3.1 HPCS fabrication and characterisation
Previous work with the HPCS material developed an optimised synthesis process which demonstrated the effectiveness of the sponge for separating oil-in-water emulsions with refined mineral oil.22 The purpose of the HPCS material is for use as a tertiary treatment solution for oil spill clean-up, this research to demonstrates and optimises the process of separating crude oil from water and attaining filtrate with a crude oil concentration less than 15 ppm.
The synthesis process was adopted from our previous work with slight modifications.22 Shown in Fig. 1 is the schematic for the synthesis of HPCS. A commercially available MF sponge was submerged in DCX solution, followed by the addition of FeCl3 to initiate the Friedel-Crafts alkylation reaction to form the hyper crosslinked polymer. Finally, a PDMS curing process was performed to stabilise the HCP on the sponge.25, 26 The optimisation was done with an aim to further decrease synthesis cost, ensuring that the process is scalable and industrially practical. The ratio of the dichloro-p-xylene (DCX) and dichloromethane (DCM) solution to MF sponge was decreased to reduce costs. Polydimethylsiloxane (PDMS), a transparent silicone-based protective coating, was applied to the sponge surface and cured as a protective layer to increase robustness of the sponge and ensure the polymer powder remained on the sponge surface. The PDMS also further contributed to the hydrophobicity of the sponge.27
The resulting sponge was a yellowish-brown colour due to the polymer coating (Fig. S1), slight degradation of the sponge occurred prior to fully coating the DCX on the surface but the MF sponge retained most of its original form. The superhydrophobicity of the HPCS was tested in previous work with contact angles determined to be above 150° with water and 0° with oil.22 Fig. 2 shows the SEM images of the pristine HCP (Fig. 2a) and the MF sponge with HCP coating (Fig. 2b-c). The HCP is in a powdery form with closely packed (Fig. 2a). While when formed in-situ on the interconnected skeleton of the sponge, the HCP particles were uniformly dispersed on the surface of the skeleton as seen in Fig. 2c-d. This allows maximising the exposure of HCP particles to oil molecules and thus increasing the absorption efficiency. Nitrogen physisorption surface area analysis was performed to determine the surface area of the HPCS material. The sorption isotherms are presented in Fig. S2 and the BET surface area was measured to be 108.75 m2/g. It should be noted that the melamine formaldehyde sponge did not contribute to the BET surface area as the macropores of the sponge do not adhere to the BET multilayer gas sorption theory. However, with the hyper-crosslinked polymer coating, the surface area for the sponge was 108.75 m2/g, this is an increase from 0 m2/g. Although the bulk HCP surface area is often higher than 1000 m2/g, the amount of HCP coating on the surface of the MF sponge is limited and so the surface area per gram of material is significantly lower. However, the porosity of the HCP is still present and is significant enough to effect hydrophobicity of the sponge. The average pore diameter of the HPCS coating was determined to be 2.66 nm with pores of diameter between 17 nm and 300 nm providing a BET surface area of 36.36 m2/g. The nanometre-scale pores drastically increase the total surface area of the sponge and should be effective for capture of the small hydrocarbons commonly found in emulsions during oil spill clean-up operations.
3.2 Adsorbed crude oil recovery
Four samples of crude oil from Canada (Hibernia, Hebron, Cold Lake and Shell ULSF) were applied to pure crude oil uptake trials. Overall, the uptake capacity of the HPCS was similar between all crude oils, despite the variation in oil viscosity. The oil which resulted in the lowest uptake capacity was the Hibernia crude oil sample at 50.14 ± 3.91 g/g and the highest uptake capacity was with the Cold Lake crude oil with 58.46 ± 5.69 g/g (Fig. 3). The results are similar despite the two oil samples being very different in terms of viscosity, 4.8 cSt for Hibernia crude and 53.0 cSt for Cold Lake crude at 50°C. The variation in uptake capacity may be due to a difference in hydrocarbon composition of the crude oils rather than viscosity. Cold lake crude oil has a higher volume percentage of aromatics than Hibernia crude oil, however, it has a lower volume percentage of naphthenes and paraffins. This likely contributes to the difference in uptake capacity due to aromatic interactions between the oils and the aromatic groups of the HCP coating.28 The Shell ultra-low sulphur fuel crude oil was by far the most viscous oil sample used during these trials but the viscosity did not seem to have a great effect on the uptake capacity of the HPCS as it achieved 50.97 ± 7.38 g/g. The lack of sensitivity of the HPCS to viscosity may be a result of the large variety of pore sizes within the sponge, ranging from large cavities in the MF sponge to the micro and nano-sized pores in the hyper-crosslinked polymer coating. Previous work with HPCS involved measuring uptake with various organic solvents where similar adsorption capacity values were reported, between 40 g/g and 80 g/g for most trialled samples.22 Overall, the adsorption capacity of the HPCS is high relative to similar materials, especially considering the low cost and ease of synthesis of the sponge. There are filters used for oil capture that have superior uptake capacity, but they frequently require very expensive and complicated synthetic processes to manufacture. Past research with a variety of pure organic solvents and oils, combined with the data collected here with crude oil mixtures, demonstrates the capability of the HPCS to adsorb a large variety of non-polar compounds with a high loading capacity. Further work to tune the pore sizes of the HPCS could result in better uptake of oils within a particular viscosity range.
3.3 Oil-in-water emulsion separation
Although the crude oil uptake capacity of HPCS was high, a more important characteristic is whether the sponge can quickly and effectively strip water of hydrocarbon contaminants. To test the effectiveness of O/W emulsion separation, experiments using 1000 ppm emulsions of medium to heavy crude oil in water were conducted to mimic conditions that would be encountered during oil spill response operations. The super-hydrophobic polymer coating of the MF sponge repels water and other polar molecules, non-polar hydrocarbons are therefore selectively captured by the HPCS as they are adsorbed within the oleophilic pores. The repelled water is pushed out of the sponge as the emulsion is pumped through the HPCS, ideally this results in a complete separation mixture of oil from water, with an oil concentration in the filtered water less than 15 ppm.
Past research with a similar material focussed on using one pure component such as n-hexane or using a sample of refined mineral oil. Crude oil is a mixture of vastly different hydrocarbons, so these trials provide valuable insight about the performance of the HPCS in realistic oil separation applications and provides data on the capture selectivity among different hydrocarbons. Trials were conducted using around 0.25 g of HPCS loaded as a filter inside a cylinder, 500 mL of 1000 ppm O/W emulsion was then pumped through the filter at a flow rate of 40 mL/min. After one cycle through the sponge the filtrate oil concentration was 5.09 ± 1.96 ppm and after 3 cycles through the sponge the filtrate oil concentration was 2.35 ± 2.03 ppm (Fig. 4.) as determined by total remaining hydrocarbon (TRH) analysis. Environmental regulations authorise that oily water must not be disposed of directly into the sea if oil concentration is above 15 ppm.29 The HPCS was able to produce filtrate as low as 2 ppm oil which is well below the 15 ppm oil threshold.
3.4 HPCS recyclability
As mentioned previously, it is hoped that the process of recycling and collecting the residual oil from the sponge can be made as simple as possible for practical applications. By applying mechanical compression as the method of removing oil and recycling the sponge, the need for washing with solvents and the subsequent distillation of these solvents is removed. This allows for a quicker, cheaper, and more environmentally safe recycling process. With repeated trials using a sample of crude oil, the sponge was recycled by mechanically compressing the sponge (Fig. 5). It was observed that after the original trial (trial 0) with a clean sponge, around 20% of the captured oil remained in the sponge following compression. This result was consistent throughout the trials indicating that without additional cleaning process the sponge will operate at 80% capacity. The oil uptake capacity in trial 1 was 81.04% of trial 0, however this uptake capacity was relatively stable during the subsequent trials, after 10 trails the sponge uptake capacity was still 94.37% of trial 1 (Fig. 6). The HPCS was robust enough to remain stable during the harsh mechanical compression process although the elastic properties of the sponge decreased over time with repeated compressions.
The HPCS demonstrated high ability and stability to capture crude oil, it is also important to analyse the long-term effectiveness of the sponge for separating O/W mixtures and provide water filtrate with an oil concentration of less than 15 ppm. A 0.25 g sponge sample was selected and used as a filter for a 500 mL 1000 ppm O/W emulsion using Hibernia crude oil. After 7 cycles through the sponge, the equivalent of around 28 cm length of filter, the filtrate was collected, and the non-polar components were manually extracted using dichloromethane. The extract was analysed with total scanning fluorescence (TSF), the excitation/emission pair used for concentration comparison was Ex 262.03 nm and Em 367.03 nm, as this was the position of the maximum fluorescence intensity peak for the 1 ppm Hibernia crude oil standard (Table S1). The first five cycles successfully removed enough oil to provide filtrate with less than 15 ppm crude oil, trial 1 had an oil concentration of 2.8 ppm and trial 2 had an oil concentration of 2.9 ppm (Fig. 7). After the fifth separation trial, the oil concentration in the filtrate was 12.8 ppm, all trials were lower than the environmental regulated 15 ppm oil concentration ceiling for sea water disposal. After 5 trials the performance of the sponge appeared to deteriorate rapidly with the filtrate ppm level rising to 46.8 ppm following the tenth trial. Performance may be improved by using a larger and longer sample of sponge, by decreasing the flow rate to increase residence time within the sponge or by increasing the number of cycles through the sponge. Another potential solution for sponge cleanliness is to perform a solvent washing process, typically seen with other sponge materials, after every 5–10 cycles to extend the life of the filter sponge. However, these initial results are encouraging for mechanical regeneration of the materials which has numerous advantages over solvent-based washing. There are some limitations with the TSF method such as the volatility of components in the analyte, recoverability during solvent extraction and instrument drift, so the concentration results were validated by total remaining hydrocarbon (TRH) analysis summarised in Table S2. Equivalent samples to trials 1 and 2 of the TSF experiments were found to have a hydrocarbon concentration of 2.35 ppm and 3.04 ppm, respectively (Table S3). These TRH values were close to the TSF derived samples which were determined to be 2.8 ppm for trial 1 and 2.9 ppm for trial 2. The correlation between TRH and TSF values may deviate further at higher oil concentrations as a quenching effect occurs during TSF analysis, but this was not investigated as the TRH was used to validate the below 3 ppm oil concentration results of the TSF experiments.
Five adsorption trials with ppm levels below the environmental standards is a promising result for the initial investigation with the HPCS sponge. The results demonstrate that although the uptake capacity of the HPCS was stable during the 10 crude oil adsorption trials, the efficiency of the oil separation from water decreased with each cycle. Future research on a larger scale could provide valuable insight about the reliability of the sponge over long periods of time and with a substantial amount of oily water.
Total scanning fluorescence (TSF) analysis was used to quantify the concentration of oil remaining in the filtrate after the separation process and to determine any change in composition of the crude oil following separation to evaluate potential selectivity in oil capture. It was observed that although the intensity of the fluorescence increased after each trial due to rising crude oil concentrations, the distribution of fluorescence on the excitation-emission matrix remained mostly the same (Fig. 8 and Fig. S3), indicating the selectivity of the sponge towards different types of oils in the crude oil mixture did not change over time. The position of the peak intensities on the filtrate fluorescence maps did not deviate much from that of the 1 ppm Hibernia crude oil standard, once again indicating the uniform capture of the large variety of oils in the crude oil mixture. Interestingly, although the position of the peak fluorescence on the 1 ppm standard map was the same as the filtrate samples, the contours around the peak of the 1 ppm standard were different. The 1 ppm standard excitation-emission matrix showed a stronger relative intensity towards lower emission wavelengths, 47.48% of the emission wavelength intensity was below or equal to 350 nm. The filtrate samples had low relative intensity below 350 nm and were instead shifted to the right of the map with significant emission fluorescence close to 450 nm. The trial 10 filtrate had 60.07% of emission wavelength intensity above 350 nm (Table S1). This may be indicative of more heavy polyaromatic hydrocarbons being left behind in the filtrate than light polyaromatic hydrocarbons, light hydrocarbons are also more volatile so may be lost during the experimental procedures.30, 31