Copyright WILEY-VCH Verlag GmbH & Co. KGaA, 69469 Weinheim, Germany, 2018.
Supporting Information for Adv. Sustainable Syst., DOI: 10.1002/adsu.201800024
Sustainable Polysulfides for Oil Spill Remediation: Repurposing Industrial Waste for Environmental Benefit Max J. H. Worthington, Cameron J. Shearer, Louisa J. Esdaile, Jonathan A. Campbell, Christopher T. Gibson, Stephanie K. Legg, Yanting Yin, Nicholas A. Lundquist, Jason R. Gascooke, Inês S. Albuquerque, Joseph G. Shapter, Gunther G. Andersson, David A. Lewis, Gonçalo J. L. Bernardes, and Justin M. Chalker*
Supplementary Information
Sustainable Polysulfides for Oil Spill Remediation: Repurposing Industrial Waste for Environmental Benefit Authors: Max J. H. Worthington,1 Cameron J. Shearer,1 Louisa J. Esdaile,1 Jonathan A. Campbell,1 Christopher T. Gibson,1 Stephanie K. Legg,1 Yanting Yin,1 Nicholas A. Lundquist,1 Jason R. Gascooke,1 Inês S. Albuquerque,2 Joseph G. Shapter,1,3 Gunther G. Andersson,1 David A. Lewis,1 Gonçalo J. L. Bernardes2,4 and Justin M. Chalker*1 To whom correspondence should be addressed:
[email protected]
Affiliations: 1) Centre for NanoScale Science and Technology; College of Science and Engineering; Flinders University; Bedford Park, South Australia (Australia)
2) Instituto de Medicina Molecular; Faculdade de Medicina da Universidade de Lisboa, Lisboa (Portugal)
3) Australian Institute of Bioengineering and Nanotechnology (AIBN), University of Queensland, St. Lucia, Queensland, Australia
4) Department of Chemistry; University of Cambridge; Cambridge (United Kingdom)
S1
Table of Contents General experimental considerations
S3
Notes on materials
S3
Up-scaled synthesis of low-density polysulfide (2.5 kg reaction mixture)
S5
Characterisation of used cooking oil used in the synthesis of the low-density polysulfide
S8
SEM analysis of low-density polysulfide
S10
NMR characterisation of low-density polysulfide
S11
Density measurements of polysulfide
S13
Surface area calculation for polysulfide
S13
Simultaneous thermal analysis (STA) of low-density polysulfide
S14
Angle-resolved X-ray Photoelectron Spectroscopy (XPS) of porous polysulfide
S15
Mechanical analysis of the low-density polysulfide
S17
Water and oil contact angle measurements
S21
Crude oil capacity of porous polysulfide
S24
SEM analysis of porous polysulfide before and after oil sorption and recovery
S25
IR analysis of low-density polysulfide before and after oil sorption and recovery
S27
Raman analysis of low-density polysulfide before and after oil sorption and recovery
S28
Re-use and recovery of low-density polysulfide and motor oil
S29
Crude oil sorption of low-density polysulfide, high density polysulfide, and elemental sulfur
S30
Large-scale oil-water separations
S31
Removal of crude oil from seawater using a continuous process
S33
References
S35
S2
General experimental considerations IR Spectroscopy: Infrared (IR) spectra were recorded on a Fourier Transform spectrophotometer using the ATR method. Absorption maxima (υmax) are reported in wavenumbers (cm-1). NMR Spectroscopy: Proton nuclear magnetic resonance (1H NMR) spectra were recorded on a 600 MHz spectrometer. All chemical shifts are quoted on the δ scale in ppm using residual solvent as the internal standard (1H NMR: CDCl3 δ = 7.26 and pyridine-d5 δ = 8.74). GC-MS: Gas chromatography-mass spectrometry (GC-MS) was carried out on a Varian CP-3800 using a Phenomonex Zebron ZB5MS, 5 %-phenyl-Arylene-95 %-dimethylpolysiloxane column (30 m long × 25 mm film thickness × 0.25 mm ID). The injection temperature was set at 220 ºC, the column temperature at 190 ºC, and the gas flow rate 1.2 mL/min. Electron ionization was used to obtained nominal masses. Raman Spectroscopy and Microscopy: Raman data was obtained using an XplorRA Horiba Scientific Confocal Raman microscope. Spectra were acquired using a 50X objective (numerical aperture 0.55) at an excitation wavelength of 638 nm. Typical integrations times for the spectra were 20 to 60 s and averaged from 1 to 3 repetitions. SEM and EDS: Scanning Electron Microscopy (SEM) images were obtained using an FEI F50 Inspect system, while corresponding EDS spectra were obtained using an EDAX Octane Pro detector. Thermogravimetric Analysis (TGA): Simultaneous Thermal Analysis (STA) was carried out on a Perkin Elmer STA8000 simultaneous thermal analyzer (STA). A sample size between 11 and 15 mg was used in each run. The furnace was purged at 20 mL/min with nitrogen, and equilibrated for 1 minute at 30 ºC before each run. Heating was carried out up to 700 ºC using a 20 ºC/min heating rate. The temperature was held isothermally at 700 ºC at the end of each experiment to oxidize remaining organic matter. Angle-resolved X-ray Photoelectron Spectroscopy: The SPECS apparatus was operated at ultra-high vacuum (UHV) with a base pressure of low 10-10 mbar. A non-monochromatic X-ray source (12kV-200 W) with a Mg anode was used for the measurements and both survey and high-resolution scans were operated at a pass energy of 40eV and 10eV, respectively. The XPS can be operated across a range of different intersection angles relative to the sample surface, with 0º as the angle when electron path is perpendicular to the sample surface. In this work, the normal angle processed is 0º while the measuring depth has been maximized. With the increase of intersection angle, the free path of electrons from deep bulk is not equivalent of escaping from the sample. Thus the measuring depth is diminished. Neutral Impact Collision Ion Scattering Spectroscopy (NICISS): Ion scattering was carried in the same analysis chamber used for XPS. The sample stage was simply rotated by 45º and directed to the ion source (SPECS) with an extracting energy of 3keV. The spectra were synchronously recorded by a time of flight (TOF) detector across a range up to 4500 channels, which is set up with an angle of 15º between the ion source. After which, the channels have been converted into flight time (up to 7 µs) for the x-scale.
Material information Porous canola oil polysulfide polymer: the porous canola oil polysulfide was prepared using the saltinclusion method described previously for mercury remediation applications (Fig. S1).1 For experiments where fryer oil was used instead of pristine canola oil, this information is noted. Low-density polysulfide polymer: The low-density canola oil polysulfide was prepared using the method described on page S5. The material is identical to the porous polysulfide however it is milled to small particles that have broken where the pores would otherwise form (Fig. S1). This material, when milled, no longer contains pores so it is simply referred to as a “low-density polysulfide.” For experiments where waste fryer oil was used instead of pristine canola oil, this information is noted.
S3
Non-porous polysulfide: A non-porous, high-density version of the canola oil polysulfide was prepared as previously described for applications in mercury capture.1 The polymerisation is very similar to the lowdensity and porous polysulfide, except that the non-porous high-density version is prepared without the use of the NaCl porogen. Fryer oil: Waste mixed-vegetable cooking oil kindly donated by McHugh’s Café, Flinders University SA. Australian Crude Oil: A thin crude oil from from the Nockatunga Oil Field (Queensland, Australia) was donated by the Australian School of Petroleum at Adelaide University Texan Crude Oil: A viscous crude oil was sourced directly from wellhead (purchased from West Texas Intermediate crude oil, Texas Raw Crude International) Motor Oil: Castrol motor oil (10W-30) A)
15 wt% Sulfur
15 wt% Canola Oil
1. Heat at 180 °C for < 1 hr
70 wt% NaCl
2. Cut product into coarse particles
1. Fine Milling
Wash
2. Wash
Low-density polysulfide
B)
Polymer-salt composite
Porous polysulfide
Fig S1: Overview of synthesis procedure for the porous polysulfide. An equal mass of sulfur and canola oil are heated with stirring to 180 °C. Sodium chloride (70% NaCl of the final total mass of the reaction mixture) is O reaction appears to form one phase. The 180 mixture °C O 20 minutes. After addedO after the vitrifies in approximately S S O (S) o < 1 hr cooling the material, it is washed with water to Sremove the salt, leaving a porous block of material. Before (S) S O (S) O p O theOpolymer-salt composite + washing, can S also beScut into a desired shape, such as a OcubeO(Fig. S2). Washing m 2.5 kg provides the porous polymer cube (see Fig.SS2S right). If the polymer-salt composite is ground finely before(S) reaction O (S) n O washing, it breaks at the salt-polymer interface where mixture pores would otherwise form (Fig. S2 left). The polymer q can be dried under vacuum, in open air, or under a stream of warm air (40 ºC). The washing and drying steps(S)r 87% alkene consumed, >98% can be repeated to ensure all salt is removed (as determined by mass balance). A detailed protocol foryield largescale preparation is provided on page S5-S7.
C)
2 cm Fig S2: Left: low-density polysulfide (0.5-2.5 mm particles). Right: Cube of porous polysulfide, approximately 5.0 mm across on each side.
S4
Up-scaled synthesis of low-density polysulfide (2.5 kg reaction mixture) The reaction apparatus was assembled in a fume hood as follows: an overhead mechanical stirrer (Heidolph Hei-TORQUE 200) was secured on an H-frame stand and equipped with a stainless steel impeller (15 cm square blade, Fig. S3). A stainless steel reaction vessel (4.7 L, 20 cm diameter) was placed on a hotplate equipped with temperature probe. The reactor handles were further secured to the H-frame stand with cable ties. Because the reaction vessel is magnetised, the reactor is further secured by attraction to the magnetic hotplate. The impeller blade was positioned several millimetres from the bottom of the reaction vessel and the temperature probe and a large plastic funnel were secured so that they would not come into contact with the rotating impeller. An image of the setup is shown in Fig. S3:
Fig. S3: Reactor apparatus (left) and impeller (right) used in large-scale synthesis of the polymer Canola oil (375.0 g, either pristine food grade or recycled used cooking oil) was added to the reaction vessel. The overhead stirrer was set to 90 rpm, and the oil was heated to 170 °C, with the temperature of the oil monitored and controlled directly with a temperature probe. Sulfur (375.0 g) was then added through the funnel at a rate such that the internal temperature did not fall below 155 °C. The addition of sulfur was carried out over approximately 5-10 min. The reaction initially appears as two transparent liquid phases: the molten sulfur and sulfur pre-polymers from ring-opening polymerisation appear as an orange or red bottom layer and the canola oil forms a light yellow top layer. At this scale, the two phases begin to react and form an opaque mixture over the duration of the sulfur addition. Once the reaction mixture appears opaque and two distinct layers are not visible, the sodium chloride porogen was added. Accordingly, sodium chloride (1750 g, finely ground in a blender) was added through the funnel at a rate such that the internal temperature did not drop below 155 ºC. Upon commencing the addition of the sodium chloride, the reaction temperature was set to 180 °C to compensate for the internal temperature drop. The full addition of sodium chloride was carried out over 15-20 min.
S5
Upon completion of the salt addition, the reaction mixture was typically an orange, opaque and relatively freeflowing slurry. Upon continued heating at 180 °C, the mixture thickens and darkens to a brown colour. The reaction was stopped when the viscosity increases to a point at which the overhead stirrer registers a torque of 40 N•cm. This change typically occurs 10-15 minutes after the addition of the sodium chloride is complete. At this stage of the reaction, some gas may be evolved (H2S) so operation in a fume hood is essential. Overheating or prolonged heating at 180 ºC also leads to additional gas evolution, so the reaction was shut down immediately when the torque of the stirrer was 40 N•cm. To stop the reaction, the stirrer and the hotplate are turned off at the power source, the cable ties are cut, and the hot plate is removed and the reaction vessel is placed on a trivet to prevent further heating. The polymer (a soft rubber) is friable, allowing straightforward removal of the impeller with a spatula. To remove the polymer from the reaction vessel, it was broken into large chunks with a large spatula or paint scraper (Fig. S4). The polymer was then processed with a mechanical grinder to provide particles between 0.5 and 3 mm. Typically >2.48 kg of the polymer salt composite was isolated at this stage (Fig. S5).
Fig. S4: The canola oil polysulfide and salt composite, after breaking down into large pieces with a spatula or paint scraper.
Fig. S5: The canola oil polysulfide salt composite is a friable rubber, easily processed into particles using a mechanical grinder.
S6
To remove the sodium chloride porogen, the polymer was washed repeatedly with DI water. In a representative procedure, the polymer (2.5 kg of the polymer salt composite) was added to a 20 L bucket and DI water was added until the total volume was 17 L. The mixture was stirred using an overhead stirrer (200 rpm, 30 min). The polymer was then isolated by filtration through a sieve (0.5 mm cut-off) and washed three more times in a similar manner. After the final wash, the polymer was filtered through a sieve (0.5 mm) and pressed with a piece of flat plastic to squeeze out excess water. The polymer was then dried in the sieve by passing warm air through the material (5-24 hours, 18 - 42 °C) using a space heater. A final drying step was carried out by placing the polymer in a plastic tray in a fume hood until the mass of the polymer was constant (1-3 days). The final mass of the product varies with water content, which can be up to 2% by mass). Typically 750 g to 768 g of the final polymer are obtained (Fig. S6).
Fig. S6: Washed and dried porous canola oil polysulfide, prepared on a 750 g scale.
S7
Characterisation of used cooking oil used in the synthesis of the low-density polysulfide Waste vegetable oil (1.00 g) was mixed with methanol (100 mL) in a 250 mL round bottom flask and cooled to 0 °C. Sodium methoxide (100 mg) was then added to the stirred mixture. The reaction mixture was stoppered and stirred vigorously at room temperature for 24 hours. Vigorous stirring is important to ensure effective mixing of the two phases present at the start of the reaction. After 24 hours, the reaction was cooled to 0 °C and quenched with 0.1 M HCl (10 mL). The mixture was transferred to a separatory funnel and then diluted with ethyl acetate (100 mL) and water (150 mL). The organic layer was isolated and then washed with water (3 x 50 mL) and brine (3 x 50 mL) before drying (sodium sulfate), filtering and concentrating under reduced pressure. Analysis by 1H NMR and GC-MS was carried out as previously reported1 and revealed clean conversion to the fatty acid methyl esters. The 1H NMR spectrum of the fatty acid methyl esters is shown in Fig. S7. The identity and distribution of the methyl esters was determined by GC-MS (Fig S8 and S9), indicating the major components were methyl oleate and methyl linoleate. Yield for fatty acid methyl esters from 1.00 g vegetable oil: 970 mg. (600 MHz, CDCl3): δ = 0.87-0.90 (m, overlapping t, CH2CH3), 1.25-1.35 (m, -(CH2)-, non-allylic/non-alpha), 1.60 (m, COCH2CH2) 1.99-2.06 (m, CH2CH=CHCH2) 2.292.31 (t, J = 7.6, COCH2), 2.77 (t, J = 6.8, CH=CH-CH2-CH=CH), 3.65 (s, CO2CH3), 5.32-5.36 (m, CH=CH). Note the integration reflects a mixture of fatty acid methyl esters. O H 3CO a
c b
O
f d
e
CH 3 g
e
H 3CO a
d
f
c
e b
e
d
d
CH 3 g
h f
methyl oleate
methyl linoleate
a d
g
b e f
h
c
Fig. S7: 1H NMR spectrum of the methyl esters derived from waste cooking oil used in the polymer synthesis.
S8
GC/MS of Used Cooking Oil Transesterification Products 4500000
Oleic
4000000
Linoleic
3500000 Counts
3000000 2500000 2000000
Solvent
1500000
Palmitic
1000000
Stearic
500000 0 1
3
5
7
9
11
13
15
17
19
Time (min)
Fig. S8: GC-MS trace of the fatty acid methyl esters derived from the waste cooking oil used in the polymer synthesis. The major fatty acid components were identified as oleate (53%), methyl linoleic (34%), palmitic (10 %), and stearic acids (3%) with other unidentified materials comprising less than 0.2% of the non-solvent peaks.
Products of transesterification of vegetable oils (fatty acids as methyl esters) 100%
100%
90%
90%
80%
Linoleic
70% 60% 50% 40% 30% 20%
Oleic Palmitic Stearic Unidentified
80% 70%
Polyunstaturated
60%
Monounsaturated
50% 40% 30%
Saturated Unidentified
20%
10%
10%
0%
0%
Fig. S9: Distribution of fatty acids that make up the triglyceride in the waste cooking oil used in the polymer synthesis.
S9
SEM analysis of low-density polysulfide Samples were loaded onto aluminium SEM stubs with adhesive carbon tape and sputter coated with platinum, achieving a surface coating of 5.01 nm Pt. Stubs were then loaded on to the sample stage of an FEI Inspect 50 SEM with EDX analyser. Low-density canola oil polysulfide, after milling and washing.
Fig. S10: Left: a single particle of low-density polysulfide; right: the same particle, at greater magnification. Fryer Oil Polysulfide
Fig. S11: Left: SEM image of low-density polysulfide prepared from used cooking oil; right: the same location, at greater magnification.
The micrographs reveal that the surfaces of these polysulfide particles are textured, but not porous. We attribute this outcome to the milling process that fractures the polysulfide-salt composite around the salt crystals.
S10
NMR analysis of low-density polysulfide prepared from pristine canola oil h h e c/d a
O
O
(S)
e
m
(S) n
c/d
O
O
O
O e
h (S) p
f b
f b
87% alkene consumed, >98% yield
(S) o
(S)
q
(S)r
HOD
pyridine-d5
g h e a b
c/d
f
Fig. S12: 1H NMR spectrum for low-density polysulfide prepared from pristine canola oil (600 MHz, pyridine-d5): δ = 0.90 (CH3), 1.31, 1.71 (signals “g” contain CHS and non-alpha CH2), 2.08-2.16 (HC=CHCH2), 2.44-2.51 (CH2C=O), 4.52, 4.68 (CH2 glycerol), 5.54 (unreacted HC=CH), 5.73 (CH glycerol). The broad peaks are consistent with the polymeric structure. Assignments were determined by COSY NMR and chemical shift. Note that the structure shown is only approximate as the triglyceride is composed of a mixture of fatty acids.
Fig. S13: 1H NMR spectrum for pristine canola oil (600 MHz, CDCl3). The integration for the olefin protons and the CH3 endgroups are shown. The ratio of the integration of the alkene protons to the CH3 endgroups
S11
(1.00 : 1.26) changes to 1.00 to 9.61 over the polymerisation. This indicates 87% of the alkenes were consumed in the polymerisation. NMR analysis of low-density polysulfide prepared from used cooking oil
pyridine-d5
g HOD e a
b
c/d
h f
Fig. S14: 1H NMR spectrum for low-density polysulfide prepared from used cooking oil (600 MHz, pyridined5): δ = 0.91 (CH3), 1.31, 1.71 (signals “g” contain CHS and non-alpha CH2), 2.09-2.14 (HC=CHCH2), 2.432.50 (CH2C=O), 4.52, 4.68 (CH2 glycerol), 5.54 (unreacted HC=CH), 5.73 (CH glycerol). The broad peaks are consistent with the polymeric structure. Assignments were determined by COSY NMR and chemical shift.
Fig. S15: 1H NMR spectrum for used cooking oil (600 MHz, pyridine-d5). The integration for the olefin protons and the CH3 endgroups are shown. The ratio of the integration of the alkene protons to the CH3 endgroups (1.00 : 1.20) changes to 1.00 to 9.02 over the polymerisation. This indicates 87% of the alkenes were consumed in the polymerisation.
S12
Density measurements of porous polymer The polymer-salt composite was cut into approximately 5.0 × 5.0 × 5.0 mm cubes before removal of salt. After removing the sodium chloride porogen in water and drying to constant mass, the actual sample dimensions were measured and the mass and volume correlated to determine density. From an average of 7 samples the density was determined to be 0.52 g/cm3 (± 0.06 g/cm3).
Surface area calculation for porous polysulfide The surface area of the polymer was estimated by calculating the surface area of the sodium chloride porogen which templates the polymer formation. Sodium chloride crystals, prepared for use in polymer synthesis as described on page S5, were sputter coated (Pt, 5.0 nm) and analysed by SEM (see representative micrograph below). Approximating the crystals as cubes, the average width of the NaCl crystal was 290 µm (± 62 µm, n = 38 crystals). This corresponds to an average surface area of ~5 × 10-7 m2 per crystal. To determine the number of NaCl crystals in the synthesis, the number of crystals per mass is required. For five samples containing >100 NaCl crystals were counted manually, and the average number of crystals per mg NaCl was found to be 15 (± 2). For every gram of polymer, 2.33 g of NaCl is used as a porogen, which corresponds to approximately 35815 NaCl crystals per gram polymer. The surface area of the sodium chloride is therefore 0.018 m2/g polymer or 18 m2/kg of polymer. As a first approximation, this surface area can be used as a surrogate for the polymer surface area. We note that the crystals might fracture through attrition in the synthesis, so this is considered a lower-bound estimate for surface area. The surface area of the polymer is likely somewhat higher as the average pore size on the final polymer was 119 ± 53 µm (Figure 2 in main text). If the NaCl crystals were fractured in the synthesis to an average of 119 µm, then the number of salt crystals would increase 14.5 times. The corresponding surface area would then be 0.044 m2/g polymer or 44 m2/kg polymer.
Fig. S16: Top: 5.0 mm porous polysulfide cubes. Bottom: SEM micrograph of the sodium chloride crystals used as a porogen in the synthesis.
S13
Simultaneous thermal analysis of low-density polysulfide
STA of low-‐density polysulfide from large-‐scale batch 35
1
25
0.8 0.7
20
0.6
15
0.4
0.5 0.3
10
Weight frac7on
Heat Flow (mW)
0.9
TGA DSC
30
0.2
5
0.1
0 50
150
250
350
450
550
0 650
Temperature (°C)
Fig. S17: Thermogravimetric analysis (TGA) and Differential Scanning Calorimetry (DSC) of the low-density polysulfide was prepared using the large-scale method described on page S5-S7. The first mass drop over 200 – 280 °C corresponds to the degredation of the polysulfide domain of the polymer (S-S bonds), and the second mass loss from 350 - 500 °C corresponds to degradation of the remaining organic matter in the polymer. The endotherm at 120 °C corresponds to the melting of free sulfur contained in the polymer (typically 10-15% by mass).1
STA of low-‐density polysulfide prepared on large scale using waste cooking oil 35
1 0.9
TGA DSC
25
0.8 0.7
20
0.6
15
0.4
10
0.3
0.5
Weight frac7on
Heat Flow (mW)
30
0.2
5
0.1
0 50
150
250
350
450
550
0 650
Temperature (°C)
Fig. S18: The STA trace of low-density polysulfide formed with used cooking oil. The profile is very similar to that obtained using pristine canola oil (Fig. S17).
S14
Angle-resolved X-ray Photoelectron Spectroscopy (XPS) of porous polysulfide Angle-resolved XPS can determine the elemental composition, valence states, and how they vary with depth. The porous polysulfide derived from pristine canola oil and also used oil were sliced into 1×10×2 mm3 samples and attached to Mo sample holders. The sample was then loaded into the vacuum chamber and brought down to a pressure of 10-8 mbar for 20 min before opening to the analysis chamber (10-10 mbar). In the analysis, charging occurred during XPS trials due to the relatively low conductivity of the polymer, which causes the decrease of kinetic energy of the emitting photoelectrons. Thus a charge compensation gun was applied at 1eV/1mV to neutralise the positive charge.2 The high resolution scans of C, O, S, Na and Cl of the polysulfide samples are illustrated in Fig. S19. The 0º angle is defined as when the electron path is perpendicular to the polymer surface.
Fig. S19: High-resolution XPS scans of A) Carbon. B) Oxygen. C) Sulfur. D) Sodium. E) Chlorine. (“Canola Oil Foam” refers to the porous polysulfide prepared from pristine canola oil, “Used Oil Foam” refers to the polysulfide prepared from used cooking oil. F) Fractional elemental concentration for various angles. G) Profile of carbon-carbon bond types at various measuring angles.
S15
Notes and discussion for the angle-resolved X-ray Photoelectron Spectroscopy (XPS) of the porous polysulfide. The carbon spectra were fit based on the contributions of C-C bonds (285 ± 0.1eV), C-O bonds (287.2 ± 0.1eV), and C=O bonds (289 ± 0.1eV).3 The oxygen spectra were fit based on the contributions of C=O (532.1 ± 0.1eV) and C-O (534 ± 0.1eV).4 The sulfur spectra were fit based on the contributions of S-S bonds (163.8 ± 0.1eV), noting that the lower abundance of C-S bonds are likely overlapping in this region.5 The sodium spectra were fit based on the contributions Na1s from NaCl (1072.0 ± 0.1eV) and another Na peak at 1074.2 ± 0.1eV (tentatively Na+ bound to the polymer).6 The chlorine spectra were fit based on the Cl2p3/2 contribution from NaCl at 198.8 ± 0.1eV and another bound form of chloride bound to the polymer (Cl2p3/2 at 201.0 ± 0.1eV).7 The depth profile by angle-resolved XPS (Fig. S19) indicates that residual NaCl porogen is present in the polymer. Interestingly, while the chloride concentration appears constant across the depth profile, the sodium concentration is lower on the surface, suggesting that some of the cation is bound inside the polymer. The depth profile by angle-resolved XPS also indicates that the relative amount of carbon is higher on the surface than in the bulk polymer. The same trend (a higher relative amount on carbon on the polymer surface than in the bulk material) was also observed by NICISS (see below). Neutral Impact Collision Ion Scattering Spectroscopy (NICISS)8, 9 allows profiling of the concentration of elements up to a 20 nm depth at a soft-matter surface. The ion source projects a beam of Helium ions on to a sample surface. Consequent backscattering of the projectiles occurs, which causes the energy loss of the projectiles. The regression energy of a scattered projectile can be determined by the flight velocity, which is measured by a time-of-flight (TOF) detector. Thus, the mass of the target element can be identified by determining the energy loss. A further step has been processed in this analysis, which accounts for another energy loss mechanism due to the small-angle-scattering and electronic excitations (stopping power) with the trajectory through samples. The linear scaling of this energy loss allows observing the depth that the projectiles are backscattered. Combining these two kinds of energy loss leads to both identifying the element and determining its depth in the material (This analysis for the porous canola oil polysulfide is shown in Fig. S20).
Fig. S20: Depth profile and surface composition of carbon and sulfur on the canola oil polysulfide, as determined by NICISS. There is a higher relative amount of carbon on the surface of the polymer down to a depth of 4 nm. The higher relative amount of carbon on the polymer surface may reflect its lower surface energy or loss of unreacted elemental sulfur from the surface of the polysulfide during sample preparation or analysis under vacuum.
S16
Mechanical testing of porous polysulfide For all mechanical analyses (pages S15to S18) a 5.0 × 5.0 × 5.0 mm cube of the porous polysulfide was used (cf. Fig. S2). Stress-strain curve with and without oil 0.05
0.04
Porous foam soaked in oil
Stress (MPa)
Porous foam 0.03
0.02
0.01
0.00 0
10
20
30
40
50
60
70
Strain (%)
Fig. S21: Stress-strain curve in compression (force applied at 0.5 N/min up to 4 N) for the porous polysulfide (blue) and porous polymer soaked in oil (red), measured by dynamic mechanical thermal analysis (DMTA). This result shows that the compression-force curve is similar for the polysulfide before and after oil sorption. The foam with oil requires a bit more force to compress, but the difference is small. This indicates that the oil can be easily removed from the foam by mechanical compression. At 70% strain the foam is compressed to the point it is almost flat and can be damaged if this compression is repeated.
S17
Repeated strain over 20 cycles 45
40
35
Strain (%)
30
25
20
15
10
5
0 0
5
10
15
20
25
30
35
40
Time (min)
Fig. S22: Strain of the porous polysulfide when stress of 0.5 N (applied at 0.5 N/min) and then down to zero force (applied at 0.5 N/min) is repeated over 20 cycles. This analysis shows that there is good repeatability of the strain (compression) and relaxation cycle. The polymer can be squeezed to 35% strain (0.5 N force) repeatedly. There is only a small increase in compression set (the permanent deformation that remains after the cycle). This experiment was carried out using a Dynamic Mechanical Analysis (DMA) compression method.
S18
Stress-strain and recovery (hysteresis)
0.014
0.012
0.2N 0.4N
0.010
0.6N
Stress (MPa)
0.8N 0.008
0.006
0.004
0.002
0.000 0
10
20
30
40
50
60
Strain (%)
Fig. S23: Strain of porous polysulfide when stress of 0.5N is applied at 1N/min up to 0.2, 0.4, 0.6 and 0.8 N force (with relaxation at 1N/min in between each compression step). This experiment examines the repeated application of force at increasingly higher compression on the same sample in the same experiment. This data shows that the polymer can be compressed to increasing amounts of strain, and it springs back quite well, although there is an offset to a fixed strain (deformation), increasing after each cycle.
S19
Stress relaxation
80
70
20%
60
40% 60%
Strain (%)
50
80% 40
30
20
10
0 0
1
2
3
4
5
6
Time (min)
Fig. S24: Stress relaxation after application of various strains to the porous polysulfide. This experiment shows that the foam can recover over a short time after application of a strain (or a force). The recovery is instantaneous for strains up to 40%, but at higher strain there is some compression set (permanent deformation) most likely due to damage of the foam structure. This is particularly noticeable at 80% strain. (Note that the 80% strain is the programmed target strain, but it only reaches 75% strain, at which point it is pressed flat).
S20
Water and oil contact angle measurements The contact angle measurements on the surfaces were determined with a contact angle goniometer (Sinterface PAT1). A droplet of Milli-Q water, diesel or motor oil was applied to the surface (approx. 5 µL) using a motor controlled syringe (water) or manual syringe (diesel, motor oil). WCA measurements were determined from using low bond axisymmetric drop shape analysis from the plugin DropSnake10 for ImageJ software (v1.48, NIH, USA).
Water Contact Angle
Water droplet on bar of porous polymer
Example contact angle measurement
10 µL water droplet on cube of porous polysulfide
Fig. S25: Average water contact angle was measured to be 130° ± 10.5, with a minimum observed angle of 111° and a maximum of 156° over 15 individual measurements.
Oil contact angle measurements
Pre-contact
0s
1s
2s
3s
Fig. S26: Diesel fuel was readily absorbed by the polymer within 3 seconds.
S21
Pre-contact
0s
5s
25 s
50 s
Fig. S27: Motor Oil (10W-30) sorption. Within 50 seconds a single motor oil bead was absorbed fully into the polymer (0.5 cm edge cubes). The same absorption rate was observed for multiple tests (each row of timelapse images is a different experiment).
Pre-contact
0 ms
100 ms
300 ms
600 ms
Fig. S28: Crude oil from the Nockatunga oil field (Western Queensland) was fully absorbed by the porous polysulfide within 1 second. (Each row of time-lapse images is a different experiment).
S22
Pre-contact
0s
1s
10 s
60 s
Fig. S29: Viscous crude oil from West Texas (Intermediate crude oil, Texas Raw Crude International) was fully absorbed by the porous polysulfide within 1 minute. (Each row of time-lapse images is a different experiment).
120 50 Diesel 100
Nockatunga Crude Oil
30
Motor Oil Texas Crude Oil
Contact Angle
80
40
20 10
60
0 0
0.5
1
1.5
2
40
20
0 0
5
10
15
20 25 30 Surface Dwell Time (s)
35
40
45
50
Fig. S30: Contact angle of a droplet of oil on the polymer surface as it changes over time for various oils. Values are an average of the left and right contact angles of a single droplet.
S23
Crude oil capacity of porous polysulfide The porous polysulfide polymer was cut into cubes (0.5 cm edges) and placed in a glass dish of crude oil such that the polymer was partially submerged. After enough time had passed for the oil to wick up to the surface of the polymer cubes (wicking time), they were removed from the oil and placed in a clean glass dish. Excess free flowing oil was allowed to run off of the cube. The cube was then removed and weighed to determine the mass of the oil bound to the polymer. Sorption capacities reported below are the average of 5 replicates. Liquid Nockatunga crude oil Texas crude oil Motor oil Diesel Water (control)
Density (g/mL) 0.76
Wicking time (s) 30
Capacity (mL/g) 0.95
Std Dev. 0.15
0.89 0.75 0.80 1.00
300 300 60 300
0.95 0.86 1.36 0.056
0.14 0.14 0.39 0.012
Fig. S31: Sorption capacity of porous polysulfide for oils, diesel and water. Capacities are reported as mL liquid (oil, diesel, or water) per gram porous polysulfide.
S24
SEM analysis of porous polysulfide before and after oil sorption and recovery 2.0 g porous polysulfide was incubated at room temperature in 10 mL motor oil (an excess) for 10 minutes to ensure maximum oil sorption. After removing the polymer particles form the oil with forceps, a portion of the saturated polymer was compressed to recover the absorbed oil. The polysulfide was analysed by SEM at four stages of this process: 1) the porous polysulfide, after it was saturated with motor oil (Fig. S32); 2) the porous polysulfide after compression to remove the bound oil (Fig. S33); 3) the untreated, pristine porous polysulfide before binding oil (control for comparison) (Fig. S34); 4) the pristine porous polysulfide, compressed similarly to sample 1 (negative control, no oil used for comparison) (Fig. S35).
3 mm
300 µm
Fig. S32: SEM micrographs of the porous polysulfide, saturated with motor oil. The smooth surface is the oil filling the pores (see Fig. S34 for comparison to see pores before oil is bound).
3 mm
300 µm
Fig. S33: SEM micrographs of the porous polysulfide after compressing to recover bound oil. The surface still appears as if oil is bound.
S25
500 µm
500 µm
Fig. S34: SEM micrographs of the porous polysulfide, before treating with oil. Left: outer surface. Right: cross section of the polysulfide showing pores
3 mm
300 µm
Fig. S35: SEM micrographs of the porous polysulfide (no oil treatment) after compression. The image shows how the shape of the polymer particle can be deformed but still retain a similar surface structure to the polymer before compression.
S26
IR analysis of porous polysuflide before and after oil sorption and recovery 2921
2853 2955 1462 1377
722
Motor Oil
Absorbance (arb.)
Oil-treated polymer
Oil-treated polymer, compressed
Compressed polymer 1742 2922
1160
2852
1458
966
722
1374
Porous polymer
3150
2650
2150
1650
1150
650
Wavenumbers (cm-1)
Fig. S36: IR spectra of motor oil and porous polysulfide before and after motor oil sorption and recovery. The IR spectra of oil-treated polymer and the polymer after compression to recover oil are highly similar. This is consistent with a layer of oil that remains bound to the surface of the polymer after the bulk oil is recovered.
S27
Raman analysis of porous polysulfide before and after oil sorption and recovery
1433 Motor Oil 2847
Intensity (a.u.)
Oil-treated polymer
Oil-treated polymer, compressed
Compressed polymer
463
145
Porous polymer
214
0
400
800
1200
1600
2000
2400
2843
2800
3200
Raman Shift (cm-1)
Fig. S37: Raman spectra of porous polysulfide treated with motor oil
Raman analysis provides information complementary to that of the IR analysis. Major peaks for the polysulfide occur below 500 cm-1, corresponding to sulfur S-S modes of vibration.1, 11 A broad signal also occurs at approximately 2900 cm-1, from 2820 to 2950 cm-1. This feature appears as a sharp bump at 2843 and a broad, secondary peak that stretches over to 2950 cm-1. Motor oil shows a similar peak stretching over 2800 to 2990 cm-1. The appearance is not quite the same however: after the initial peak there is not a second bump, but a slow descending slope that eventually drops off from 2930 to 2990 cm-1 (see Fig. S38). A similar signal is seen in the oil-treated polymer, indicating oil on the surface of the polymer. The oil-treated and compressed polymer also shows this slope rather than a second broad peak, suggesting a layer of motor oil is retained on the surface of the polysulfide after bulk oil recovery (this result is consistent with both SEM and IR analysis). Fig. S38 shows a magnified and normalised view of these peaks for comparison. A baseline correction was applied to these spectra using cubic spline fit and then applying a quadratic correction.
2700
2800
2900
Raman Shift
3000
3100
(cm-1)
Fig. S38: Raman signals from 2700 - 3100 cm -1, normalised and magnified from Fig. S37 (same colour coding)
S28
Re-use and recovery of porous polysulfide and motor oil To a mixture of 5.00 mL water and 1.00 mL motor oil was added 1.00 g of the porous polysulfide (2.5 – 5.0 mm diameter particles). After 5 minutes, the polymer was removed and compressed to remove the trapped oil. The recovered polymer was then re-used in the same experiment 5 times with the same outcome.
2nd use
3rd use
4th use
5th use
Fig. S39: The porous polysulfide binds to motor oil on water and aggregates. The polymer-oil aggregate can be removed directly or isolated by filtration. Compressing the oil-polymer aggregate allows the recovery of the motor oil. The polymer can be re-used (see image) and still remove the same amount of oil from water. With repeated compression to recover the oil, the porous polysulfide begins to break into smaller particles. This change in particle size does not compromise its affinity for oil, with a single 1.00 g polymer sample removing 1 mL oil 5 times in successive samples.
S29
Crude oil sorption of low-density polysulfide, high-density polysulfide, and elemental sulfur Crude oil (1.00 g, Nockatunga oil field) was layered on to deionised water (8.0 mL) in each of four vials. One of the vials was swirled gently as the low-density polysulfide was added. The polysulfide was added in several portions until no free crude oil was observed on the surface of the water. The sorption was typically complete within a few seconds. The same protocol was repeated for the non-porous polysulfide (prepared without the sodium chloride porogen)1 and also for elemental sulfur. Only 500 mg of the low-density polysulfide was required to capture the oil. When using the non-porous polysulfide, 1.19 g was required to remove the oil from the water. When using elemental sulfur, 1.23 g was required. The low-density polysulfide aggregated rapidly with the crude oil (seconds) and floated on the surface of the water. More of the non-porous polysulfide sorbent was required to bind all of the crude oil and the aggregate was denser and less buoyant than the lowdensity polysulfide. In some replicates of this experiment, the aggregate formed from the crude oil and the non-porous polysulfide sank to the bottom of the vial. Elemental sulfur also binds to the oil and forms a dense aggregate that sinks in water. This series of experiments demonstrates that 1) elemental sulfur has an affinity for crude oil, but forms a dense and low-buoyancy aggregate. 2) The non-porous canola oil polysulfide has an affinity for crude oil, but forms a dense and low-buoyancy aggregate and 3) The low-density canola oil polysulfide has a greater crude oil sorption capacity than the non-porous polysulfide and elemental sulfur. 4) The low-density polysulfide has greater buoyancy than both the non-porous polysulfide and elemental sulfur, when each are aggregated with crude oil.
A
B
C
D
Fig. S40: A) 1.00 g of Nockatunga crude oil on water. B) 500 mg of low-density polysulfide aggregated with 1.0 mL of Nockatunga crude oil. C) 1.19 g of the non-porous polysulfide aggregated with 1.0 mL of Nockatunga crude oil. D) 1.23 g of elemental sulfur aggregated with 1.0 mL of Nockatunga crude oil.
S30
Large-scale oil-water separations Two oil spill remediation tests were carried out on a larger scale. The first test was the removal of motor oil from deionised water using the porous polysulfide prepared from pristine canola oil (Fig. S41). The second test was the removal of crude oil (Intermediate crude oil, Texas Raw Crude International) from seawater using the low-density polysulfide prepared from waste cooking oil (Fig. S42). A video of the crude oil sorption is provided as Supplementary Movie S1.
Removal of motor oil from deionised water using the porous polysulfide A
B
C
D
E
F
G
H
Fig. S41: Screenshots from video of experiment. A) 500 mL of deionised water was added to a glass dish. B) 100 mL motor oil layered on top of water. C) 100.0 g porous polysulfide (made from pristine canola oil, 2.55.0 mm particle size) was added to the oil-water mixture and the system was incubated for 5 minutes. D) t = 0 for remediation procedure. E) t = 5 minutes for remediation procedure. F) The aggregate formed from the oil and polysulfide was removed with net (1 mm diameter holes in net mesh). G) The motor oil is no longer visible as free oil on the surface of the water after it is captured by the polymer. H) All polysulfide removed, with no free oil visible.
S31
Removal of crude oil from seawater using the low-density polysulfide prepared from used cooking oil A video of this experiment is provided as Supplementary Movie S1.
A
B
C
D
E
F
G
H
Fig. S42: Screenshots from video of experiment. A) 1.50 L of seawater from Brighton Beach, South Australia was added to a glass dish. B) 100 mL crude oil (Intermediate crude oil, Texas Raw Crude International) was added to the seawater. C) 100.0 g of low-density polysulfide (prepared from waste cooking oil, particle size between 0.5 and 3 mm) was added to the oil-seawater mixture. D) t = 0 for the remediation procedure. E) t = 1 minute of the remediation procedure (the oil visible in the image has permeated the polymer). F) The oilpolysulfide aggregate can be removed with a net (1 mm diameter holes in net mesh). G) The black polysulfide is saturated with oil. The brown polysulfide is not saturated. H) Seawater after oil-polysulfide removed with net.
S32
Removal of crude oil from seawater using a continuous process Filtration apparatus: The low-density polysulfide (30.0 g, synthesised from waste cooking oil) was packed into a PVC pipe (25 mm internal diameter PVC compression fitting). The polymer was enclosed using mesh (PVC coated fibreglass yarn; 0.25 mm diameter wire; 18 × 30 strands per inch) on the inflow end and cotton fabric (2 ply) at outflow end. Vinyl tubing (25 mm diameter) was affixed to both ends (all components are shown in Fig. S43 and the assembled filter is shown in Fig. S44).
Fig. S43: Components for a filtration apparatus used to separate crude oil and seawater.
Fig. S44: Assembled filtration apparatus used to separate crude oil and seawater Continuous separation of crude oil and seawater: A mixture of seawater (100.0 g, obtained from Brighton Beach, South Australia) and crude oil (10.00 g, Texas Raw Crude) was poured through the filter. The clear seawater eluted through outflow end and the crude oil was retained on the filter. A video of this experiment is available as Supplementary Movie S2. Screen shots of the experiment and images of the polymer before and after oil capture are shown in Fig. S45-S46.
S33
Fig. S45: Screen shots from a video of the continuous separation of crude oil and seawater using a filter containing the low-density polysulfide as the oil sorbent. A video of this experiment is available as Supplementary Movie S2.
Fig. S46: Left: The low density polysulfide (prepared from waste cooking oil) before the crude oil and seawater separation. Right: The porous polysulfide after oil capture. The darker material is polymer that captured the crude oil. Note that excess polymer was used in the filter.
S34
References 1.
2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
M. J. H. Worthington, R. L. Kucera, I. S. Abuquerque, C. T. Gibson, A. Sibley, A. D. Slattery, J. A. Campbell, S. F. K. Alboaiji, K. A. Muller, J. Young, N. Adamson, J. R. Gascooke, D. Jampaiah, Y. M. Sabri, S. K. Bhargava, S. J. Ippolito, D. A. Lewis, J. S. Quinton, A. V. Ellis, A. Johs, G. J. L. Bernardes and J. M. Chalker, Chem. Eur. J., 2017, 23, 16219-16230. J. B. Metson, Surf. Interface Anal., 1999, 27, 1069-1072. G. Beamson and D. Briggs, High resolution XPS of organic polymers: The Scienta ESCA 300 database, John Wiley & Sons, Chichester, 1992. G. Beamson and D. Briggs, XPS of Polymers Database, Surface Spectra, Manchester, 2nd edn., 2000. R. S. C. Smart, W. M. Skinner and A. R. Gerson, Surf. Interface Anal., 1999, 28, 101-105. K. Siegbahn, U. Gelius, H. Siegbahn and E. Olson, Phys. Scripta, 1970, 1, 272-276. A. G. Wren, R. W. Phillips and L. U. Tolentino, J. Colloid Interface Sci., 1979, 70, 544-557. C. R. a. G. G. Andersson, Rev. Sci. Instrum., 2010, 81, 113907. C. Ridings and G. G. Andersson, Nucl. Instr. Meth. Phys. Res. B, 2014, 340, 63-66. A. F.Stalder, T. Melchior, M. Müller, D. Sage, T. Blu and M. Unser, Colloids Surf. A, 2010, 364, 7281. M. P. Crockett, A. M. Evans, M. J. H. Worthington, I. S. Albuquerque, A. D. Slattery, C. T. Gibson, J. A. Campbell, D. A. Lewis, G. J. L. Bernardes and J. M. Chalker, Angew. Chem. Int. Ed., 2016, 55, 1714-1718.
S35
