Kin Kit Fong ,Inn Shi Tan *,Henry Chee Yew Foo ,Man Kee Lam ,Adrian Chiong Yuh Tiong ,Steven Lim

1 Department of Petroleum Engineering,Faculty of Engineering and Science,Curtin University,CDT 250,98009 Miri,Sarawak,Malaysia

2 Department of Chemical Engineering,Faculty of Engineering and Science,Curtin University,Miri 98009,Malaysia

3 Chemical Engineering Department,Universiti Teknologi PETRONAS,Seri Iskandar 32610,Malaysia

4 HICoE-Centre for Biofuel and Biochemical Research,Institute of Self-Sustainable Building,Universiti Teknologi PETRONAS,Seri Iskandar 32610,Malaysia

5 Department of Chemical Engineering,Lee Kong Chian Faculty of Engineering and Science,Universiti Tunku Abdul Rahman,Kajang 43000,Malaysia

6 Centre for Photonics and Advanced Materials Research,Universiti Tunku Abdul Rahman,Kajang 43000,Malaysia

Keywords:Graphene oxide κ-Carrageenan Adsorbents Demulsification Composites Waste water

ABSTRACT The rising production of produced water from oilfields had been proven to bring detrimental environmental effects.In this study,an efficient,recyclable,and environmental-friendly reduced graphene oxide immobilized κ-Carrageenan hydrogel composite (κCaGO) was fabricated as an alternative sorbent for crude oil-in-water demulsification.Polyethyleneimine(PEI)was employed to form a stable hydrogel composite.The conditions for the immobilization of graphene oxide(GO)on PEI-modified κ-Carrageenan(κC)beads were optimized appropriately.An immobilization yield of 77%was attained at 2%PEI,2 h immobilization activation time,and pH 6.5.Moreover,the synthesized κCaGO is capable of demulsification with an average demulsification efficiency of 70%.It was found that the demulsification efficiency increases with salinity and κCaGO dosage,and it deteriorates under alkaline condition.These phenomena can be attributed to the interfacial interactions between κCaGO and the emulsion.Furthermore,the κCaGO can be recycled to use for up to six cycles without significant leaching and degradation.As such,the synthesized κCaGO could be further developed as a potential sorbent substitute for the separation of crude oil from produced water.

1.Introduction

Produced water is an oilfield waste obtained in a high quantity of up to 4 million gallons per well during the production of petroleum and gas[1].The produced water may originate from one of these two causes,which are the naturally trapped water from the underground formations or the production activity of hydraulic fracturing or polymer flooding.As the water was believed to be initially present in the rock formation before the invasion of oil during its migration process,the reservoir rocks would have absorbed both saline water and hydrocarbons,which are more commonly known as connate water or formation water[2].Upon the commencement of petroleum production,the connate water and formation water will then be known as produce water.For enhanced oil recovery efforts,fracturing fluid is pumped into the formation under high pressure to increase the mobility of the hydrocarbon for extraction[3].However,the high pressure generated inside the rock formation will lead to the mixing of fracturing fluid,brine,hydrocarbons,and formation water at the subsurface,which is then brought back up to the surface through the production wellbore as produced water.

On the other hand,polymer flooding is also used as an alternative enhanced oil recovery operation.This method uses its flood water comprising of surfactant and water-soluble polymer to be injected into the well to displace the petroleum and gas and increase its production efficiency[4].As the main component of the injected solution is water,it will also become a part of the produced water during this recovery effort.Therefore,the generation of produced water is inevitable for petroleum production,which calls for the need of good treatment methods for the produced water.

The increasing production of produced water has caused great concern to many due to its harmful contaminants and other detrimental effects on the environment if disposed in onshore or offshore areas[5,6].Heavy metal contaminants may be of low levels in the environment,but it can be severely toxic when the concentration increases[7].However,produced water is unavoidable due to the excessive need and reliance on petroleum for the economy and technology.This would result in the need for cost-effective and efficient treatment for wastewater treatment.Recently,graphene-based materials had become wellknown for its appealing adsorption qualities,which resulted in the application of graphene oxide for oil separation.Graphene oxide(GO)is also amphiphilic,which can be used to further enhance the separation between any type of liquid and water.However,graphene oxides are often not recycled due to its easy dispersion in water and difficulty in the separation of adsorption substrates[8].Due to the lack of sustainability,frequent fabrication is required leading to higher costs.Apart from that,graphene oxide also possesses large surface areas,multilayered structure,and abundance of oxygenic functional groups that help them to synthesize easily with other chemicals and even solidify radionuclides from wastewater[9–11].

Polyethyleneimine(PEI)is an inexpensive polymer with low toxicity,which could also be used in oil removal and separation application.When polyethyleneimine is synthesized with graphene oxide,its application may be applied in the adsorption of anionic species as well.On the other hand,the synthesis of PEI and polysaccharides has also shown good mechanical traits and having selective sorption properties.κ-Carrageenan(κC)is a versatile polysaccharide with high adsorption capabilities,which makes them suitable for the separation of oil and water.κC is known to have poor environmental stability and poor gel strength,hence leading to degradation during the oil removal process[12].However,it could be improved to attain hydrophobicity and as a linker with the immobilization of a functional polymer such as polystyrene and PEI[13].Therefore,the development of synthesis between GO,PEI,and κC may be able to improve their flaws in terms of leaching and degradation.Subsequently,this can provide an improved sorbent alternative for effective oil-in-water demulsification.

A simple,green,and low-cost approach was adopted for the fabrication of reduced GO hydrogel composite(κCaGO)in this study.The assembled κCaGO was then characterized for the physicochemical properties using Fourier Transform Infrared Resonance(FTIR)Spectra to determine the presence of functional groups for bonding mechanism,Ultraviolet–Visible Spectroscopy(UV–Vis)to analyze the adsorption capability,Scanning Electron Microscopy(SEM)to observe the surface morphology,Thermogravimetric Analysis(TGA)to analyze the heat capacity,Brunauer–Emmett–Teller(BET)method to calculate the specific surface area(SSA),and Barrett–Joyner–Halenda(BJH)method to calculate the pore size distribution.Finally,drop shape analysis(DSA)was also used to determine its hydrophilicity.In addition to that,the modified κC beads were subjected to varying parameters,which include the concentration of PEI,activation time,and effect on its pH value to ensure the best immobilization yield of GO on the surface of the hydrogel beads.Moreover,to ensure its practical application,the emulsion separation was tested through different parameters(effect of salinity,dosage,pH,and reusability).

2.Materials and Methods

2.1.Materials

In the present work,red macroalgae (Eucheuma cottonii) was purchased from Natural Health Farm.The macroalgae was rinsed with distilled water before used for the subsequent experiments.Concentrated sulfuric acid was purchased from MERCK(Malaysia),graphite powder,glutaraldehyde solution(GA)(25%),PEI,dichloromethane anhydrous(DCM),phosphoric acid,potassium permanganate,hydrochloric acid,hydrogen peroxide,potassium chloride were purchased from Sigma-Aldrich(Malaysia).Four formulations were used upon their synthesis which are formulation 1:κC,formulation 2:κC added with GA (κC/GA),formulation 3:κC/GA synthesized with PEI (κC/GA/PEI),and formulation 4:κC/GA/PEI coated with GO(κC/GA/PEI/RGO).

2.2.Synthesis of GO

GO was synthesized based on the improved Hummers method[14]and revised in our laboratory,whereby NaNO3was excluded while the amount of oxidizing agent,KMnO4,was increased.A 9:1 mixture of concentrated H2SO4and H3PO4was prepared under constant agitation.1%(w/v)of graphite powder and 6%(w/v) of KMnO4was slowly added and stirred vigorously.Subsequently,the reaction mixture was kept under constant agitation at 50 °C for 6 h to complete the oxidation process.The reaction mixture was then cooled to room temperature and transferred to a beaker containing 200 ml deionized flaked ice.1.5% (v/v)of 30% H2O2was added into the reaction mixture and stirred for 30 min to stop the oxidation process.The mixture was centrifuged using a refrigerated centrifuge(Hettich Refrigerated Centrifuge)for 50 min at 4000 r·min−1to remove the residual graphite.The brownish-black paste was washed with 200 ml of 5 wt%hydrochloric acid,followed by repeated washing with deionized water until a neutral pH value was obtained.The graphite oxide produced was then added with deionized water and transferred to a blue cap bottle.The mixture was placed in an ultrasonic bath (Desen Digital Ultrasonic Cleaner 4.0 L 100 W DSA100-SK2) for 1 h to exfoliate the graphene sheets from the graphite oxide structure.The product was then centrifuged at 4000 r·min−1for 40 min to remove the non-exfoliated graphite oxide.Finally,it was dried overnight in a drying oven (UM500,Memmert,German)at 60°C.

2.3.κC extraction

The Eucheuma cottonii was initially cut into smaller pieces for weighing before washed with distilled water.Then,3% (w/v) of dry Eucheuma cottonii was boiled for 2 h at a constant temperature of 90 °C.The κC extract was then separated from its residue through vacuum filtration.The extracted residue was dried at 40°C until a constant weight was reached.It was then pulverized and reboiled for an additional 30 min with 2 L of distilled water before separation through vacuum filtration.The extracted filtrate was known to be the extracted κC[15].

2.4.Preparation and modification of κC hydrogel beads

Initially,a 2%(w/v)κC solution was prepared by mixing the κC beads and deionized water at 70°C using a hotplate and was stirred until the κC beads were completely dissolved.Then,0.25%of GA was added into the κC solution to be used as the crosslinking agent.It was placed in a thermostatic shaking water bath(Model KW-1000 DC,DFS)for 24 h at 50°C.The solution was then transferred into a 50 ml burette,and it was added dropwise to a 0.3 mol⋅L–1potassium chloride solution,which is stirred using a magnetic stirrer.The preparation method would yield uniform κC beads with an approximate diameter of about 2.5 mm.The modified κC beads were kept in the solution to solidify at 25°C for 3 h.The beads were then filtered out,and it washed thoroughly with distilled water twice.

The crosslinking step for the modified κC and PEI started by washing the modified κC beads with distilled water.The modified κC beads were then soaked in 1%GA solution for 2 h.The GA was used as a catalyst to activate the immobilization process between the κC and PEI through covalent bonding.GA was also used as a spacer in this covalent immobilization process.It allowed intense covalent binding on multiple points,which could facilitate good rigidity and better stability to the immobilization of PEI to κC[16,17].The activated κC beads were rinsed with distilled water to remove any unreacted GA solution.This activation step for the beads was conducted with a PEI concentration of 2%(w/v).

2.5.Synthesis of reduced graphene oxide hydrogel composite(κCaGO)

1% (w/v) synthesized GO was dispersed in an ultrasonic bath for 30 min.Then,the modified κC beads with PEI were washed with deionized water to remove any unreacted PEI solution.Then,the hydrogel beads were added into the dispersed graphene oxide mixture and maintained in a thermostatic shaking incubator at 40°C for 2 h.During this incubation process,the self-assembly of graphene oxide sheets on the PEI layer occurred through hydrogen bonding,which led to the formation of the reduced graphene oxide/polyethyleneimine(RGO/PEI)layer surrounding the hydrogel.The final synthesized hydrogel was annotated as fabricated κCaGO.The κCaGO was rinsed with deionized water and stored for further characterization and demulsification tests.The effect of different synthesis formulation was studied,as shown in Table S1,and the best immobilization yield process was used for subsequent demulsification tests.

2.6.Analytical methods

The κCaGO beads obtained from Section 2.2 to Section 2.5 were characterized.All the samples prepared,as described above,were subjected to oven drying.The κCaGO beads obtained in Section 2.5 would undergo a demulsification test.

2.6.1.FTIR spectroscopy

FTIR spectroscopy was used to detect the presence of the functional group in the beads.The infrared spectra of all formulations were obtained using a Fourier transform infrared spectroscopy(IR-Prestige-21,Shimadzu,Japan).FTIR spectra of the samples pressed in KBr pellet were obtained at a resolution of 2 cm−1between 4000 cm−1and 500 cm−1at ambient temperature.The characteristic peaks were recorded.

2.6.2.Ultraviolet–visible spectroscopy(UV–Vis)

The UV–Vis spectroscopy was performed on the UV–Vis Spectrophotometer,Lamda 25 UV/Vis Double Beam.This characterization studied on the absorptions band of the sample.The functional group can be obtained by determining the peak curve at the respective wavelength.Deionized water was used as the blank sample to make a reference point for the sample.The sample of 1%(mass)was then diluted with deionized water and analyzed.The maximum absorption band was observed based on the preset wavelength range of 200–360 nm on the Perkin-Elmer Lamda 25 Spectrophotometer.

2.6.3.Scanning electron microscopy(SEM)

For the characterization of formulation 3(κC/GA/PEI)and formulation 4 (κC/GA/PEI/RGO),the surface morphology can be obtained through observation of the micrograph.The samples were initially dried and pounded into dry powder.The dry powder was then mounted on a sample holder where it was coated with a conductive metal.The sample would then undergo scanning electron microscopy (SEM,Hitachi S-3000 N)at an accelerating voltage of 5 kV in a vacuum environment upon spluttering Pt-Pd on the surface of the sample.

2.6.4.Thermal gravimetric analysis(TGA)

The TGA/DSC analysis was performed on the Mettler-Toledo TGA/DSC1(METTLER TOULEDO,Switzerland).This characterization method studies on the weight loss of residues.The significance of temperature on weight loss can be studied for their complete heating profile of residues.The sample used for the analysis was approximately 15 mg,with their temperature set from 50°C to 600°C.High purity nitrogen was used under the set condition of 100 ml·min−1with the heating rate of 10°C·min−1.

2.6.5.Brunauer–Emmett–Teller(BET)

The specific surface area was calculated using a BET method via nitrogen(N2)adsorption–desorption isotherms measurement on a physical adsorption instrument (Micromeritics,ASAP 2920) at 77 K.According to the nitrogen adsorption data,the pore size distributions(PSD)of formulation 4(κC/GA/PEI/RGO)was obtained by using BJH method.

2.6.6.Drop shape analysis(DSA)

A drop shape analyzer was used to identify the interaction between a fluid and a solid phase.This analysis was performed on the DSA25 Model,Kruss,Germany,whereby the contact angle was determined using the pre-installed plug-in from the software.The samples were tested for their oleophobicity and hydrophobicity.10 ml syringe with a steel needle was used to conduct the sessile test method with water and crude oil as its fluids.The data was obtained and used for further analysis.

2.7.Crude oil–water emulsion preparation and characterization

2.7.1.Preparation of crude oil-in-water emulsion

The crude oil samples were obtained from Crude Oil Terminal and were heated at 60°C for 1 h to ensure that it dispersed well and had good mobility.The emulsions were prepared by using brine as the continuous p hase with different salinity of 5000 40,000 and100,000 mg·L−1,and the crude oil samples were at thedispersed phase to simulate thecondition of a low,average and high emulsion salinity.The oilin-wateremulsion was prepared at 5%(mass)andwas subjected to stirringusinganultrasonic homogenizer(LabsonicM,Sartorious)at6000 r·min−1for10 min.The concentrationofoilinwater wasmeasured at 5%(w/v).

2.7.2.Demulsification test

The demulsifying capability was evaluated using the synthesized κCaGO in the prepared emulsion of 6 mg·ml−1.The emulsion was poured into a separating funnel for 15 min before transferred into a centrifuge tube.The mixture was thoroughly shaken with the assistance of a vortex mixer for 3 min to ensure that the κCaGO and emulsions were well-mixed.A cylinder containing oil-in-water emulsion with the same water volume was used as a reference.Ambient conditions were set for the emulsion whereby the oil and water separation was observed.The reference oil-in-water emulsion and the κCaGO driven demulsification process was monitored and photographed with an optical microscope(Nikon Eclipse E200).The emulsion samples were placed on a cleaned glass slide and covered with a coverslip.

Next,the residual oil content of the separated emulsion samples was determined through their absorbance in an Ultraviolet–Visible Spectroscopy(UV–Vis)at a path length of 1 cm.DCM was used as its extraction solvent and was taken as its reference.Through a standard calibration graph,the relationship between the sample's absorbance and the DCM-diluted oil concentration was determined to be linear at the wavelength of 260 nm.The κCaGO was used to treat the crude oil in water emulsions and trialed with varying parameters such as pH,salinity,and dosage of κCaGO.The demulsified samples were collected from each centrifuge tube after agitation and gravity settling of 15 min.Subsequently,the water extracted from the samples was diluted with DCM.Then,the residual oil concentration was obtained through the pre-determined standard calibration curve.The oil concentration in the demulsified water sample was calculated using the following equation:

where Co(mg·L−1)is the concentration of oil,mo(mg)is the mass of oil in the standard curve,and Vw(ml) is the volume of water.The demulsification efficiency was calculated through the difference between the initial and final absorbance value of the mixture.The equation is as shown below:

where DE(%)is the efficiency value of the demulsification process,Co(mg·L−1)is the value of the initial oil content in the prepared emulsion,Ci(mg·L−1) is the value of the residual oil content after the demulsification process with κCaGO.Each sample was tested three times,and the reported values are the average values of the test results.

2.7.3.Reusability study

The reusability of the κCaGO was defined through the repeated testing on its oil separation capability.This study would help to identify the potential threat of leaching that might occur if used repeatedly.To evaluate the reusability,the κCaGO that undergo its demulsification test was extracted through filtration and washed with DCM to ensure that the oil was fully extracted from its surface.Then,it was rinsed with deionized water to remove any excess DCM present on the surface of the κCaGO.The demulsification test was repeated for 10 cycles on the same κCaGO sample,and its demulsification efficiency measurement was recorded for each cycle.The measurements were compared and used to evaluate the extent of its reusability of the κCaGO sample.

3.Results and Discussion

3.1.Optimization of the immobilization process

In the effort of obtaining the best synthesis conditions for modified κC composite beads with GO,the influences,and conditions that govern its immobilization process was investigated.Fig.1(a)reveals that the concentration of PEI plays an important role in the effect of immobilization.The effect of the increasing PEI concentrations from 1%to 2%led to an increase in immobilization yield of 60%to 77%.The results could be explained through the increase of active amine group available for conjugation through covalent grafting with the carboxyl and epoxy group of the GO[18].The number of GA present on the modified κC beads at the given concentration would have most likely be consumed by the amine groups of 2%PEI and that the subsequent increase in the concentration of PEI would have led to an excess amount of stoichiometric PEI molecules that caused a limiting effect on the crosslinking reactions[19].The presence of the excess PEI acted as a hindrance for the crosslinking reaction of the hydrogel formation.A higher amount of amine groups will form a more compacted and denser microstructure on the crosslinking sites of the hydrogel,which would lead to unstable swelling and disintegration[20].In a consequence of that,the apparent immobilization activity decreased,subsequently causing the decrease in active amine group sites required for bonding with graphene oxide.

It was also expected that there would be degeneration in the hydrogel due to its unreacted amine groups on the κCaGO hydrogels as it was unlikely that a total reaction of amines in a PEI chain with the activated κC due to the reduction of mobility and conformation of polymer chain during the crosslinking reaction,which leads to a detrimental reactivity[19].Therefore,the PEI concentration of 2%was taken as the optimized value and was tested for the optimization of the immobilization activation time.

The immobilization yield of graphene oxide is closely related to the amount and presence of active amine groups of PEI surrounding the modified κC hydrogel beads.Therefore,the required activation time for κC-PEI-RGO was studied as an important parameter affecting the availability of active sites for the duration of 0.5,1,2,3,and 4 h.As observed in Fig.1(b),the increase in duration resulted in an increase in immobilization yield.The optimum attained activation time was within the 2 h mark with maximum immobilization yield of 77%whereby the results obtained beyond it led to a minor decrease of immobilization yield.It was noteworthy that the immobilization activation time played a significant role in the synthesis of κCaGO through its graphene polymerization.Stronger bonds between GO and the modified κC beads would result in better stability in terms of thermal,mechanical as well as degradation time[21].Therefore,if insufficient activation time is provided for this immobilization process,the graphene oxide may not be fully reduced into the RGO layer surrounding the κCaGO.Graphene oxide,which has a negatively charged surface,will result in good solubility in water,which may cause leaching when conducting the demulsification tests.

Fig.1.Effect of(a)PEI concentration,(b)activation time,and(c)pH value of PEI solution on the immobilization yield.

On the other hand,this could potentially result in the shrinkage or swelling of the hydrogel with prolonged activation time.This was due to the differences between the hydrogel and the external solution,which led to a lower immobilization yield[22].Since a prolonged activation time might lead to structural damage to the hydrogel beads while short activation time might cause a lack of GO attachments on the modified κC hydrogel beads,further experiments were conducted at the optimized condition of 2 h.

The influence of pH values on the synthesis of κCaGO hydrogel beads could be observed from Fig.1 (c).The immobilization yield shows higher activity in the acidic condition compared to basic conditions.In an acidic environment,lower pH value resulted in the gradual protonation of ionized -COO−groups which caused larger depositions of graphene oxide.This would lead to increased water solubility in an acid environment and decreasing water solubility when rising to a neutral state[23].

However,upon alteration of the pH value from 7 to 8,the occurrence of protonation of carboxyl groups and its partial removal of the oxygencontaining functional groups,which are the bonding sites with the amine group of PEI on the modified κC hydrogel beads,lead to the decrease in immobilization yield [24].Besides,a high pH value might cause GO to micellize due to the electrostatic repulsions and subsequently became a hindrance for the immobilization of GO on the surface of the modified κC hydrogel beads [25].Therefore,the optimum pH value obtained from the 5 cases was pH 6.5 and it was used for the subsequent experiments.

3.2.Synthesis mechanism of κCaGO hydrogel composite

Based on the production of κC and glutaraldehyde(GA)as shown in Fig.2(a),it was reported that the production of κC films and 0.25%GA solution was done through immersion in a thermostatic water bath at 50°C for 24 h.It was observed that the GA was able to crosslink with the hydroxyl groups of the κC when the levels of GA were larger than 0.027 g GA per 1 g of the polymer [26].The resulting hydrogel was also identified to be sensitive to the pH values.Next,the reaction between the activated κC beads and branched PEI was schematically represented,as shown in Fig.2(b).

In order to obtain the κCaGO hydrogel composite,the bonding between RGO/PEI composite was studied to ensure the success of the insitu synthesis of GO and PEI that surrounded the modified κC hydrogel.Henceforth,Fig.S1 shows an illustration of the fabrication procedure that would occur between GO films and PEI.The synthesis between GO and PEI would produce covalent and hydrogen bondings,which allowed it to withstand higher stress on the interface.Since PEI films are brittle at low temperatures,the RGO/PEI composite can help to increase its ductility due to its cryofracture surface [27].Therefore,the conjugation surface of RGO on the surface of the κCaGO will indirectly improve its ductility and stability of the hydrogel beads as it undergoes tests for demulsification.As a result,the amide bonds between the−NH2of PEI and −COOH group of the GO had led to the reduction of oxygen functional groups to form RGO/PEI composite.

Finally,the crosslinking of the proposed κCaGO composite was formed between the oxygen-containing groups of the GO sheets with the amino groups of the PEI and shown in Fig.2(c).As discussed previously,the successful conjugation between the carboxyl groups of graphene oxide and the amino groups of PEI would lead to instantaneous reduction and was further characterized using functional group analysis(FT-IR).

3.3.Characteristics of κCaGO

The Fourier-Transform Infrared(FT-IR)Spectroscopy analysis to investigate the new functional groups was conducted on the synthesis processes of κCaGO to ensure its successful conjugation.Fig.S2(a–b)introduces the FTIR Spectra of graphite to graphene oxide using the improved hummers method.Fig.S2(c–f)shows the FTIR spectra of formulation 1(κC)as its control,formulation 2(κC/GA),formulation 3(κC/GA/PEI),and formulation 4(κC/GA/PEI/RGO),respectively.Based on Fig.S2 (b),the FTIR spectrum of graphene oxide showed an abundance in oxygen functional groups,especially at the stretching of the hydroxyl group from 3200 to 3700 cm−1and at 1404 cm−1of O-H and C-OH.Other functional groups present include anhydride CO-OCO group at 1140 cm−1,C-O group at 1252 cm−1,aromatic C=C group from unoxidized sp2bonds at 1618 cm−1,and the stretching band of C=O at 1722 cm−1.This infrared spectrum was in agreement with the GO spectra and depicted the result of a successful synthesis of graphene oxide [28,29].As the synthesized graphene oxide was shown to be similar to the literature,the improved hummers method was used in the subsequent formulation.

Fig.S2 (c–f),is a combined functional group analysis from the starting point of κC beads,the addition of glutaraldehyde as support,crosslinking with PEI and the final fabrication of κCaGO.Hence,they allow the identification and determination of the successful synthesis process of κCaGO hydrogel composite.Based on formulation 1,the peak bands located at 842,924,1065,and 1252 cm−1are observed to be attributed to properties of κC,which are D-galactose-4-sulfate,3,6-anhydro-D-galactose,linkage of glycoside and stretching of sulfate esters,respectively.A medium band at 1642 cm−1shows the stretching of strong C=O amide group properties while the broadband observed at 3200–3600 cm−1is caused by the stretching of-OH groups in the polysaccharide.It was determined that the infrared spectra of extracted carrageenan from Eucheuma cottonii exhibited similar bonding groups of kappa-carrageenan[30].

In formulation 2,the chemical reaction between the hydroxyl group of κC and the crosslinker GA could be observed through the reduction in absorption bands of the hydroxyl group.This reduction of adsorption bands indicated that the reaction between κC and GA was successful,whereby hydroxyl groups were consumed during the crosslink reaction.Subsequently,the loading of PEI on the activated κC beads was observed in the IR spectrum of formulation 3[31].Formulation 3 exhibits a significant decrease in the peak intensity at 1065 cm−1while the peak at 842 cm−1disappeared,which indicates the successful interaction of the protonated amine group of PEI together with the sulfate groups of the κC.It can also be identified as the κC-PEI polyelectrolyte complex.Upon the addition of graphene oxide into formulation 3,the peaks at 2842 and 2945 cm−1were observed.These peaks indicate the attributes of the stretching of C-H.It can also be observed that the peak at 1722 cm−1,which corresponds to the C=O stretching vibration disappeared upon the conjugation of graphene oxide onto the polyelectrolyte complex.Apart from that,the presence of bands at 842 and 1246 cm−1represents the stretching vibration of C4-O-S and the symmetric vibration of O=S=O,respectively,which confirms its ability of self-assembly and successful synthesis[32].With that,κCaGO was shown to exhibit the necessary changes in its function group upon synthesis and was used for the demulsification tests.

The morphological properties of the formulation 3 and 4 samples were studied through a SEM.The morphology of formulation 3,as shown in Fig.S3(a)was shown to be smooth with a crimpy structure,which indicated that PEI had successfully adhered on the surface of the beads.The morphology of the beads after the coating of GO,as shown in Fig.S3(b)led to a rough and uneven surface pattern,which may be attributed by the agglomeration of RGO and PEI.

The thermal behavior of GO and formulation 1 to 4 were identified through the thermogravimetric analysis (TGA).The samples were freeze-dried before tested.For further understanding of the thermal stability of the functional groups present,the TGA analysis was conducted between the temperatures of 30–600 °C.In Fig.S4,the first stage of thermal decomposition was observed in all samples at the range of 30–160°C.The weight loss of approximately 10%–25%was because of the loss of water content.At higher temperatures,the second stage of decomposition began,which showed a more distinct variation of mass loss.This significant weight loss was ascribed by pyrolysis of the functional group-containing oxygen,which released steam,carbon monoxide,and carbon dioxide[33].Any temperature beyond 200 °C would be the final degradation step,as most of its functionalities would be lost.At this stage,graphene oxide underwent a reduction process to form thermally RGO.Furthermore,the weight loss that occurs within the heating range of 300–600°C was not so intensive as thermally reduced GO would lead to higher thermal stability[34].

Fig.2.Schematics of(a)κ-carrageenan and glutaraldehyde,(b)PEI and activated κ-carrageenan,and(c)Proposed reduced graphene oxide hydrogel composite.

On the other hand,the second stage of deposition for formulation 1–4 shows a subsequent weight loss between 25%–50%loss(after the loss of hydrated water)between 160 and 400°C which was due to the fragmentation of carbohydrate backbone and the removal of-OSO3[35].The treatment of κC with GA and PEI shows higher thermal stability,and this was due to the crosslinking that occurred.In the expense of GA and PEI,the increasing trend of weight loss could be observed.In formulation 4,the addition of GO reduced the weight loss,which showed enhanced thermal stability upon chemical modification.This improvement was attributed to the decomposition of amino groups.Subsequent temperatures after 400°C showed negligible weight losses which were presumed to be due to the elimination of oxygen functionalized groups.With these cumulative results obtained,the chemical modification of κCaGO had shown enhanced its thermal stability.

N2adsorption–desorption isotherms and pore size distribution of formulation 4 sample are displayed in Fig.3.The N2adsorption–desorption isotherms of the formulation 4 sample can be classified as type V isotherm and H1 hysteresis loops,based on the IUPAC classification.H1 hysteresis loops are often associated with porous materials exhibiting a narrow distribution of relatively uniform pores.Hence,it can be stated that PEI/RGO porous structures were formed,mainly through the crosslinking between amino groups of PEI polymers and oxygen-containing groups of RGO sheets.The mesopores diameter of reduced graphene oxide composites calculated by the BJH method varies from 1.65 to 20.7 nm,with an average size of about 5.7 nm.From the literature,it was known that the hydrothermally reduced graphene oxide composites have a surface area,<100 m2·g−1,depending on exfoliation conditions such as temperature,atmosphere,and reaction time[36].

A drop shape analysis was conducted on graphene oxide sheets to examine its hydrophilic and oleophilic properties,as shown in Fig.S5.Upon contact with the GO films,the water has a contact angle of 38°,which indicated its hydrophilic nature.Moreover,the oil contact angle upon contact with the fabricated GO sheets was 6°.After 10 s,there was no contact angle visible because the oil was fully adsorbed into the GO sheet.This adsorption mechanism also proved and agreed with the literature that GO was oleophilic in nature[37].GO shows amphiphilic properties,and when it reaches the oil/water interface to interact with the molecules of asphaltenes and/or resins,the protective stabilizing film surrounding the emulsions will be destroyed.As a result,oil droplets would start coalescing and float up,forming the oil phase.Hence,this application of GO as an outer layer of the hydrogel for oil adsorbent could be made.

3.4.Demulsification performance of κCaGO

The demulsification for crude oil in water emulsion was studied through a bottle test technique.The performance study of κCaGO was conducted on three parameters–salinity,pH,and dosage of κCaGO.With regard to Fig.S6,it was observed that the initial crude oil-inwater emulsion had a yellowish cloudy color while the separated phase was seen to be of a pale cloudy color.Upon the addition of κCaGO,the emulsion mixture was agitated and allowed to settle for 10 min gravitationally.As the turbidity of the emulsion was observed to be reduced,it could imply that the stability of the crude oil was destroyed,indicating successful demulsification.It could also be observed that the added κCaGO hydrogel composite was observed to be floating in the oil phase.This shows that the surrounding layer of GO on the κCaGO was reduced during the synthesis with the amine group of PEI due to its oleophilic nature.

3.4.1.Factors affecting the demulsification efficiency

Several parameters,such as the effects of salinity,the dosage of κCaGO,and pH,were studied on the demulsification efficiency.Researchers had shown that the salt content in an aqueous medium would affect the interfacial properties of the emulsion and influence its stability[38,39].Hence,the crude oil emulsions were prepared for three scenarios which were for low salinity,produced water and high salinity with the brine water salinity of 5000 mg·L−1,40,000 mg·L−1,and 100,000 mg·L−1,respectively to study its effects and influences on the demulsification efficiency of κCaGO.It was determined that the efficiency of the demulsifier increased together with the rise of salinity in the brine water,as shown in Fig.S7 (a).The demulsification performance at low salinity brine was at 21.09%.As salinity increased,the efficiency of the separation increased correspondingly at 37.26%and 54.66%,respectively.A similar trend was observed in Fig.S7(b).With an increased dosage of κCaGO of 1 mg·ml−1,5 mg·ml−1and 10 mg·ml−1,its demulsification efficiency increased at 60.06%,65.97%and 76.77%respectively.These results indicated that salinity and dosage of κCaGO possessed a significant impact on the performance of demulsification.The possible reason might be that the interfacial tension properties of the emulsion were altered due to its increase in salinity.

Fig.3.Nitrogen adsorption–desorption isotherms of the formulation 4(κC/GA/PEI/RGO).The inset shows their BJH pore-size distribution curve.

Next,it was observed in Fig.S7(c)that the efficiency of the oil separation in the acidic and neutral was at 80.52%and 82.16%efficiency.However,when a strong base with a pH value of 11 was introduced,the demulsification efficiency decreased to 65.3%.The decrease in performance could be attributed to the fact that the hydroxyl and carboxyl groups of the surrounding RGO sheets could be deprotonated at high alkalinity,which led to the electrostatic repulsion of the oil droplets and κCaGO.Therefore,the optimal condition of the usage of κCaGO would be when the emulsions were subjected to the high salinity and neutral pH condition.In addition,the higher the dosage,the faster the demulsification would be as more contact area of κCaGO would be breaking the emulsion apart.Hence,these results indicated that κCaGO could be used as an efficient demulsifier when these optimal conditions are met.

3.4.2.Reusability study

The ability to reuse a demulsifying agent plays a major role in any practical oil–water separation process.The reusability of a sample will help to reduce the cost of production and ensure that the demulsifier is used to its maximum potential.The synthesized κCaGO hydrogel composite was tested 6 times under optimum condition to examine on its demulsification efficiency.It was observed that the demulsification efficiency of the oil in the separated water phase in the initial 2 cycles was at around 68%,as shown in Fig.S8.Subsequent cycles showed a more prominent decrease in oil and water separation,but they still suggested good reusability as the separation performance dropped at a low rate of approximately 2%to 4%with each successive cycle.The degradation was probably due to the constant agitation throughout each demulsification test,which could cause the structural integrity failure of the hydrogel.Apart from that,the decrease of the demulsification efficiency was mainly caused by natural surfactants such as asphaltenes and resins present in the crude oil that had attached to the RGO sheet surrounding the κCaGO.This occurrence would cause difficulty in removing the asphaltene and resin layer,especially after multiple usages.Despite the degradation,it was still a relatively good demulsifier and could be recyclable for more than 6 times as long as its structural integrity of the κCaGO hydrogel beads was maintained.

3.4.3.Demulsification mechanism

In order to understand the demulsification mechanism driven by κCaGO on the crude oil emulsion,the morphology of the emulsion was studied before and after its demulsification process through an optical microscope.Generally,the oil droplets present in a stable emulsion had a size of less than 10 μm diameter[40].Fig.S9(a)shows the stable emulsions produced were of similar size when compared to the literature.Fig.S9(b)shows the separated water phase after the addition of the demulsifier.It could be observed that very minute oil droplets with small scale oil flocculation could be seen after the settling of 15 min.This also indicated a decrease in oil concentration when compared to before demulsification.Fig.S9(c)showed the separated oil phase after thorough shaking with κCaGO and was observed to have more significant oil concentration and aggregated oil floccules present.

The stabilization of oil-in-water emulsion is generally achieved by a rigid asphaltenes film at the oil and water interphase [41].GO sheets have a large area of π-conjugated hydrocarbons on its basal planes,which allows the adsorption of the rigid asphaltene films surrounding the oil–water interface[40].As shown in Fig.S10,the immobilized RGO sheets surrounding the κCaGO beads will then adsorb the asphaltene molecule to its π-conjugated hydrocarbons.Hence,the protective film encapsulating the oil droplets will be destroyed,which leads to the coalescence of oil droplets and results in an effective oil and water separation.

Recent studies had shown that GO had been used for the demulsification of crude oil in water.Table S2 shows the findings and results of the demulsification efficacy of recent studies in comparison to the efficiency of κCaGO beads used in this study.Firstly,GO in the form of dispersion was used as a demulsifier and was shown to have poor demulsification efficiency of up to only 7.04%after 15 min[16].When GO is freeze-dried,that form allowed the efficiency to improve to 38.6%.Following that,amine-modified graphene oxide had an efficiency of 62.25%,which was a great improvement compared to its original two counterparts[16].Another study showed that when RGO and TiO2were modified into an aqueous solution,it was able to improve its demulsification efficiency of up to 79.1%[17].However,the results were obtained after 30 min of settling time,which was double the time taken compared to the initial few tests.In this study,κCaGO beads could reach up to 82.16% efficiency in bottle suspension demulsification test despite only 10 min in settling time.However,researchers had used GO in the form of a membrane or column method to demulsify oil and water,which allowed them to improve its efficiency further to>95%[16,42].Regardless of that,κCaGO had shown to exhibit high demulsification performance despite only 10 min of settling time when compared to similar bottle test demulsification.

4.Conclusions

In this work,a new demulsifying agent,fabricated κCaGO,was successfully synthesized through the crosslinking method,with 77%of immobilization yields.The stability of κCaGO against pH,salinity,and leaching were able to demulsify emulsified crude oil of up to 82.16%efficiency.It was also shown to have good reusability of up to 6 times with a low degradation rate after each successive use.Therefore,these performances indicated that the κCaGO could potentially be used as a substitute adsorbent material for the effective demulsification performance of crude oil in water emulsion.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to acknowledge for the financial supports given by Fundamental Research Grant Scheme (FRGS/1/2019/TK02/CURTIN/03/2)from Ministry of Higher Education(MOHE),Malaysia.

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cjche.2020.08.027.