to the photo-induced photo-Fries rearrangement of the phenyl ester groups, can be detected.
Fig. 6 Comparison of FTIR spectra of poly(1-co-2) obtained by different illumination steps: solid line: prior to irradiation; dotted line: after UV irradiation (E = 20 J cm−2 and λ > 300 nm); dashed line: after 2nd illumination step (E = 0.85 J cm−2 and λ = 254 nm). |
In Fig. 7 the depletion of the nitro peak during illumination with UV-light of >300 nm (first illumination step) and the evolution of a ketone band during the subsequent second illumination step with UV-light of 254 nm are depicted as a function of the irradiation time. This graph also reveals the wavelength selective photo-reactivity of poly(1-co-2), i.e. no photo-product of the photo-Fries rearrangement is formed in the first illumination procedure.
Fig. 7 Conversion of nitro groups (squares, P = 13.2 mW cm−2) and formation of o-hydroxy ketone (circles, P = 178 μW cm−2) during the two step irradiation. (1st step λ > 300 nm and 2nd step λ = 254 nm). |
Although, the exact amount of photo-generated carboxylic acid groups cannot be determined due to overlapping infrared signals, an estimation of the yield of hydroxy ketones can be performed using the infrared absorbance coefficients of low molecular model compounds as described in previous contributions.3c,5
For that purpose, hydroxy acetophenone (A1631 = 345 L mol−1 cm−1) and phenyl acetate (A1750 = 420 L mol−1 cm−1) were used as model compounds. Using this method a maximum yield of ortho-hydroxy ketone was estimated to be 30% after the second illumination step (see Fig. 7).
Fig. 8 Cauchy fits of the dispersion of the refractive index of poly(1-co-2) before (solid line) and after the 1st illumination step (dotted line) with UV-light of >300 nm (E = 18.2 J cm−2); after 2nd illumination step (dashed line) using UV-light of 254 nm (E = 4.1 J cm−2). |
According to the investigations on poly-1, the irradiation of poly(1-co-2) with UV >300 nm leads to a significant decrease of the refractive index, which can be attributed to the formation of carboxylic acid groups. The value of the decrease (Δn589 = −0.017) is approximately 1/3 of the change observed in the homopolymer poly-1 (Δn589 = −0.047) and is in good accordance with the molar fraction of o-nitrobenzyl groups in poly(1-co-2). This fact reveals the possibility of a selective tailoring of the expected refractive index change in ROMP derived polymers by the choice of the molar ratio of the applied monomers.
The second irradiation of this film with UV-light of 254 nm wavelength results in an increase of the refractive index by approximately Δn589 = +0.031, which can be assigned to the photo-induced generation of o-hydroxy ketones. The finding is consistent with the observed refractive index changes in previously reported polymeric materials (considering that poly(1-co-2) contains approximately 66 mol% of photo-Fries reactive groups).3c,5
This journal is © The Royal Society of Chemistry 2013 |