Study of polyamide composite reverse osmosis membrane degradation in water under gamma rays

Journal of Membrane Science 480 (2015) 64–73

Contents lists available at ScienceDirect

Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci

Study of polyamide composite reverse osmosis membrane degradation in water under gamma rays Nicolas Combernoux a,b, Luc Schrive a, Véronique Labed a, Yvan Wyart b, Emilie Carretier b, Adeline Benony-Rhodier a, Philippe Moulin b,n a

Commissariat à l’Energie Atomique et aux Energies Alternatives, CEA, DEN, DTCD, SPDE, F 30207 Bagnols sur Cèze, France Aix Marseille Université, CNRS, Centrale Marseille, M2P2 UMR 7340, Equipe Procédés Membranaires (EPM), Europôle de l’Arbois, BP80, Pavillon Laennec, Hall C, 13545 Aix en Provence Cedex, France b

art ic l e i nf o

a b s t r a c t

Article history: Received 17 December 2014 Received in revised form 7 January 2015 Accepted 8 January 2015 Available online 19 January 2015

This study aims to investigate the impact of irradiation on the behavior of Polyamide (PA) composite reverse osmosis (RO) membranes. Irradiations were performed for two doses (0.1 and 1 MGy) in wet conditions under an oxygen atmosphere, with a gamma 60Co source. Irradiation effect on RO membranes performances (NaCl rejection, permeability) was studied before and after irradiation. ATR-FTIR, XPS, AFM, FESEM microscopy, ion chromatography were also used to characterize structural modifications. Results show that NaCl rejection of RO membranes irradiated at 1 MGy decreased until 64% and permeability increased by a factor of three. Nevertheless, membranes irradiated at 0.1 MGy did not exhibit any change in theirs permselectivity properties. Advanced analysis techniques confirmed that the firsts effects of gamma rays on RO membranes occurred between 0.1 and 1 MGy. Results emphasize that gamma rays effects on the RO membranes led to the breaking of amide and ester bonds at 1 MGy. These breakings resulted in loss of hydrogen bonds between polyamide chains, and consequently to a relaxation of the polyamide network. Finally, modifications of the polysulfone layer underneath were highlighted. Both relaxation of the polyamide network and modifications of the polysulfone layer could be involved in the drop of the permselectivity properties. & 2015 Elsevier B.V. All rights reserved.

Keywords: Gamma irradiation Reverse osmosis Polyamide membrane XPS

1. Introduction Nuclear power industry provides 3% of the world's energy, while in France it reaches roughly 80% [1]. As a result, nuclear facilities generate a significant amount of radioactive liquid effluents needing further treatment to meet environmental requirements. Evaporation, precipitation and ion exchange are most commonly used in industrial scale for the decontamination of these effluents [2]. However, membrane processes for decontamination have shown a significant growth in the past decades since they exhibit many advantages over conventional processes [3]. Among them reverse osmosis might be suitable for radioactive liquid waste treatment [4]. Reverse osmosis is a well-known process extensively used in desalination to produce drinking water [5,6], and in other applications [7,8]. Reverse osmosis membranes are Thin Film Composite (TFC) materials usually made of a superposition of three layers: an active one with aromatic polyamide supported by a polysulfone microporous layer and a non-woven polyester bottom structure.

n

Corresponding author. Tel.: þ 33 4 42 90 85 01; fax: þ 33 4 42 90 85 15. E-mail address: [email protected] (P. Moulin).

http://dx.doi.org/10.1016/j.memsci.2015.01.019 0376-7388/& 2015 Elsevier B.V. All rights reserved.

However, accurate information about the composition of commercial TFC RO membranes remains unknown through manufacturer data. Only a limited number of studies have carried out characterizations of some commercial RO membranes. Most commonly, studies focus mainly on the characterization of PA active layer, since the active layer governs separation properties [9–11]. As widely reported in the literature, organic materials are known to undergo degradations under gamma radiations [12]. Particularly for nuclear industry, gamma rays effects on polymer are studied on Ion Exchange Resin (IER) regarding long term behavior [13] and polymer materials for their stability [14,15]. Gamma rays are highly penetrating ionizing radiations that could induce a production of radical species into the polymers structure. Evolution of these species via radical or ionic mechanisms leads to crosslinking or relaxation reactions and gas production [16]. Only few studies focus on polyamide degradations under irradiation, and comparisons between studies are difficult because of the wide variety of polyamide type materials and irradiation conditions. Moreover, behavior of linear polyamide under irradiation is most commonly studied to estimate radiolytic yields, gas production, radical species and degradation mechanisms instead of modification of separation properties for membrane processes [17–20].

N. Combernoux et al. / Journal of Membrane Science 480 (2015) 64–73

65

Fig. 1. Experimental set-up for filtration experiments.

A limited number of studies reported the effect of gamma irradiation on TFC RO or NF membranes [21–24]. According to these papers, first effects of gamma rays degradation occur above the dose of 0.1 MGy,1 with variations of membrane permeability and selectivity. These results show some similitude with those from studies carried out on the ageing of RO membranes under chlorine exposition [25]. Only Nakase et al. [24] have reached higher doses and found out a link between structural and separation properties. However, their study was carried out with accelerated electron and not gamma rays which depending on conditions may result in different effects [26]. Finally, the interest in RO membranes for nuclear effluents treatment is related to the Fukushima-Daiichi accident. Indeed, one step of the radioactive water decontamination process was carried out with RO membranes. Consequently, the aim of the present work is to study the modifications of physical and chemical properties of both polyamide and polysulfone layers, and link them to separation performances for two doses (0.1 and 1 MGy) of gamma rays. The 0.1 MGy dose was defined in accordance with the literature on RO membranes. A 10 factor was then applied to set the upper dose of 1 MGy in order to effectively observe noticeable effects of gamma irradiation on RO membranes. These two thresholds were also in accordance with the feedback from IER degradation studies under gamma irradiation.

2. Materials and methods 2.1. Materials Reverse osmosis membranes SE (GE Osmonics) were used in this study. Membranes were purchased with a size of 19 cm  14 cm, in order to fit with the SEPA CFII cell and suit the irradiation glass tube size. This membrane was used because it may be suitable for the treatment of an effluent defined by one of our industrial support. A cross-flow filtration cell (SEPA CFII, GE Osmonics) was used for membrane performance measurement (NaCl rejection and permeability). A feed spacer of 1.2 mm and a permeate collector surrounded membrane sample inside the cell and provided 138 cm2 of effective surface filtration. Feed solution was incorporated into a 4 L stainless steel thermostated cylindrical tank. A constant feed solution temperature of 25 1C70.5 was kept by adjusting the chiller temperature set point. pH (Mettler Toledo, FiveGo) and conductivity 1

1 Gy (Gy) is a unit traditionally use in irradiation. A gray represent the dose absorbed by the material and corresponding to 1 J kg  1.

(Hannah Instrument, HI9865) of the feed solution were also monitored. Feed solution circulation through the filtration cell was carried out by hydracell pump (Wanner Engineering). Applied pressure was adjusted by a back pressure regulator (Tescom) and measured by two pressure gauges (Keller). Feed flow rate across the filtration cell was adjusted with a by-pass valve to obtain a recirculating flow around 250 L h  1. Fig. 1 provides an illustration of the experimental setup used in the study. 2.2. Sample preparation In the following, the term “membrane” will design an entire membrane, whereas the term samples will refer to cut membranes. All membranes were rinsed with pure water (resistivity 15 MΩ cm) and soaked in pure water baths for 24 h at 8 1C to remove preservation agents before both irradiation and/or filtration experiments. Samples used for polymer characterization (attenuated total reflection Fourrier transform infrared spectroscopy ATR-FTIR and X-ray photoelectron spectroscopy XPS) were dried in vacuum for at least 48 h before analysis. Membranes for characterization experiments were cut into 4 bands and then placed into a 100 mL glass tube, and 20 mL of pure water was added to the glass so that samples were entirely immerged. Nitrogen bubbles were used to remove dissolved oxygen and carbon dioxide from water before to be poured into the glass. Irradiation tubes were emptied of atmospheric air by vacuum aspiration. Vacuum absolute pressure was capped at 15 mbar to limit the sample moisture content. Then irradiation tubes were backfilled with pure oxygen for aerobic conditions. This operation was repeated three times to remove the residual air before sealing the tube at an absolute pressure around 900 mbar. Membranes for filtration experiments were directly and entirely placed into 250 mL glass tubes. Pure water purged with nitrogen bubbles was poured into the glass to totally immerge membranes samples. These samples were not sealed and water was naturally gas saturated with atmospheric conditions. Before and after irradiation experiments, samples were stored into a MilliQ water bath at 8–10 1C. Water baths were periodically renewed during the storage. In order to avoid artefact measurement, all irradiation samples were doubled. 2.3. Irradiation conditions Gamma irradiation was carried out using 60Co source in an industrial facility. This irradiation conditions were chosen regarding

66

N. Combernoux et al. / Journal of Membrane Science 480 (2015) 64–73

diamond crystal to the spectrum. Membrane active layer or support layer were then pressed tightly against the crystal plate for analysis. Each spectrum results in an average of 56 scans in the range of 4000–800 cm  1 at 1 cm  1 resolution. ATR corrections were applied after measurement.

Table 1 Sum up of irradiations conditions carried out in the study. Dose (MGy)

Dose rate (kGy h  1)

0.1

0.5

1

0.5

Type of glass tube Sealed Opened Sealed Opened

Membrane size

Conditions

Sample Oxygenþ water Membrane Airþ water Sample Oxygenþ water Membrane Airþ water

Number of glass tube 2 2 2 2

the large amount of samples. Indeed, a sophisticated installation was necessary in order to performed irradiation of several samples in the same time. This facility provides a constant dose rate, approximately 0.5 kGy h  1. The total dose was set to 0.1 or 1 MGy. A sum up of irradiation conditions carried out in this study is given in Table 1. Irradiation was performed at room temperature (24 1C73). The cumulated energy absorbed in the sample itself was measured by dosimeters with a known response to irradiation, proportional to the exposure time, as described by Traboulsi et al. [13]. Furthermore, the samples were irradiated homogeneously by rotating them during exposure. Irradiation time ranges from few days to several months according to the total absorbed dose targeted. 2.4. Membrane characterization 2.4.1. Permeability and salt rejection Pure water at pH  6 was first filtered through the membrane at low applied pressure of 5 bar and the applied pressure was then slightly increased to 32 bar to allow membrane compaction. This pressure was maintained constant for at least 24 h until the water flux remained constant. Then, the membrane pure water permeability was determined. Afterwards, the feed tank was filled with a 1 g L  1 NaCl solution for salt rejection measurement with a transmembrane pressure (TMP) of 28 bar. Concentrations of both feed and permeate were deducted from conductivity measurements with a calibration curve, in order to calculate the observed rejection with following equation: Robs ¼ 1–C p =C a

ð1Þ

where Robs represents the observe rejection (%), Cp the permeate concentration (mol L  1) and Ca the retentate concentration (mol L  1) 2.4.2. Atomic force microscopy (AFM) Surface analysis was performed using an atomic force microscope (SPM Lab) operated in tapping mode using commercially available Si cantilevers (Model TESP-MT, Veeco, Santa Barbara, CA). The manufacturer specified a cantilever frequency of 320 kHz and a constant force of 42 Nm-1. A scan rate of 1.0 Hz was used for a surface scan of 5 μm  5 μm. At least 6 replicates at different areas of a sample were performed to avoid artefacts. 2.4.3. Field emission scanning electron microscopy (FE-SEM) FE-SEM images of the cross-section of virgin and irradiated SE membranes were carried out with a hot cathode ZEISS Supra S5 field emission scanning electron microscope. Before imaging, membranes were submerged in liquid nitrogen, cut by a sharp blade, and then coated with platinum. 2.4.4. Attenuated total reflection-Fourier transform infrared spectroscopy (ATR-FTIR) ATR-FTIR experiments were carried out using an iS50 FTIR spectrometer (Nicolet) equipped with an ATR element (diamond crystal plate) and Omnics 9.2 software (Thermo Electron Corporation). Background spectrum was recorded prior to each experiment to avoid contribution of carbon dioxide, water vapor and

2.4.5. X-ray photoelectron spectroscopy (XPS) XPS experiments were performed on an ESCALAB 250 XPS spectrometer (Thermo Electron Corporation) using Al Kα monochromatic X-ray source (1486.6 eV). The analysis area was 40 mm diameter. Survey scans were collected for binding energies ranging from 0 to 1200 eV. High resolution scans were obtained by averaging 10 scans for C (1s), O (1s), and N (1s) peaks with a resolution of 0.2 eV. Sample charging was compensated by an electron flood gun operated at 2 eV. High resolution spectra were subtracted by Shirley-type background. Deconvolution of high resolution spectra was conducted using CasaXPS software with Gaussian–Lorentz functions. 2.4.6. Characterization of outgoing species during irradiation in the presence of water Water inside irradiation glass tubes was analyzed in order to follow the outgoing species from the RO membrane. Thus, formates, acetates, oxalates, phthalates, chlorides, sulfates and nitrates were analyzed. A Dionex ICS 2100 ion chromatography system was used in gradient mode. The eluent was a gradient of KOH from 3 to 60 mM with flow rate of 0.6 mL min  1. Calibration was carried out using aqueous solutions of H2C2O4, HCOOH, CH3COONH4 (CPAChem), K2SO4, NaNO3, NaNO2, NaCl (VWR) and phthalic acid (Fluka) in the concentration range between 0 and 20 mg L  1. A Dionex ICS 3000 ion chromatography system was also used in 2D mode to quantitatively analyze amines leached from membranes to the supernatant solution. Each column was used in isocratic mode. The eluent was a 30 mM and 2 mM aqueous solution of methanesulfonic acid (Merck) with a flow rate of 1 mL min  1 for the first and second column respectively. Calibration was performed using aqueous solutions of NH4Cl (VWR), CH3NH3Cl, (CH3)2NH2Cl, (CPachem). (CH3)3NHCl (Fluka) in the concentration range between 0 and 20 mg L  1 for each compound except ammonium was in the range of 0–2.5 mg L  1.

3. Results and discussion 3.1. Pure water permeability and salt rejection First, the pure water permeability was assessed for a virgin membrane stored in an ultrapure water bath for 10 days and two months at 8–10 1C. No significant difference was measured compared to a virgin membrane without a storing time. Consequently, it was supposed that membrane performances stayed constant within the exposure times considered in this study. Pure water permeability and NaCl rejection of virgin and irradiated RO membranes are compared in Fig. 2. At 0.1 MGy, the permeability and rejection showed a slight decrease but values remained in the error range. However at 1 MGy permeability exhibited an increase from 2.9 L h  1 m  2 bar  1 to 7.0 L h  1 m  2 bar  1 where as the NaCl rejection decreased from 98.6% to a value around 64%. The permeability and rejection followed the same variation with the dose as that reported by Nakase et al. [24], particularly for the slight decrease at low doses. Those results confirmed that reverse osmosis membranes remained stable for doses lower than 0.1 MGy. However, the degradation of membrane performances occurred before 1 MGy in the present study, whereas Nakase et al. [24] reported a start of the degradation near 2 MGy. This difference might come from the irradiation source type: gamma rays versus accelerated electrons in their study.

N. Combernoux et al. / Journal of Membrane Science 480 (2015) 64–73

67

Fig. 3. RMS roughness of virgin and irradiated at 0.1 and 1 MGy RO SE membranes.

Increase of the membrane roughness might be involved in the decrease of membrane performances and might come from a possible decrease of the top skin layer.

Fig. 2. Pure water permeability (a) and NaCl rejection (b) of virgin and irradiated at 0.1 and 1 MGy RO SE membranes.

In order to explain the rejection changes, first a visual inspection was done. Irradiated membranes (0.1 and 1 MGy) did not exhibit any changes: the color remains a typical brown for the active layer and white for the support layer. Also, no scars were to be found on the surface. Consequently, it stressed the need to carry out more accurate investigations in order to highlight evidence of polymer degradation. In the following SEM, AFM, ATR-FTIR and XPS have been performed for top surface characterization and ion chromatography for the evaluation of ionic species formation during the irradiation. 3.2. Textural analysis of SE membrane surface 3.2.1. Active layer topography analysis with AFM AFM analyses were performed to investigate the topography of the membrane surface. RMS measurement and top surface topography for virgin and irradiated membranes are presented in Figs. 3 and 4 respectively. Roughness increased gradually with the irradiation dose from 11.2 nm for the virgin membrane to 21.4 nm for the 1 MGy irradiated membrane. The value obtained for virgin membrane sample was in agreement with results of Tang et al. for coated reverse osmosis membranes [27]. The two times increase of RMS value for 1 MGy irradiated membrane could be explained by a decrease of the active layer thickness of the membrane probably influenced by the polysulfone underlayer. This result was confirmed by SEM analysis (see Section 3.2.2). In the same way, AFM images clearly revealed a sharper topography for irradiated membranes. The growing size of the apparent “sphere like” forms on the membrane led to a rougher surface.

3.2.2. FE-SEM observation of membrane cross-section FE-SEM experiments were performed in order to clarify the hypothesis of a skin layer thickness decrease. Images were taken on the cross-section of virgin and irradiated membranes to investigate this hypothesis. Fig. 5a–c shows the virgin membrane, the 0.1 MGy and 1 MGy irradiated membranes by FE-SEM observations, respectively. First, the virgin membrane image presented the expected structure: a dense thin skin top-layer and a porous underneath layers representing respectively polyamide and polysulfone similar to RO membranes presented elsewhere [28]. Investigation of the cross-section of irradiated RO membrane revealed some changes. The polysulfone intermediate layer showed a less organized network with growing dose. This observation was evident in the Fig. 5b and c where the porous network was more packed. This phenomenon could be a consequence of crosslinking reactions occurring in polysulfone, as observed for polysulfone membrane [29] and polysulfone film [30] under gamma rays in air. 3.3. Chemical composition of membranes surface 3.3.1. Chemical bonds characterization by ATR-FTIR ATR-FTIR spectra of virgin and irradiated membranes over the range 4000–2400 and 2000–400 cm  1 are presented in Fig. 6. In the range 1800–800 cm  1, both polyamide and polysulfone layers were analyzed because of the depth penetration of the infrared ray. Assignation of interesting peak was carried out according to the bibliography [9,10]. More accurately, three major peaks of the polyamide layer were followed: a band at 1651 cm  1 corresponding to CQO stretching of amide groups, a band at 1721 cm  1 assigned to CQO stretching of esters groups and a large peak centered at 3440 cm  1 assigned to O–H/N–H elongation of the polyamide layer. Wavenumber of the amide peaks underlined a presence of both semi-aromatic and fully-aromatic structures for this membrane as showed by Akin and Temeli [9]. In addition, the band at 1721 cm  1 confirmed the presence of a coating layer, probably PVA type to initiate esterification reactions and interfacial polymerization during the membrane manufacturing. For 0.1 MGy and 1 MGy membranes, the bands at 1651 and 1721 cm  1 exhibited strongly decreases of their absorbance values, as well as O–H/N–H bands at 3450 cm  1. These findings showed that the skin layer of the RO membrane was affected by gamma radiation. To attest the absorbance decrease, a method described by Regula et al. [31] was used. Absorbance height of the interesting

68

N. Combernoux et al. / Journal of Membrane Science 480 (2015) 64–73

Polyamide skin layer

Polysulfone underneath layer

Polyamide skin layer

Polysulfone underneath layer

Polyamide skin layer

Polysulfone underneath layer

Fig. 5. FE-SEM image of virgin (a) and irradiated at 0.1 MGy (b) and irradiated at 1 MGy (c) SE membranes cross section.

Fig. 4. AFM pictures and RMS measurement of virgin and irradiated at 0.1 and 1 MGy SE membranes (scale: 5 mm  5 mm).

bands was divided by the absorbance height of a characteristic band of polysulfone representing the stretching of the aryl band (C–O–C)

at 1242 cm  1. These results are given in Fig. 7. For the band at 3411 cm  1, the relative absorbance decreased as soon as the dose reached 0.1 MGy, and kept decreasing for 1 MGy. More precisely, the variation of N–H bands at 3411 cm  1 was likely to be coming from a loss in hydrogen bonds inside the polyamide structure, according to Sengupta et al. [20]. This hypothesis was confirmed by the slight decrease of relative absorbance of amide bonds at 0.1 MGy. However when the dose reached higher values, the sharp drop at 1 MGy showed that chain breakings also occurred leading to a major decrease in N–H/O–H stretching at 3411 cm  1. Regarding the 1721 cm  1 band, the relative absorbance dropped immediately for 0.1 MGy and 1 MGy. Breakings of ester bonds could occur but this change did not result in a switch in wavenumber as indicated by Kang et al. [32]. Thus, the environment of carbonyl groups was unchanged. A possible mechanism explaining chain breakings occurring has been described by Buttafava et al. [33] and is presented in Fig. 8. This mechanism could explain the decrease of hydrogen bonds by the loss of ester groups and the shift of wavenumber. Amide bonds breakings could also follow a same kind of mechanism as ester groups.

N. Combernoux et al. / Journal of Membrane Science 480 (2015) 64–73

69

Fig. 6. FTIR spectra of non-irradiated, 0.1 MGy and 1 MGy SE membranes for wavenumber ranging from 400 to 2000 cm  1 (a) and 2400–4000 cm  1 (b).

Finally, FTIR spectra highlighted radicals' reaction occurring in the membrane, leading to chain breakings of the polyamide layer. The network structure was relaxed due to a loss of hydrogen interchains bonds, and the breaking of ester and amide bonds. Amide bonds were also assumed to be more resistant to radiation than ester bonds since the breaking occurred above 0.1 MGy.

3.3.2. Elemental composition of top membrane surface by XPS XPS results showed that membranes contained mostly carbon, nitrogen and oxygen. The elemental composition was assessed based on the intensity of S (2p), C (1s), N (1s) and O (1s) peaks centered around 169, 285, 400 and 532 eV, respectively. Atomic composition and O/N, C/O and C/N ratio are reported in Table 2. First, the virgin membrane elemental composition highlighted a weak nitrogen percentage compared to carbon or oxygen, and

consequently a O/N value of 3.1. This finding matched with results reported in the literature [9]. A O/N value above 2.0 results in the addition of PVA-type monomer in the active layer of the membrane for a better fouling resistance [34]. This evidence attested by XPS was consistent with ATR-FTIR spectra. Values of O/N and C/N ratio remained steady at 0.1 MGy and soared for 1 MGy, whereas C/O ratio were constant at a value around 3.5 for this two doses. These findings led to the conclusion that the 10 nm top layer underwent a loss of nitrogen atoms. Thus, the top surface of the skin layer was the place of radical reactions leading to breakings of amide bonds. The surface behaved similarly as the rest of the active layer. Observation of ester bonds breakings were harder, since it resulted in both carbon and oxygen evolution, whereas C/O ratio remained steady. Elemental analysis results also showed an increase in the S percentage, probably due to a decrease of the skin polyamide layer.

70

N. Combernoux et al. / Journal of Membrane Science 480 (2015) 64–73

Fig. 7. ATR-FTIR relative absorbance of three peaks for virgin, 0.1 MGy and 1 MGy irradiated SE membranes.

leading to a loss of amide bonds, ester bonds and a significant decrease in binding energy shift. XPS results also proved that ester groups were less stable than amide groups due to the decrease of both binding shift and percentage of the second peak starting from 0.1 MGy. Crossing XPS and FTIR also revealed a relaxing of the polyamide network due to breakings of ester and amide bonds and the loss of inter-chain hydrogen bonds. Finally, XPS data seemed to show a decrease of the thickness of the polyamide layer because of an increasing amount measurement of sulfur, as showed by AFM. 3.4. Determination of outgoing species during irradiation by ion chromatography Fig. 8. Possible mechanism for the breaking of ester bonds in the polyamide active layer [33].

High resolutions scans provided additional information regarding the chemical environment of carbon atoms. Fig. 9 represents the spectra obtained by the deconvolution of C (1s) for virgin and irradiated membranes. Three peaks were identified: a major peak at 284.8 eV corresponding to a carbons without adjacent withdrawing electrons such as C–C or C–H, a second peak at 286.4 eV assignable to carbons with low electron withdrawing environment (C–O or C–N) and a third peak at 288.2 eV assignable to carbons with strong electron withdrawing environment (O–CQO or N– CQO). Binding energy shifts (δBE) were 1.6 eV and 3.4 eV for the second and third peaks, respectively. The binding energy shift of 1.6 eV is an additional argument for the presence of a coating layer according to Tang et al. [10]. Then, a binding energy shift of 3.4 eV for the third peak is typical for fully aromatic polyamide membrane suggesting that the SE membrane was also composed of aromatic polyamide. Data from XPS high resolution spectra are reported in Table 3 with values from the literature. Percentage of peak II and peak III decreased with the irradiation dose as well as the binding energy shift. These results suggested a loosing of the polyamide layer added to changes in the electron environment of amide carbons and esters carbons. These two findings could be matched with ATR-FTIR spectra: a preferential radical attack to amide groups and ester groups occurs. Breakings of amide bonds or ester bonds led to a decrease in the binding energy shift because of the loss of stability bring by aromatic rings. Then, the evolution of radicals after bonds breakings resulted in outgoings of carboxyl or NH3 as proved by ionic chromatography, with a similar mechanism as described for anionic IER [35]. Consequently, XPS results confirmed ATR-FTIR observations: membrane underwent chain breakings during gamma irradiation

Ion chromatography was carried out to identify outgoing species coming from the degradation of the membrane during irradiation. Results are shown in Fig. 10. Traces of dimethyl amine and trimethyl amine were also found but not quantifiable. Acetate, formate and phthalate species were measured; the presence of these compounds could be a consequence of polymer degradations by radical mechanisms, as for gamma irradiation. Traces of chloride were also noticed, probably coming from the rest of TMC (Trimethyl Chloride) in the membrane, used during the active layer synthesis. Then, ion chromatography revealed the presence of high amount of sulfate and ammonium ions, indicating not only the degradation of active polyamide layer but also the polysulfone layer underneath. Higher the dose was, higher the concentrations of these two species were. The presence of ammonium and nitrate ions confirmed the hypothesis of amide bonds breakings observed by FTIR and XPS analysis. Finally, the high quantity of sulfate ions compared to ammonium ions had to be correlated with the membrane structure—the polysulfone layer thickness was 50 mm whereas polyamide active layer was around 0.1–0.2 mm.

4. Summary and discussion Various and complementary characterizations were performed in this study to provide a deep analysis of the RO membrane behavior under gamma rays in the presence of water and oxygen or air. The whole results presented herein started from the evidence of separation and permeability properties alterations beginning between 0.1 and 1 MGy. More precisely, a sharp decrease of NaCl rejection and increase in water permeability was assessed for Osmonics SE membranes irradiated at 1 MGy. Accurate investigation of chemical compositions by XPS and FTIR highlighted breakings of amide bonds and ester bonds in the active polyamide layer. This conclusion was

N. Combernoux et al. / Journal of Membrane Science 480 (2015) 64–73

71

Table 2 Atomic composition obtained by XPS analysis of virgin SE membrane and irradiated under 0.1 and 1 MGy in sealed glass tube.

Virgin membrane 0.1 MGy membrane 1 MGy membrane

%C

%O

%N

%S

O/N

C/N

C/O

70.07 1.3 73.4 7 0.6 74.3 7 0.9

22.8 7 1.4 18.8 7 1.0 22.2 7 0.1

6.8 7 1.0 7.17 0.5 2.0 7 1.0

0.4 7 0.1 0.6 7 0.1 1.17 0.2

3.3 2.6 10.9

7.1 10.3 36.4

3.1 3.9 3.4

Fig. 9. XPS spectra of C(1s) component for (a) virgin SE membrane, (b) 0.1 MGy SE membrane and (c) 1 MGy SE membrane.

72

N. Combernoux et al. / Journal of Membrane Science 480 (2015) 64–73

5. Conclusion

Table 3 Binding energy shift and peak area for high resolution XPS spectra of C(1s). Membrane

Experiment (this study)

Coated fully aromatic [10] Coated piperazine based [10]

Peak I

SE membrane 0.1 MGy 1 MGy SW30HR BW30 SG membrane DK membrane

Peak II

Peak III

δBE (eV)

%

δBE (eV)

%

δBE (eV)

%

0

57.9

1.6

30.0

3.4

12.1

0 0 0 0 0

65.2 71.9 81.0 64.7 73.3

1.3 1.2 1.6 1.5 1.2

23.8 20.1 12.6 31 15.9

3.3 3 3.3 3 3.3

11.0 7.9 6.5 4.4 10.8

0

68.6

1.1

16.2

2.8

15.2

Gamma rays irradiation of the reverse osmosis SE polyamide membrane were carried out in water with aerobic conditions for two doses (0.1 and 1 MGy) with a dose rate of 0.5 kGy h  1. Characterization of reverse osmosis membranes after irradiation showed the evidence of degradation for 1 MGy. More accurately, the active polyamide layer underwent breakings of amide and ester bonds, leading to a relaxing or an alteration of the polyamide network and an increase in roughness of the surface. Structure modifications via chain breakings were related to a sharp drop in permselectivity properties of the membrane. Consequently, amide and ester bonds were also identified as the more fragile in polyamide top layer of RO membranes. Whereas 1 MGy membranes were affected by gamma rays, membranes irradiated at 0.1 MGy highlighted only minor structure modifications without an alteration of their permselectivity properties. This result proves that RO membranes are radiation resistant until at least a dose level around 0.1 MGy under gamma rays and a dose rate of 0.5 kGy h  1. Further investigations to assess the evolution of properties between 0.1 and 1 MGy are in progress. Finally, a deep modification of the polysulfone intermediate layer was attested.

Acknowledgments

Fig. 10. Ion chromatography results for outgoing of ions in the water bath during irradiation for two doses.

The authors acknowledge AREVA and EDF for providing financial supports to this study. The authors also acknowledged Valerie Flaud from ICG Montpellier for déconvolution of XPS spectra, Karine Ressayre, Charlene Vallat and Thierry Combaluzier from CEA Marcoule/DTCD for performing Ion Chromatography, FE-SEM analysis and irradiation support respectively. References

confirmed by the ion chromatography analysis of outgoing species during the irradiation: ammonium and carbonyl salts (formate, acetate, and phthalate) were identified. As reported in the literature, carbonyl bounds (esters or amides), C–O and C–N simple bonds are widely implied in rejection mechanisms of RO membranes [36]. Indeed, these bonds enable to create a dense highly reticulated interchains polyamide network by the formation of hydrogen bonds. Consequently, the alteration of membrane separation properties might come from a relaxing of the polyamide network due to the breaking of ester and amide bonds. This less crosslinked network leads to an increase in inter-chain voids, and consequently a higher permeability, and weaker rejection [37]. Then, AFM analysis revealed a growing of membrane roughness with the irradiation dose. Consequently, the polyamide top 10 nm layer was not only relaxing but also totally degraded. XPS elemental analysis confirmed this hypothesis since nitrogen atoms were hardly missing (only 2.0% at 1 MGy instead of 6.8% for virgin membrane) and sulfur atoms had quasi-doubled. Thus, this phenomenon highlighted a non-uniform decrease of the active top 10 nm layer, leading in fine to a more peak and valley structure of the top surface. FE-SEM pictures of the top surface gave a visual observation of the cross-section. However, this images were not enough accurate to confirm a slight decrease of the skin layer thickness. Nevertheless, these pictures showed a different behavior of polysulfone with irradiation dose, probably reticulation reactions, that might be deeply studied. Finally, all these arguments seemed to prove that a possible mechanism for ester or amide bonds breakings could be similar to the one presented in Fig. 8 [33]. Also, it can be inferred that amides and esters are the more sensitive bonds in polyamide RO membrane under gamma irradiation.

[1] C. Poinssot, S. Bourg, N. Ouvrier, N. Combernoux, C. Rostaing, M. VargasGonzalez, J. Bruno, Assessment of the environmental footprint of nuclear energy systems. Comparison between closed and open fuel cycles, Energy 69 (2014) 199–211. [2] R.O.A. Rahman, H.A. Ibrahium, Y.-T. Hung, Liquid radioactive wastes treatment: a review, Water 3 (2011) 551–565. [3] D. Rana, T. Matsuura, M.A. Kassim, A.F. Ismail, Radioactive decontamination of water by membrane processes — a review, Desalination 321 (2012) 77–92. [4] A.G. Chmielewski, M. Harasimowicz, B. Tyminski, G. Zakrzewska-Trznadel, Concentration of low- and medium-level radioactive wastes with three-stage reverse osmosis pilot plant, Sep. Sci. Technol. 36 (2001) 1117–1127. [5] L.F. Greenlee, D.F. Lawler, B.D. Freeman, B. Marrot, P. Moulin, Reverse osmosis desalination: water sources, technology, and today's challenges, Water Res. 43 (2009) 2317–2348. [6] C. Fritzmann, J. Löwenberg, T. Wintgens, T. Melin, State-of-the-art of reverse osmosis desalination, Desalination 216 (2007) 1–76. [7] K. Häyrynen, E. Pongrácz, V. Väisänen, N. Pap, M. Mänttäri, J. Langwaldt, R.L. Keiski, Concentration of ammonium and nitrate from mine water by reverse osmosis and nanofiltration, Desalination 240 (2009) 280–289. [8] H. Ozaki, K. Sharma, W. Saktaywin, Performance of an ultra-low-pressure reverse osmosis membrane (ULPROM) for separating heavy metal: effect of interference parameters, Desalination 144 (2002) 287–294. [9] O. Akin, F. Temelli, Probing the hydrophobicity of commercial reverse osmosis membranes produced by interfacial polymerization using contact angle, XPS, FTIR, FE-SEM and AFM 278 (2011) 387–396Desalination 278 (2011) 387–396. [10] C.Y. Tang, Y. Kwon, J.O. Leckie, Effect of membrane chemistry and coating layer on physiochemical properties of thin film composite polyamide RO and NF membranes: I. FTIR and XPS characterization of polyamide and coating layer chemistry, Desalination 242 (2009) 149–167. [11] A. Canãs, M.J. Ariza, J. Benavente, Characterization of active and porous sublayer of a composite reverse osmosis membrane by impedance spectroscopy, streaming and membrane potentials, salt diffusion and X-ray photoelectron spectroscopy measurement, J. Membr. Sci. 183 (2001) 135–146. [12] R.L. Clough, High energy radiation and polymers: a review of commercial processes and emerging applications, Nucl. Instrum. Methods Phys. Res. Sect. B: Beam Interact. Mater. At. 185 (2001) 8–33. [13] A. Traboulsi, V. Labed, V. Dauvois, N. Dupuy, C. Rebufa, Gamma radiation effect on gas production in anion exchange resins, Nucl. Instrum. Methods Phys. Res. Sect. B: Beam Interact. Mater. At. 312 (2013) 7–14.

N. Combernoux et al. / Journal of Membrane Science 480 (2015) 64–73

[14] V. Plaček, V. Hnát, R. Pejša, T. Kohout, Testing of polysulfone for applications in nuclear facilities, J. Appl. Polym. Sci. 109 (2008) 2395–2399. [15] S. Diao, K. Jin, Z. Yang, H. Lu, S. Feng, C. Zhang, The effect of phenyl modified fumed silica on radiation resistance of silicone rubber, Mater. Chem. Phys. 129 (2011) 202–208. [16] R.L. Clough, K.T. Gillen, M. Dole, Radiation resistance of polymers and composites, in: D.W. Clegg, A.A. Collyer (Eds.), Irradiation Effects on Polymers, Elsevier Applied Science, Elsevier, London, UK, 1991. [17] B. Li, L. Zhang, ESR approach to free radicals traped in irradiated polyamide1010, Radiat. Phys. Chem. 49 (1997) 575–579. [18] J.S. Félix, M. Monteiro, J.E. Manzoli, M. Padula, D. Pezo, J. Romero, C. Nerín, Identification and migration of degradation compounds from irradiation of multilayer polyamide 6 films for meat foodstuffs and cheese, Anal. Bioanal. Chem. 391 (2008) 847–857. [19] W. Feng, F.M. Hu, L.H. Yuan, Y. Zhou, Y.Y. Zhou, Radiation crosslinking of polyamide 610, Radiat. Phys. Chem. 63 (2002) 493–496. [20] R. Sengupta, V.K. Tikku, A.K. Somani, T.K. Chaki, A.K. Bhowmick, Electron beam irradiated polyamide-6,6 films—I: characterization by wide angle X-ray scattering and infrared spectroscopy, Radiat. Phys. Chem. 72 (2005) 625–633. [21] A.G. Chmielewski, M. Harasimowicz, Influence of gamma and electron irradiation on transport properties of nanofiltration and hyperfiltration membranes, Nukleonika 42 (1997) 857–862. [22] J.M. Arnal, M. Sancho, G. Verdú, J.M. Campayo, J.I. Villaescusa, Treatment of 137C liquid wastes by reverse osmosis part I. Preliminary tests, Desalination 154 (2003) 27–33. [23] E.E.M. Oliveira, C.C.R. Barbosa, J.C. Afonso, Selectivity and structural integrity of a nanofiltration membrane for treatment of liquid waste containing uranium, Membr. Water Treat. 3 (2012) 231–242. [24] Y. Nakase, T. Yanagi, T. Uemura, Irradiation effects on properties of reverse osmosis membrane based on crosslinked aromatic polyamide, J. Nucl. Sci. Technol. 31 (1994) 1214–1221. [25] A. Antony, R. Fudianto, S. Cox, G. Leslie, Assessing the oxidative degradation of polyamide reverse osmosis membrane—accelerated ageing with hypochlorite exposure, J. Membr. Sci. 347 (2010) 159–164. [26] A.G. Chmielewski, M. Haji-Saeid, S. Ahmed, Progress in radiation processing of polymers, Nucl. Instrum. Methods Phys. Res. Sect. B: Beam Interact. Mater. At. 236 (2005) 44–54.

73

[27] C.Y. Tang, Y. Kwon, J.O. Leckie, Effect of membrane chemistry and coating layer on physiochemical properties of thin film composite polyamide RO and NF membranes II. Membrane physiochemical properties and their dependence on polyamide and coating layers, Desalination 242 (2009) 168–182. [28] P.S. Singh, A.P. Rao, P. Ray, A. Bhattacharya, K. Singh, N.K. Saha, A.V.R. Reddy, Techniques for characterization of polyamide thin film composite membranes, Desalination 282 (2011) 78–86. [29] N.N. Rupiasih, P.B. Vidyasagar, Comparative study of effect of low and medium dose rate of γ irradiation on microporous polysulfone membrane using spectroscopic and imaging techniques, Polym. Degrad. Stab. 93 (2008) 1300–1307. [30] K. Murakami, H. Kudo, Gamma-rays irradiation effects on polysulfone at high temperature, Nucl. Instrum. Methods Phys. Res. Sect. B: Beam Interact. Mater. At. 265 (2007) 125–129. [31] C. Regula, E. Carretier, Y. Wyart, M. Sergent, G. Gésan-Guiziou, D. Ferry, A. Vincent, D. Boudot, P. Moulin, Ageing of ultrafiltration membranes in contact with sodium hypochlorite and commercial oxidant: experimental designs as a new ageing protocol, Sep. Purif. Technol. 103 (2013) 119–138. [32] G.-D. Kang, C.-J. Gao, W.-D. Chen, X.-M. Jie, Y.-M. Cao, Q. Yuan, Study on hypochlorite degradation of aromatic polyamide reverse osmosis membrane, J. Membr. Sci. 300 (2007) 165–171. [33] A. Buttafava, G. Consolati, M. Mariani, F. Quasso, U. Ravasio, Effects induced by gamma irradiation of different polyesters studied by viscometry, thermal analysis and positron annihilation spectroscopy, Polym. Degrad. Stab. 89 (2005) 133–139. [34] C.Y. Tang, T.H. Chong, A.G. Fane, Colloidal interactions and fouling of NF and RO membranes: a review, Adv. Colloid Interface Sci. 164 (2011) 126–143. [35] A. Leybros, A. Roubaud, P. Guichardon, O. Boutin, Supercritical water oxidation of ion exchange resins: degradation mechanisms, Process Saf. Environ. Prot. 88 (2010) 213–222. [36] S.Y. Kwak, S.G. Jung, S.H. Kim, Structure-motion-performance relationship of flux-enhanced reverse osmosis (RO) membranes composed of aromatic polyamide thin films, Environ. Sci. Technol. 35 (2001) 4334–4340. [37] S.H. Kim, S.Y. Kwak, T. Suzuki, Positron annihilation spectroscopic evidence to demonstrate the flux-enhancement mechanism in morphology-controlled thinfilm-composite (TFC) membrane, Environ. Sci. Technol. 39 (2005) 1764–1770.