Documente Academic
Documente Profesional
Documente Cultură
A R T I C L E I N F O A B S T R A C T
∗
Corresponding author.
E-mail address: guillaume.raynel@aramco.com (G. Raynel).
http://dx.doi.org/10.1016/j.polymdegradstab.2017.09.021
Received 19 July 2017; Received in revised form 24 September 2017; Accepted 26 September 2017
Available online 28 September 2017
0141-3910/ © 2017 Elsevier Ltd. All rights reserved.
G. Raynel et al. Polymer Degradation and Stability 146 (2017) 69–77
Although POD is considered to be very chemically stable, the hy- either CDCl3 or DMSO-d6. For DMSO-d6 samples, heating at elevated
drolysis of POD is known to occur, leading to the opening of the 1,3,4- temperatures was employed at times to homogenize the sample. All
oxadizaole ring in the presence of acids [3,29]. In this work, a similar spectra were acquired on a Varian 500 MHz NMR spectrometer,
mechanism is detected in F-POD. The electron withdrawing effect of the equipped with a 5 mm liquids probe. The 1H NMR spectra were ac-
hexafluoroisopropylidene group weakens the carbon bonds in the ox- quired with a single 90° pulse, a 60 s recycle delay and a 5 s acquisition
adiazole ring, leading to a further degradation of the molecule through time. Generally, a line broadening of 0.5 Hz was applied to the FID
hydrolysis in acidic media. The vulnerability of F-POD is demonstrated before Fourier transformation. The spectra were typically calibrated to
in different acidic media, along with how this degradation property can the TMS signal (0 ppm).
be explored as a method to crosslink the polymer.
2.3.2. FTIR spectroscopy
2. Experimental Samples for FTIR were analysed as received in solid form. For the
sample dissolved in CDCl3, the solvent was evaporated prior to the test.
2.1. Materials FTIR spectroscopic data were collected for all samples using a Varian
FT-IR spectrophotometer equipped with mercury cadmium telluride
The synthesis of F-POD is reported elsewhere [19,28]. Acetic acid (MCT) detector and Golden gate accessory with an average of 128 scans
(Sigma Aldrich, USA, > 99.8%), hydrochloric acid (Fisher scientific, at a resolution of 4 cm−1.
USA, 37%), deuterated chloroform (CDCl3) (Merck Sharp and Dohme,
Montreal, > 99.8% D), deuterated dimethylsulfoxide (DMSO-d6) 2.3.3. Gel permeation chromatography
(Wilmad Glass Co., USA, > 99.8% D), tetrahydrofuran (THF) (Sigma Samples for GPC were prepared by dissolving the solid material in
Aldrich, USA, > 99.9%), hexylamine (Sigma Aldrich, USA, 99%), 1,8- THF (2 mg/mL). The samples were left for 30 min to ensure complete
diaminooctane (Sigma Aldrich, USA, 98%) and iron sulfide (Sigma dissolution and then filtered through a 0.45 μm syringe filter prior to
Aldrich, USA, > 95%) were used without further purification. Ultra- injection of 100 μL. The Agilent 1260 Infinity GPC/SEC system was
pure water was purified on a Milli-Q® Integral 15 water system. equipped with a degasser, an isocratic pump, an autosampler, a ther-
mostatted column compartment and a refractive index detector. The
2.2. F-POD degradation pump was used to achieve a THF eluent flow rate of 1 mL/min through
two GPC columns (type 100 Å and 10,000 Å), each 30 cm long and
2.2.1. Aqueous acidic F-POD degradation (pH 1.5 & 5.5) connected in series. The columns were packed with highly cross-linked
A small sample of F-POD (25 mg) was added to 20 mL of an aqueous polystyrene-divinylbenzene (PhenogelTM) 5 μm particles. The tempera-
solution of hydrochloric acid (31.7 mmol/L) and acetic acid ture of the column oven was maintained at 30 °C. Calibration was
(872.5 mmol/L). The pH of the solution was 1.5. Another F-POD sample achieved using Agilent EasiCal polystyrene standards (molecular
was added to 20 mL of an aqueous solution of hydrochloric acid weights: 580; 1390; 2780; 4830; 9970; 19,540; 51,150; 113,300;
(31.7 mmol/L) and iron sulphide (187.1 mg, 2.13 mmol). The pH was 224,900 and 483,400 g/mol) dissolved in THF. The average molecular
5.5. Both mixtures were left overnight at 25 °C. The treated F-POD weight of the samples was calculated from the plot of molecular weight
samples were then rinsed with pure water and left to dry overnight. versus relative retention time, using the integrated Agilent software.
Both samples were analysed by differential scanning calorimeter (DSC),
thermogravimetric analyser (TGA) and attenuated total reflection 2.3.4. Differential scanning calorimeter
Fourier-transform infrared spectroscopy (ATR FTIR) spectroscopy. The Glass transition temperatures were determined by a TA instruments
average molecular weights of the samples were subsequently de- DSC Q20 coupled with refrigerated cooling system, following ASTM
termined using gel permeation chromatography (GPC). D3418 standard method. Scans were run with 5 mg of sample under a
nitrogen purge (20 cm3/min) as follow a heating cycle at a heating rate
2.2.2. Non-aqueous acidic F-POD degradation of 10 °C/min to 400 °C, a cooling cycle at a rate of 100 °C/min to
An F-POD sample (25 mg) was added to 1 mL of CDCl3. This liquid 20.0 °C, and finally a heating cycle at a heating rate of 10 °C/min to
mixture was analysed by NMR spectroscopy immediately after dis- 400 °C. Reported values were obtained from a second heating run.
solution and by FTIR spectroscopy after being left overnight at 25 °C.
The remaining solution was left to evaporate at 25 °C. The resulting 2.3.5. Thermogravimetric analysis
beige solid was not soluble in CDCl3. The solid was partially dissolved The samples were analysed using a TA instruments TGA Q500 that
at 60 °C in DMSO-d6 and was analysed again by NMR spectroscopy. was programmed to run at a heating rate of 10 °C/min between 20 °C
and 800 °C under a nitrogen atmosphere to determine the weight losses
2.2.3. Reaction of F-POD with hexylamine of the samples. About 5 mg of each sample was uploaded into TGA
An F-POD sample (23.1 mg) was added to 1 mL of a deuterated sample pan.
chloroform solution with 29.2 mg (286 μmol) of hexylamine. The
mixture was left overnight at 25 °C. An off-white solid precipitated 3. Results & discussions
immediately. This precipitate was analysed by FTIR and NMR spec-
troscopy in DMSO-d6. The water-insoluble poly(2,5-(4,4’-(hexafluoropropylidene)di-
phenyl)-1,3,4- oxadiazole) (F-POD in Fig. 1) was exposed to acidic
2.2.4. Crosslinking reaction of F-POD aqueous solutions (pH 1.5 and 5.5). In parallel, a similar amount of F-
An F-POD sample (26.2 mg) was added to 1 mL of a deuterated POD was added to pure water as a reference sample.
chloroform solution with 1.8 mg (12.2 μmol) of 1,8-diaminooctane. The
mixture was left overnight at 25 °C. An off-white solid precipitated
immediately. This precipitate was analysed by FTIR and NMR spec-
troscopy in DMSO-d6.
70
G. Raynel et al. Polymer Degradation and Stability 146 (2017) 69–77
Fig. 4. GPC chromatograms for treated F-POD (red) and untreated F-POD (blue). (For
interpretation of the references to colour in this figure legend, the reader is referred to the
Fig. 2. Reference sample of F-POD (left) and F-POD treated in acidic conditions (right) in web version of this article.)
pure water.
Fig. 3. DSC thermograms for untreated F-POD (blue) and treated F-POD (red). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version
of this article.)
71
G. Raynel et al. Polymer Degradation and Stability 146 (2017) 69–77
Fig. 5. TGA thermograms for untreated F-POD (blue) and treated F-POD (red). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version
of this article.)
intensity (19% ± 1% in pH 5.5 and 23% ± 1% in pH 1.5) for F-POD hydrolysis to give the corresponding perfluoroalkanoic acids and hy-
samples exposed to aqueous acidic solution. This effect was more sig- drazinium salt in acidic conditions; or the resultant perfluoroalkanoate
nificant as the pH of the aqueous solution decreased. Furthermore, the salt and hydrazine in basic conditions [31,32].
infrared spectra peak at 1725 cm−1 marginally increased. This nascent The build-up of the carboxylic acid functionality at the F-POD's
peak was attributed to the C=O stretching of carboxylic acid. Brown surface indicates a similar mechanistic path of degradation (Scheme 1)
et al. showed that the hydrolysis of 2,5-bisperfluoroalkyl-1,3,4-ox- as observed in the hydrolysis of 2,5-bisperfluoroalkyl-1,3,4-oxadiazoles
adiazoles occurs readily in basic or acidic conditions to yield bis(per- in acidic conditions. In this scenario, the carbons, which constitute the
fluoroalkanoic acid) hydrazides, which can undergo a subsequent 1,3,4-oxadiazole ring, become more electrophilic due to the electron
Fig. 6. Infrared spectrum superposition of untreated F-POD (blue), F-POD treated with acid (red pH 5.5 & purple pH 1.5). (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)
72
G. Raynel et al. Polymer Degradation and Stability 146 (2017) 69–77
Fig. 7. 1H NMR spectra of F-POD in CDCl3 (a), F-POD degraded in DMSO-d6 (b), hexylamine reaction product in DMSO-d6 (c), and cross-linked F-POD in DMSO-d6 (d).
withdrawing effect of bis(trifluoromethyl)methylene groups in the para the various aqueous solutions used to clean the treated F-POD. With
position. The nucleophilic attack of water on F-POD (1) gives, as an time, the hydrazinium salts (pKa: 8.1 [32–34]) in equilibrium with its
intermediate, the polyhydrazide (2). Readily, a carbonyl group of 2 base form, will slowly oxidize in the presence of oxygen to produce
undergoes another nucleophilic addition of water to form the corre- nitrogen and water [35]. The presence of hydrazine in F-POD, as a
sponding acid (3) and hydrazide (4). Finally, the hydrolysis of 4 yields decomposition product was discussed by Gomes et al. [28] during the
3 and a hydrazinium ion (5). sulfonation of F-POD to explain the increase of the N/C ratio from 0.119
The hydrazinium salts formed can be assumed to be washed away in to 0.138. In comparison, it was reported that poly(1,4-phenylene-1,3,4-
73
G. Raynel et al. Polymer Degradation and Stability 146 (2017) 69–77
Fig. 8. FTIR spectra of reference F-POD, F-POD after NMR in CDCl3, reaction with hexylamine and cross-linked F-POD.
oxadiazole) (6) hydrolyzed to the matching polyhydrazide (7) in acidic non-polar solvent, such as chloroform. The spectrum showed the main
conditions (25–75% sulfuric acid) at room temperature (Scheme 2) [3]. product (80% of F-POD 1, 1H NMR: 8.19 & 7.61 in CDCl3) some traces
The degradation of 6 into diacid (8) occurs in acidic conditions at of unreacted diacid (11% of 3, 1H NMR: 8.05 & 7.49 ppm in CDCl3) and
higher temperature (i.e., reflux of aqueous sulfuric acid solution) [29]. polyhydrazide (9% of 2, 1H NMR: 8.11 & 7.54 ppm in CDCl3) [28,36].
A 1H NMR of the reference F-POD material was performed in CDCl3 An FTIR measurement of the resultant deuterated solution was per-
(top spectrum of Fig. 7). The F-POD polymer dissolved fully in aprotic formed and showed an unexpected partial decomposition (Fig. 8). The
74
G. Raynel et al. Polymer Degradation and Stability 146 (2017) 69–77
13
Fig. 9. C{1H} and 1H NMR spectra of 11 in DMSO-d6.
peak of the oxadiazole C=N stretching at 1665 cm−1 disappeared, the limiting factor of this acidic hydrolysis was water (12.6%, 3.6 mmol
while four peaks were observed to increase: the peak previously en- of water for 28.6 mmol of 1), only a partial decomposition of F-POD
countered in the degradation of F-POD in acidic water (i.e. 1725 cm−1), was observed.
attributed to the C=O stretching of carboxylic acid (3); two peaks at The terminal amino group of the carboxylic hydrazide (4) is a good
1645 cm−1 and at 1619 cm−1, attributed to C=O stretching of car- nucleophile [42], which may attack the 2 (or 5) position(s) on the
boxylic hydrazide functionality existing in 2 and 4; and a peak at 1,3,4-oxadiazole ring, as shown on Scheme 3, to form a 1-((4,4’-(hex-
3071 cm−1, attributed to N-H stretching of hydrazine species (5) [37]. afluoropropylidene)diaryl)amidinyl)-2-((4,4’-(hexafluoropropylidene)
It is posited that the F-POD polymer in CDCl3 reacted with the trace diaryl)carboxyl) hydrazine (9) [32,43]. The change of solubility of the
residual water in the solvent (0.066 %w of water in chloroform [38] at resulting solid (i.e., no longer soluble in CDCl3) indicates that the linear
20 °C) [39,40], contributing to the opening of 1,3,4-oxadiazole [41]. As F-POD 1 was transformed through partial acid hydrolysis into a
75
G. Raynel et al. Polymer Degradation and Stability 146 (2017) 69–77
reticulated polymer (e.g. branched or cross-linked polymer). The F-POD acidic conditions (pH 5.5). This effect should be considered for appli-
decomposition in wet acidic chlorinated solvent resulted in a beige cations where F-POD is used as a material in contact with acidic water
coloured solid, which was slightly soluble in DMSO-d6. The 1H NMR (e.g., membranes for waste water treatment or fuel cell membranes in
spectrum (Fig. 7(b)) showed the diacid 3 remaining from the reference humid environments). In addition, we also demonstrated that when F-
F-POD material as the main product (56%, 1H NMR: 8.05 & 7.48 ppm in POD is dissolved in aprotic non-polar solvent (i.e., chloroform), the
DMSO-d6), with minor products of unreacted F-POD (24% of 1, 1H residual water present in the solvent is capable of inducing degradation.
NMR: 8.17 & 7.60 ppm in DMSO-d6) and carboxylic hydrazide (20% of The initially linear polymer is transformed into a reticulated polymer
4, 1H NMR: 8.27 & 7.66 ppm in DMSO-d6). (e.g., branched or cross-linked polymer), considerably losing its solu-
The formation of 1,2,4-triazole from 2,5-disubstituted-1,3,4-ox- bility, except for partial solubility in a highly polar solvent such as
adiazole [43–45], or 1,2-disubstituted carboxylic hydrazide [46], DMSO. Furthermore, the product of degradation yielded the substituted
usually requires heating to circumvent the sterics between the aromatic amidinyl carboxyl hydrazide, or the substituted bis-amidinyl hydrazide
protons in ortho position of the triazole ring and the R group attached in the presence of an excess of amines under mild reaction conditions.
to the nitrogen, which would be positioned at the 4 position on the 2,5- Lastly, the crosslinking of F-POD using a diamine compound (1,8-dia-
disubstituted-1,2,4-triazole ring. These steric interactions don't allow minooctane) was successfully induced by taking advantage of the me-
for maximum orbital overlap of the phenyl-triazole-phenyl aromatic chanistic pathway of degradation shown in this work.
rings, consequently destabilizing slightly the molecule (i.e., increasing
its energy of formation) [47]. The reaction of F-POD (1) with an excess Acknowledgments
of hexylamine (10) shown in Scheme 4 was performed to investigate
the potential ring formation, which would yield the corresponding The authors would like to thank Mr. Amer A. Tuwailib, Dr. Saroj K.
1,2,4-triazole (12) at room temperature. The resulting off-white powder Panda and Mr. Thunayyan A. Al-Qunaysi from the Saudi Aramco
was analysed by FTIR and NMR spectroscopies (IR spectrum on Fig. 8 Research and Development Center for their support with sample ana-
and 1H NMR spectrum on Fig. 7). lysis; Mr. Phil Embleton from the Saudi Aramco Public Relations
An FTIR was performed on the sample and is shown at the middle of Department for proofreading and Dr. Regis Vilagines from the Saudi
Fig. 8. The peak of the oxadiazole C=N stretching at 1665 cm−1 has Aramco Research and Development Center for endorsing this research.
disappeared. In addition, two peaks increased: one peak at 1735 cm−1,
attributed to C=N stretching of amidinyl group of 11; one peak at References
1377 cm−1, attributed to C-N stretching of amidine species (11). The
1
H NMR spectrum of the reaction product (Fig. 7(c) and Fig. 9) showed [1] C.S. Abshire, J. Claude, Marvel, Some oxadiazole and triazole polymers, Die
a shielding of the aromatic peaks in 11 (1H NMR: 7.90 & 7.27 ppm in Makromol. Chem. 44 (1961) 388–397, http://dx.doi.org/10.1002/macp.1961.
020440132.
DMSO-d6) compared to F-POD (1H NMR: 8.17 & 7.60 ppm in DMSO-d6). [2] A.H. Frazer, W. Sweeny, F.T. Wallenberger, Poly(1,3,4-oxadiazoles): a new class of
Furthermore, the 13C NMR spectrum of this product (Fig. 9) showed a polymers by cyclodehydration of polyhydrazides, J. Polym. Sci. Part A Gen. Pap. 2
peak at ∼168 ppm in DMSO-d6, attributed to C=N of the two amidinyl (1964) 1157–1169, http://dx.doi.org/10.1002/pol.1964.100020313.
[3] B. Schulz, New aspects of the solid-state structures and electrochemical properties
group of 11, compared to the C=N peak of F-POD at 153 ppm in CDCl3. of aromatic poly(1,3,4-oxadiazole)s, Mol. Cryst. Liq. Cryst. Sci. Technol. Sect. A.
When the NMR sample was heated at 50 °C, no change in the NMR Mol. Cryst. Liq. Cryst. 240 (1994) 135–141, http://dx.doi.org/10.1080/
spectrum was observed. The cyclisation of the hydrazide intermediate 10587259408029724.
[4] B. Schulz, M. Bruma, L. Brehmer, Aromatic Poly(1, 3, 4-Oxadiazoe)s as advanced
(e.g. 9, 11, or 14) to form the corresponding 1,2,4-triazole didn't occur materials, Adv. Mater 9 (1997) 601–613, http://dx.doi.org/10.1002/adma.
at temperatures lower than 50 °C, which corroborate with the structures 19970090804.
of the products in the various proposed mechanisms. [5] R.L. Leonard, J.D. Veitch, T.M. Veazey, Process for the Production of Flame
Retarded Acrylic Fibers, US 3193602 A (1962).
Using a similar mechanistic approach to the nucleophilic attack of
[6] A.H. Frazer, F.T. Wallenberger, Poly(1,3,4-oxadiazole) fibers: new fibers with su-
carboxylic hydrazide onto F-POD (Schemes 3 and 4), the cross-linking perior high temperature resistance, J. Polym. Sci. Part A Gen. Pap. 2 (1964)
reaction of F-POD shown in Scheme 5 was performed with of 1,8-dia- 1171–1179, http://dx.doi.org/10.1002/pol.1964.100020314.
minooctane (13) and F-POD (1). The resulting powder (14) was off- [7] B. Schulz, Y. Kaminorz, L. Brehmer, New aromatic Poly(1,3,4-oxadiazole)s for light
emitting diodes, Synth. Met. 84 (1997) 449–450, http://dx.doi.org/10.1016/
white in colour and was analysed by FTIR and NMR spectroscopies (IR S0379-6779(97)80826-0.
spectrum on Fig. 8 and 1H NMR spectrum of the soluble fraction on [8] Q. Pei, Y. Yang, 1,3,4-Oxadiazole-Containing polymers as electron-injection and
Fig. 7(d)). blue electroluminescent materials in polymer light-emitting diodes, Chem. Mater 7
(1995) 1568–1575, http://dx.doi.org/10.1021/cm00056a025.
An FTIR was performed on the sample and is shown at the bottom of [9] S. Janietz, S. Anlauf, A. Wedel, New n-type rigid rod full aromatic poly(1,3,4-ox-
Fig. 8. The peak of the oxadiazole C=N stretching at 1665 cm−1 was adiazole)s and their application in organic devices, Synth. Met. 122 (2001) 11–14,
found to have disappeared. In addition, four peaks increased related to http://dx.doi.org/10.1016/S0379-6779(00)01320-5.
[10] Y. Kaminorz, B. Schulz, S. Schrader, L. Brehmer, OLEDs based on new oxadiazole
14: one peak at 1594 cm−1, attributed to C=N stretching of the ami- derivatives, Synth. Met. 122 (2001) 115–118, http://dx.doi.org/10.1016/S0379-
dinyl group; two peaks at 1545 cm−1 and at 1502 cm−1, attributed to 6779(00)01367-9.
C=O stretching of carboxylic hydrazide functionality; and a peak at [11] E.R. Hensema, B. Gebben, M.H.V. Mulder, C.A. Smolders, Polyoxadiazoles and
polytriazoles as new heat and solvent resistant membrane materials, Bull. Des
1379 cm−1, attributed to C-N stretching of amidine species. The 1H Sociétés Chim. Belges, Wiley-VCH Verlag GmbH & Co. KGaA, 1991, pp. 129–136, ,
NMR spectrum of the soluble fraction of the crosslinked product (15, http://dx.doi.org/10.1002/bscb.19911000205.
Fig. 7(d)) showed a shielding of the aromatic peaks (1H NMR: [12] E.R. Hensema, M.H.V. Mulder, C.A. Smolders, On the mechanism of gas transport in
rigid polymer membranes, J. Appl. Polym. Sci. 49 (1993) 2081–2090, http://dx.doi.
7.90 & 7.28 ppm in DMSO-d6) compared to 11 (1H NMR:
org/10.1002/app.1993.070491204.
7.90 & 7.27 ppm in DMSO-d6 – Fig. 7(c)) and F-POD (1H NMR: [13] E.R. Hensema, M.E.R. Sena, M.H.V. Mulder, C.A. Smolders, Gas separation prop-
8.17 & 7.60 ppm in DMSO-d6). Furthermore, the cross-linked product, erties of new polyoxadiazole and polytriazole membranes, Gas. Sep. Purif. 8 (1994)
which wasn't fully soluble in DMSO-d6, swelled with the addition of the 149–160, http://dx.doi.org/10.1016/0950-4214(94)80025-1.
[14] S.M.J. Zaidi, S.F. Chen, S.D. Mikhailenko, S. Kaliaguine, Proton conducting mem-
deuterated solvent (Fig. 8). branes based on polyoxadiazoles, J. New Mater. Electrochem. Syst. 3 (2000) 27–32.
[15] D. Gomes, J. Roeder, M.L. Ponce, S.P. Nunes, Characterization of partially sulfo-
4. Conclusions nated polyoxadiazoles and oxadiazole–triazole copolymers, J. Memb. Sci. 295
(2007) 121–129, http://dx.doi.org/10.1016/j.memsci.2007.02.046.
[16] D. Gomes, J. Roeder, M.L. Ponce, S.P. Nunes, Single-step synthesis of sulfonated
In this work, the degradation of F-POD in the presence of aqueous polyoxadiazoles and their use as proton conducting membranes, J. Power Sources
acidic media is demonstrated by a variety of analytical techniques. The 175 (2008) 49–59, http://dx.doi.org/10.1016/j.jpowsour.2007.09.090.
[17] K.P. Lee, T.C. Arnot, D. Mattia, A review of reverse osmosis membrane materials for
F-POD degradation increased as the pH of the aqueous acidic solution desalination—Development to date and future potential, J. Memb. Sci. 370 (2011)
decreased. This surface degradation can be as significant as 20% in mild
76
G. Raynel et al. Polymer Degradation and Stability 146 (2017) 69–77
77