Oscillatory Rheology and Surface Water Wave Effects on Crude Oiland Corn Oil Gels with (R)‑12-Hydroxystearic Acid as GelatorV. Ajay Mallia,† Daniel L. Blair,‡,§ and Richard G. Weiss*,†,§

†Departments of Chemistry and ‡Physics and §Institute for Soft Matter Synthesis and Metrology, Georgetown University,Washington, DC 20057-1227, United States

*S Supporting Information

ABSTRACT: The mechanical properties of films of gels composed of a crude oil or corn oil and (R)-12-hydroxystearic acid aredescribed using oscillatory rheology and water-surface waves. The integrity of these gel films are contrasted with those of neat oilfilms subjected to the same types of mechanical testing. The oil-based gels are thixotropic, and we quantify their post-recoveryyield. A simple model is proposed to describe the loss of integrity when the gels are subjected to high amplitude surface waves.Our technique provides a novel method for quantifying the role of dynamic surface perturbations on the mechanics ofviscoelastic films, providing an illustrative model system for testing coagulants and dispersants designed to mitigate oil spills.

■ INTRODUCTION

Crude oil and other chemical spills continue to produce long-term and perhaps permanent damage to the environment.1 Spillsmay occur during extraction of oil from the ground, at refineries,at storage facilities, and during transportation of refinedproducts.2

Remediation of crude oil spills is a very complex process andmay require one or more methods, such as dispersion,solidification, digestion by microorganisms or skimming.2

Although it is not possible to mitigate completely the deleteriouseffects of a spill, chemical methods, including dispersion throughthe use of polymeric surfactants3 or superabsorbent polymers,4

offer attractive options. However, the process of dispersal doesnot permanently remove the oil from the environment; for thisreason, there is increasing interest in the design and use ofmaterials that efficiently trap crude oil as a mechanism tominimize environmental impact.Gels composed of a liquid and a low molecular-mass organic

gelator (LMOG) are a class of self-assembled systems that formwhen the LMOGs aggregate via van derWaals, intermolecular H-bonding, or π−π stacking (in the case of gelator having aromaticmoieties) into rodlike objects (e.g., fibers, strands, nanotubes, ortapes).5 Some LMOGs are known to be able to solidifyselectively an oil phase from a mixture of oil and water.6

Examples of such gelators include amino acid amphiphiles,6a

closed-chain saccharides,6b and dialkanoate derivatives ofsaccharides.6d However, many factors may affect the ability ofan LMOG to coagulate and gel crude oil selectively.(R)-12-hydroxystearic acid (HSA) is known to be a very

efficient gelator of a wide variety of liquids7 and has been usedindustrially as a lubricant and to modify the properties of bulkpolymers.8 Here, we report the properties of gels employingHSAas the gelator and a crude oil or a vegetable oil as the liquidcomponent. Corn oil was selected arbitrarily to show that themethodology employed is applicable to other liquids that mightbe spilled. These studies are complemented by measurements ofthe mechanical effects on them measured by rheology and bywater-surface waves on films of the gels and the liquids in the

absence of the gelator. HSA is an efficient, low molecular-massgelator that forms helical fibrillar aggregates and gelates aromaticliquids, alkanes, carbon tetrachloride, chloroform, acetone,ethanol and silicone oil.9 In a previous report, the crystal latticeof HSA in its gelated state has been tentatively assigned to bemonoclinic, in which four HSA molecules are packed in a “head-to-head” arrangement.7b The hydroxyl groups on the chiralcenters of HSA are connected by unidirectional hydrogenbonding sequences in their 1-dimensional (1D) fibers.10 The gelmelting temperatures (Tgel) and stabilities (as indicated by theperiod between when the gels were prepared in sealed containersand when they underwent visually detected phase separationafter being left at ∼24 °C) of 2 wt % HSA in n-alkane gels areknown to increase slightly upon increasing the chain length of theliquid from hexane to decane.9 The HSA gels with both crude oiland corn oil are found to exhibit reversible mechano-responses(i.e., they are thixotropic),9 and they are the materials usedthroughout this work.

■ EXPERIMENTAL SECTION

A BK Precision Function Generator, Crown PSA-2 Self-Analyzing Amplifier, Stanford Research System model SR830DSP lock-in amplifier and shear accelerometer (PCB Piezo-tronics, model J 353B68, voltage sensitivity = 102.5 mV/g,frequency range = 1−8000 Hz, output bias level = 14.9 V) wereused to detect the oscillating frequency of waves generated by aVibration Test System VTS 100-8 shaker through the use of acustom built Teflon trough (7.5 cm × 2.5 cm × 4.5 cm; 80 cccapacity). The sinusoidal parametric driving amplitude is directlyrelated to the voltage signal of an accelerometer that is mountedparallel to the oscillation direction. Using the output voltage andthe calibration characteristics of the accelerometer, we calculate

Received: November 10, 2015Revised: January 9, 2016Accepted: January 10, 2016

Article

pubs.acs.org/IECR

© XXXX American Chemical Society A DOI: 10.1021/acs.iecr.5b04267Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

the linear amplitude of the trough over a frequency range of 15−30 Hz.Polarized optical micrographs (POMs) were recorded on a

Leitz 585 SM-LUX-POL microscope equipped with crossedpolars, a Leitz 350 heating stage, a Photometrics CCD camerainterfaced to a computer, and an Omega HH503 microprocessorthermometer connected to a J-K-T thermocouple. Samples forPOM analyses were flame-sealed in 0.4 or 0.5 mm path-length,flattened Pyrex capillary tubes (VitroCom, Inc.). Rheologicalmeasurements were made using an Anton Paar MCR 301 stress-controlled rheometer with a Peltier based temperature controllerwith parallel plates (25 mm diameter) with a nominal gap of d =0.5 mm. Data were collected and analyzed using Rheoplus/32Service V3.10 software.Materials. HSA (Arizona Chemicals) was purified as

reported (mp, 80.2−80.3 °C (lit.,11 80.5−81.0 °C)).9 Corn oil(Mazola) and crude petroleum oil (Surrogate, SO-20111116-MPDF-003, A0064A-OL-OIL, supplied by British Petroleum)were used as received. Water was purified by passing through LTTechnology CF 101 and CF102 cartridge filters, a SANITRONUV water purifier, and by reverse osmosis that was monitored byHydro and Ecodyne controllers.Preparation of HSA−Corn Oil Gels. Weighed amounts of

corn oil and 1 wt % of HSA were placed into a screw cappedsample vial and the mixture was heated at ca. 80 °C until asolution/sol was obtained. The vial was then placed directly intoan ice−water bath for 10 min and allowed to return to roomtemperature.Preparation of HSA−Crude Oil Gels. A mixture of 2, 5, or

10 wt % HSA was added to the crude oil and stirred at 45−50 °Cfor 10min. The resulting sol was left undisturbed at 25 °C for 1 h.Preparation of HSA in Crude Oil “Aged” Gels. Aliquots

(1.0 g) of 2, 5, or 10 wt % HSA in crude oil gels, prepared asabove, were placed in scintillation vials (7 mL; o.d., 17 mm; i.d.,13 mm; and height 54 mm; height of the gel = 9 mm) andweighed before and after being left open to the air for 1 week.

Deposition of 10 wt % HSA−Crude Oil Gels on Water.HSA (1 g) and crude oil (10 g) were mixed and heated to 45−50°C for 10 min to form a solution/sol. About 5 g of the sol wereplaced on ∼80 mL of an ice−water mixture within the Teflonchamber and the mixture was left there undisturbed at 24 °C for24 h (during which time the ice melted) to obtain a ∼ 0.2 cmlayer of the gel on water.

Deposition of 1 and 10 wt % HSA−Corn Oil Gels onWater. The same procedure as described above was employedon HSA (100 mg or 1 g) and corn oil (10 g) that were heatedinitially to 80 °C to form a clear solution/sol. The layerthicknesses were ∼0.2 cm.

■ RESULTS AND DISCUSSIONRheological and POM Studies of HSA in Crude Oil Gels.

In the present study, gels of 1 and 10 wt % HSA in corn oil and 2,5, and 10 wt % HSA in crude oil have been investigated in arheometer and in a wave trough over water, so that theircomportment under different forms of disturbance can becompared with each other and with the neat liquids (i.e., in theabsence of HSA).Frequency sweep studies (ω = 0.1−100 rad/s at 0.05% strain)

of the gels at 25 °C indicate that the storage modulus (G′) andloss modulus (G″) values are independent of the appliedfrequency within the linear viscoelastic regions (LVRs) (Figure 1and Figures S1−S3); the samples are true gels. The mechanicalstrength of the gels decreased upon decreasing the HSAconcentration in both oils. For example, at 1 rad/s; G′ ≈ 55and 250 kPa for 1 and 10 wt %HSA in corn oil gels (Figure 1 andFigure S1), respectively, andG′≈ 0.5, 15, and 27 kPa for 2, 5, and10 wt % HSA in crude oil gels, respectively. Figures S4−S6 showstrain sweeps for the corresponding 1 and 10 wt % HSA in cornoil gels and the 2, 5, and 10 wt % HSA in crude oil gels.Rheological properties of a gel material are independent of strainup to yield strain, and beyond yield strain the rheologicalbehavior is nonlinear. At a frequency of 1 rad/s, the yield strainfor the 1 wt %HSA in corn oil gel was at 2.6% strain, a value more

Figure 1. Log−log frequency sweeps (0.1% strain, A and B) and strain sweeps (1.0 rad/s, C andD) for a 10 wt %HSA in corn oil gel (A and C) and 10 wt% HSA in crude oil gel (B and D) at 25 °C. G′, ■; G″, red ●.

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.5b04267Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

B

than two times larger than that of the gel with 10 wt % HSA(Figure 1C and Figure S4 and S7). For HSA in crude oil gels, thelargest yield strain was observed for 2 wt % HSA; that value isnearly two times larger compared to the measured yield strainsfor gels comprising 5 and 10 wt % HSA in crude oil (Figure 1Dand Figure S6). The rheologically determined quantity tan(δ) =G″/G′ determines the relative elasticity of viscoelastic materials;tan(δ) values less (greater) than unity indicate a more solid-(liquid-) like behavior, respectively. The lower tan(δ) observedfor 1 wt % HSA in corn oil compared to that for 10 wt % HSA incorn oil is indicative of a higher relative elastic modulus for thelower HSA concentration gels (Figure S7). Interestingly, asimilar correlation was observed for HSA in crude oil gels.The POM image of a freshly prepared 10 wt % HSA in crude

oil gel (Figure 2A) shows fibillar assemblies. Because theappearance of the gel did not change markedly after being “aged”,Ostwald ripening is not important in these systems (Figure 2B).Distinct fibrillar textures were not observed for 1 and 10 wt %HSA in corn oil gels (Figures S8 and S9).The percentage weight loss from about 1 g each of 2, 5, and 10

wt % HSA in crude oil gels was measured after they were kept at25 °C for 1 week in the air (Table S1). Increasing theconcentration of HSA in a crude oil gel decreases the evaporationof volatiles from the crude oil. We hypothesize that the moreextensive self-assembled fibrillar networks (SAFINs) at higherHSA concentrations decreases the rate at which the crude oildiffuses to the air interface as a result of increased interfacialinteractions between the gelator fibers and the liquid surfaces atthe micrometer (or submicrometer) distance scales.The mechanical properties of the aged HSA in crude oil gels

were studied using rheology at 25 °C. Frequency sweeps at 0.1%strain (Figures S10−S12) indicate that, like the fresh samples, theaged ones remain gels. TheG′ values for the 10, 5, and 2 wt % gelsat 1 rad/s were ca. 240, 30, and 0.2 kPa, respectively; agingstrengthened the 10 and 5 wt % gels and weakened the 2 wt % gelfor reasons that are not clear at present. The similarity betweenthe corresponding strain sweeps of the fresh and aged HSA incrude oil gels (Figures S13−S15) indicates that evaporation ofvolatiles in the crude oil does not significantly affect the linearviscoelastic regime. This conclusion is consistent with theobservation of little change in the appearance of the fresh gel afteraging it (Figure 2).Surface Wave Dynamics of Multiphasic Fluid/Gel

Layers: Corn Oil, HSA−Corn Oil Gels, Crude Oil, andHSA-CrudeOil Gels. Fluids and even granular materials that aresinusoidally accelerated exhibit linear and nonlinear surface

waves.12 We utilize these waves to explore the stability of the thinfluid and gel films found in Figures 3 and 4. Corn and crude oil

(both in a neat state and with HSA added) are placed on thesurface of a water bath resulting in continuous thin films (h∼ 0.2cm). Through vertical sinusoidal acceleration, parametric surfacewave states are generated. The amplitude of the sinusoidalacceleration is slowly increased while maintaining a constantfrequency. Breaking amplitudes are calculated using the peakheights of the fluid at which the continuity of the corn oil or crudeoil layer was lost (Tables 1 and 2).Table 1 summarizes the breaking amplitudes and drop sizes at

different applied frequencies for the crude oil layer deposited onwater. At the breaking amplitudes, droplets similar to those forcrude oil layer were observed (Figure 4B).For both oils, the sizes of the droplets appeared to be

independent of applied frequency over the range investigated(Table 1). At an applied frequency of ∼20 Hz, the breakingamplitude of corn oil as well as crude oil deposited on water was0.04mm, but the observed droplets of crude oil were larger (0.2−0.7 cm in diameter; Table 2) compared to those from the corn oillayer (0.1−0.3 cm in diameter; Table 2). At an applied frequencyof ∼25 Hz, similar breaking amplitudes and droplet sizes wereobserved from the corn oil and crude oil layers. At this point,there are insufficient data to correlate the specific dependence of

Figure 2. POMs (taken with a full-wave plate) at 25 °C of a 10 wt % HSA in crude oil gel: (A) freshly prepared and (B) “aged” by allowing it to remainopen to the air for 1 week. Both images show SAFINs with fibrillar networks. The colors are a function of the optical alignment of the samples, and arenot physically important here. Scale bar = 20 μm.

Figure 3. A∼0.2 cm thick layer of corn oil on water (A) before applyingwave action and (B) at the breaking amplitude (0.04 mm at 20 Hz).

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.5b04267Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

C

the perturbations on the droplet sizes because the latter areexpected to change with factors such as film thickness, interfacialinteractions, and liquid viscosity.In the experiments conducted with a corn oil layer on water,

droplets of ∼0.1−0.5 cm diameter were formed at the breakingamplitudes (Figure 3B). The sizes of the droplets have beenmeasured directly from photographs similar to that in Figure 3B.We assume that all are on the surface. We found no evidence thatnew droplets were formed over the periods of the experiments(∼20 min). In essence, there was no evidence for agglomerationof smaller droplets into larger ones. Increasing the amplitudecreated inhomogeneities in the oil layer thickness which facilitatebreakage of the layer at a critical shaking amplitude (especially,where the layers are thinnest; Figure 5).Several investigations of the conditions needed to break a

layer, form stable layers, or prepare continuous jets (when oneliquid was added to a second immiscible liquid, leading to thegeneration of droplets) have been reported.13,14 For example,droplets were formed in a microfluidic device upon injection ofan immiscible 15 wt % aqueous solution of tripotassiumphosphate into capillaries containing a continuous aqueousphase of 17 wt % of aqueous polyethylene glycol.15 On the basis

of the formation of droplets at the breaking amplitudes of layers,we suggest that the dynamics of formation and stability of an oillayer on the surface of water depend mainly on the relativemagnitudes of the viscous forces and velocity at the interfacebetween the liquids.13 The externally imposed velocity field andshear rate at large distances from a drop of corn oil or crude oilwithin the vicinity of an existing drop increase interfacial tensionthat generates drops of different dimensions.13

To understand the effect of gelation on the response of the oilsto surface wave action, aliquots of the gels were deposited onwater and the amplitude of the shaking, as quantified by theaccelerometer, was slowly increased as before at a constantfrequency until the gel layers yielded. For a 1 wt % HSA in cornoil gel (Figure 6A), the amplitude of the waves at which the layer

was broken was 0.09 mm at 20 Hz (Figure 6B). The breakingamplitude was increased to 0.2 mm for the 10 wt % HSA corn oilgel (Figures S16 and S17). As described before and shown inFigure 3B, the breaking amplitude of a∼ 0.2 cm corn oil layer onwater was 0.04 mm. Similarly, the breaking amplitude of a ∼0.2cm thick layer of a 10 wt % HSA in crude oil gel on water wasfound to be 0.1 mm at 20 Hz (Figure 7). Here again, no dropletswere observed at the breaking amplitude, and the layer wasirreversibly broken into large fragments that remained on or nearthe water surface. At 20 Hz applied frequency, the breaking

Figure 4. An∼0.2 cm crude oil layer on water (A) before applying waveaction and (B) at the breaking amplitude (0.04 mm at 20 Hz).

Table 1. Summary of Breaking Amplitudes and Droplet Sizesat Different Applied Frequencies for Crude Oil Deposited onWater

applied frequency, Hz breaking amplitude, mm droplet diameters, cm

20 0.04 0.2−0.725 0.03 0.1−0.330 0.03 0.2−0.3

Table 2. Summary of Breaking Amplitudes and Droplet Sizesat Different Applied Frequencies for Corn Oil on Water

applied frequency, Hz breaking amplitude, mm droplet diameters, cm

15.4 0.06 0.1−0.515.2 0.05 0.1−0.520.3 0.04 0.1−0.325.3 0.02 0.1−0.5

Figure 5. Schematic cartoon of the shaking trough in cross section with athin corn or crude oil layer placed on the surface of water and verticallyoscillated at a fixed frequency and amplitude. The thin layers remaincontinuous until a critical amplitude, Ac (as shown in the dashed box),when the oil layer becomes unstable and breaks into droplets on thesurface of the water.

Figure 6. An ∼0.2 cm layer of a 1 wt % HSA−corn oil gel deposited onwater (A) before applying wave action and (B) at the breakingamplitude (0.09 mm and 20 Hz). Space bar = 1 cm.

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.5b04267Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

D

amplitude of the neat crude oil layer deposited on water was 0.04mm (Figure 4B).The difference between the droplets observed at the breaking

amplitudes of the neat oil layers and the solid-like chunks of thelayers from the gels is clearly a manifestation of the much greaterelasticity within the HSA gels. Upon increasing the amplitudes,the gel layers on water experience increased strain on the crests,leading to breakage of the gel (Figure 8). To understand betterhow the different perturbations affect the gels, the trough andrheology studies were compared.

From the strain sweep analysis, the storagemodulus (G′) valueof 10 wt %HSA in crude oil gel (G′= 300 kPa at LVR, 1 rad/s and25 °C; Figure 1) is about five times larger than that of the 1 wt %HSA in corn oil gel (60 kPa at LVR, 1 rad/s and 25 °C; Figure 1),whereas the yield strain of the 10 wt % HSA in crude oil gel isabout 7 times smaller than that of the 1 wt % HSA in corn oil gel.The breaking amplitude of 1 wt %HSA in the corn oil gel and the10 wt % HSA in crude oil gel were nearly same (0.1 mm at 20Hz). This observation indicates the lack of a direct relationshipbetween the G′ values of the gels and their breaking amplitudes.Although the exact height of the waves at the breaking

amplitudes has not been determined, we speculate that the gelsmay be in their linear viscoelastic regimes at the 20 Hz frequencyemployed. Additional experiments to examine this hypothesis areplanned for the future.

Thixotropic Properties of HSA in Crude Oil Gels. Thepossibility that the gels from HSA in the two oils are thixotropicwas explored. In other studies with organogels composed ofLMOGs with structures very similar to that of HSA, (R)-12-hydroxystearamide and (R)-12-hydroxy-N-(ω-hydroxyalkyl)-octadecanamides, we have observed extensive viscoelasticrecovery after the cessation of destructive strain; the gels arethixotropic.16 There, it was found that the value of the destructivestrain did not change the % of recovery. However, the breakingamplitudes of the gels may not be directly compared with thethixotropic properties described here if our assumption is correctthat destructive strains were not applied in the troughexperiments at the breaking amplitudes. In the thixotropicexperiments, rheological measurements were conducted on freshHSA in crude oil gels at 25 °C under initial conditions at whichthey were in their linear viscoelastic regimes (i.e., a strain of 0.1%and angular frequency of 1 rad/s) for 150 s to establish baselinevalues for G′ and G″. Figure S18 shows the evolution of G′ andG″ for the 10 wt % HSA in crude oil gel after applying a 30%strain, a condition that leads to loss of its viscoelasticity. Then,the original conditions were reapplied in order to monitor therecovery of the viscoelastic properties. Approximately 75% of theoriginal G′ value, from 228 000 Pa to 170 000 Pa, was recoveredeventually. After application again of the 30% destructive strain,only about 70% of the previous G′ value was recovered, andsimilar recoveries were observed in subsequent cycles (Figure 9).

To investigate the nature of the thixotropy, the rheologicaldata were analyzed using a stretched exponential model (eq 1),17

where τ represents the recovery time and m is a dimensionlessconstant. The temporal fraction of recovery can be expressed in

terms of G′ as ′ − ′′ ∞ − ′

G t GG G

( ) (0)( ) (0)

where G′(0), G′(t), and G′(∞) are

the values at times = 0, t, and ∞, respectively. This equation issimilar to the Avrami equation18,19 in its logarithmic form.

Figure 7. A ∼0.2 cm layer of a 10 wt % HSA−crude oil gel on water (A)before applying wave action and (B) at the breaking amplitude (0.1 mmat 20 Hz). Scale bar = 1 cm.

Figure 8. Schematic cartoon of the shaking trough in cross section with athin corn or crude oil gel layer placed on the surface of water andvertically oscillated at a fixed frequency and amplitude. We anticipatethat capillary waves are suppressed due to the viscoelasticity of the gellayer and that, at the critical amplitude, Ac, the gel layer will either crackor break up near the pinning boundary condition as shown in the dashedbox.

Figure 9. G′ (■) and G″ (red ●) values at 25 °C for a 10 wt % HSA incrude oil gel as a function of time and application of different strains andfrequencies. Linear viscoelastic region (LVR): γ = 0.1%, ω = 1 rad/s.Destructive strain (DS): γ = 30%,ω = 1 rad/s. Rotational strain was keptat 0% for 0.05 s before changing from the DS to LVR conditions.

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.5b04267Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

E

τ− ′ ∞ − ′′ ∞ − ′

= −⎡⎣⎢

⎤⎦⎥

G GG G

m t mln ln( ) (t)( ) (0)

ln ln(1)

From theG′ values for the first cycle in Figure S21 and eq 1, τwascalculated to be 55 ± 4 s (R2 = 0.98; Figure S19).Figure S20 shows the evolution of G′ and G″ after applying a

30% strain to the 5 wt %HSA in crude oil gel, conditions that leadto loss of its viscoelasticity. After application of the non-destructive conditions, ∼82% of the original G′ value, 14 700 Pa,was recovered. However, in subsequent destruction-recoverycycles, virtually no additional losses to the value of G′ weredetected (Figure S21). From analyses of data from the averageG′values of 4 consecutive recovery curves (Figure S21), τ wascalculated to be 53 ± 9 s (Figure S22). Figure S26 shows theevolution of G′ and G″ after applying a 30% strain to the 2 wt %HSA in crude oil gel. Only ∼44% of the original G′ value wasrecovered for this sample (from 540 to 246 Pa). From the G′values of the recovery curves (Figure S23), τ was calculated to beτ = 73 ± 40 s (R2 = 0.84); Figure S24). Similar τ values of ∼50 s,observed for the 5 and 10 wt % HSA in crude oil gels, show thatthe recovery times are not acutely dependent on theconcentration of HSA, at least within this concentration range.Considerably larger error in τ value, 73± 40 s obtained for 2 wt %HSA in crude oil gel may be due may be the mechanically weakernature of the 2 wt % HSA in crude oil gel.Also, rheological measurements were conducted on HSA in

corn oil gels at 25 °C. The initial conditions were in their linearviscoelastic regimes (i.e., a strain of 0.05% and angular frequencyof 1 rad/s) for 150 s to establish baseline values for G′ and G″.Figure S25 shows the evolution of G′ and G″ for the 10 wt %HSA in corn oil gel after cessation of the application of 100%strain, a condition that leads to loss of its viscoelasticity.Eventually, ∼75% of the original G′ value, 300 kPa, wasrecovered. After application again of the 100% destructive strain,only about 90% of the previous G′ value was recovered, andsimilar losses were observed in subsequent cycles (Figure S26).From the G′ values for the average of the first three cycles inFigure S26 and eq 1, τ was calculated to be 17 ± 4 s (R2 = 0.99,Figure S27). A somewhat greater degree of thixotropic recovery(>99% of G′) was observed for a 1 wt % HSA in corn oil gel afterthe cessation of the destructive strain (Figure S28), although its τvalue, 15± 3 s (R2 = 0.99, Figure S29), was very similar to that ofthe 10 wt % gel. We suspect that the higher degree of recovery ofthe 1 wt % gel than that of the 10 wt % gel is related to theG′ andG″ values. The less extensive SAFIN of the 1 wt % gel is weakerand, thus, more easily reconstituted.Thixotropic experiments were also conducted on the

corresponding aged gels. By contrast, only ∼25%, ∼60%, and∼25% of the initial G′ values were recovered after destruction,respectively, of the 2, 5, and 10 wt % aged gels (Figures S30−S32). These observations, that the volatile components of crudeoil are necessary to maintain viscoelasticity, are important to thedesign of systems to recover oil from spills.

■ CONCLUSIONSPossible correlations between the mechanical properties ofmolecular gels composed of crude oil or corn oil and HSA havebeen explored using oscillatory rheology and surface waves onwater. POM studies indicate that the SAFINs of HSA in corn oil(1 and 10 wt %) and 10 wt % HSA in crude oil gel have fiberlikenetworks. Frequency sweep studies of the gels at 25 °C indicatethat the G′ and G″ values are independent of the appliedfrequency at LVR. The mechanical strength of the gels decreased

upon decreasing the HSA concentration in both oils. The G′value of 10 wt %HSA in crude oil gel is five times larger than thatof the 1 wt % HSA in corn oil gel, whereas the yield strain of the10 wt % HSA in crude oil gel is about 7 times smaller than that ofthe 1 wt % HSA in corn oil gel. The sizes of the droplets thatappear at the breaking amplitudes for the thin films (0.2 cm) ofcorn oil or crude oil deposited on water were independent of theapplied frequency. Breaking amplitudes calculated for a 1 and 10wt %HSA corn oil gel were 0.09 mm and 0.2 mm, respectively, at20 Hz. Upon increasing the amplitude, the gel layers on waterexperienced increased strain on the crests, leading to breakage ofthe gels. Similarly, the breaking amplitude of a ∼0.2 cm thicklayer of a 10 wt %HSA in crude oil gel on water was 0.1 mm at 20Hz.In addition, the HSA in crude oil and HSA in corn oil gels

exhibit thixotropic properties. They recover a large part of theirviscoelasticity within seconds after the cessation of destructiveshear strains. The thixotropic recovery for the 10, 5, and 2 wt %HSA in crude oil gels after application a γ≈ 30% strain were∼75,∼82, and∼44%, respectively, based on the originalG′ value. Thethixotropic recovery for 10 wt % HSA in corn oil gel was ∼75%based on the original G′ value. Thixotropic recovery valuesdecreased to ∼25, ∼60, and ∼25%, respectively, for the aged 10,5, and 2 wt % gels. The observation that the volatile componentsof crude oil are necessary to maintain viscoelasticity is importantto the design of systems to recover oil from spills. In the future,we will adopt a room temperature gelation protocol by dissolvingHSA in small amounts of ethanol or methanol to avoid the needfor heating and cooling. We have found that this methodologydoes work with other gelators;6d it will be explored with crude oil(and vegetable oil) in future investigations.

■ ASSOCIATED CONTENT

*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.iecr.5b04267.

POM images, strain, and frequency sweep rheology curvesof HSA in corn oil and HSA in crude oil, and thixotropyplots of HSA in crude oil gels (PDF)

■ AUTHOR INFORMATION

Corresponding Author*E-mail: weissr@georgetown.edu.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

We thank the National Science Foundation (Grant CHE-1147353) and the Gulf of Mexico Research Initiative for theirsupport of this research. We also thank Mr. Franklin Feingold(REU summer student) and Mr. Bibek Samanta (Bose-Khoranasummer scholar) for performing some of the initial experiments.

■ REFERENCES(1) Rudzinski, W. E.; Aminabhavi, T. M. A Review on Extraction andIdentification of Crude Oil and Related Products Using SupercriticalFluid Technology. Energy Fuels 2000, 14, 464.(2)Nguyen, S. T.; Feng, J.; Le, N. T.; Le, A. T. T.; Hoang, N.; Tan, V. B.C.; Duong, H. M. Cellulose Aerogel from Paper Waste for Crude OilSpill Cleaning. Ind. Eng. Chem. Res. 2013, 52, 18386.

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.5b04267Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

F

(3) Al-Sabagh, A. M.; Atta, A. M. Water-Based Non-Ionic PolymericSurfactants as Oil Spill Dispersants. J. Chem. Technol. Biotechnol. 1999,74, 1075.(4) Yuan, X.; Chung, T. C. M. Novel Solution to Oil Spill Recovery:Using Thermodegradable Polyolefin Oil Superabsorbent Polymer(Oil−SAP). Energy Fuels 2012, 26, 4896.(5) (a) Pozzo, J.-L.; Clavier, G. M.; Desvergne, J.-P. Rational Design ofNew Acid-Sensitive Organogelators. J. Mater. Chem. 1998, 8, 2575.(b) Tamaru, S-i.; Nakamura, M.; Takeuchi, M.; Shinkai, S. RationalDesign of a Sugar-Appended Porphyrin Gelator that is Forced toAssemble into a One-Dimensional Aggregate. Org. Lett. 2001, 3, 3631.(c) Mieden-Gundert, G.; Klein, L.; Fischer, M.; Vogtle, F.; Heuze, K.;Pozzo, J.-L.; Vallier, M.; Fages, F. Rational Design of Low MolecularMass Organogelators: Toward a Library of Functional n-Acyl-ω-aminoAcid Derivatives. Angew. Chem., Int. Ed. 2001, 40, 3164.(6) (a) Bhattacharya, S.; Ghosh, Y. K. First Report of Phase SelectiveGelation of Oil fromOil/Water Mixtures. Possible Implications TowardContaining Oil Spills. Chem. Commun. 2001, 185. (b) Vemula, P. K.;John, G. Crops: A Green Approach Toward Self-Assembled SoftMaterials. Acc. Chem. Res. 2008, 41, 769. (c) Vemula, P. K.; Li, J.; John,G. Enzyme Catalysis: Tool to Make and Break Amygdalin Hydro-gelators fromRenewable Resources: A DeliveryModel for HydrophobicDrugs. J. Am. Chem. Soc. 2006, 128, 8932. (d) Jadhav, S. R.; Vemula, P.K.; Kumar, R.; Raghavan, S. R.; John, G. Sugar-Derived Phase-SelectiveMolecular Gelators as Model Solidifiers for Oil Spills. Angew. Chem., Int.Ed. 2010, 49, 7695.(7) (a) Terech, P.; Rodriguez, V.; Barnes, J. D.; Mckenna, G. B.Organogels and Aerogels of Racemic and Chiral 12-Hydroxyoctadeca-noic Acid. Langmuir 1994, 10, 3406. (b) Terech, P.; Pasquier, D.;Bordas, V.; Rossat, C. Rheological Properties and StructuralCorrelations in Molecular Organogels. Langmuir 2000, 16, 4485.(c) Tachibana, T.; Mori, T.; Hori, K. Chiral Mesophases of 12-Hydroxyoctadecanoic Acid in Jelly and in the Solid State. I. A New Typeof Lyotropic Mesophase in Jelly with Organic Solvents. Bull. Chem. Soc.Jpn. 1980, 53, 1714. (d) Tachibana, T.; Mori, T.; Hori, K. ChiralMesophases of 12-Hydroxyoctadecanoic Acid in Jelly and in the SolidState. II. A New Type of Mesomorphic Solid State. Bull. Chem. Soc. Jpn.1981, 54, 73. (e) Tachibana, T.; Kitazawa, S.; Takeno, H. Studies ofHelical Aggregates of Molecules. II. The Sense of Twist in the FibrousAggregates from the Alkali Metal Soaps of Optically Active 12-Hydroxystearic Acid. Bull. Chem. Soc. Jpn. 1970, 43, 2418. (f) Tachibana,T.; Kambara, H. Studies of Helical Aggregates of Molecules. I.Enantiomorphism in the Helical Aggregates of Optically Active 12-Hydroxystearic Acid and its Lithium Salt. Bull. Chem. Soc. Jpn. 1969, 42,3422. (g) Tachibana, T.; Kambara, H. Enantiomorphism in the HelicalAggregate of Lithium 12-Hydroxystearate. J. Am. Chem. Soc. 1965, 87,3015.(8) (a) Tsau, J. S.; Heller, J. P.; Pratap, G. Thermally ReversibleOrganogels of 12-Hydroxystearic Acid. Polym. Prepr. 1994, 35, 737.(b) Terech, P. Small-Angle-Scattering Study of 12-HydroxystearicPhysical Organogels and Lubricating Greases. Colloid Polym. Sci. 1991,269, 490.(9) Mallia, V. A.; George, M.; Blair, D. L.; Weiss, R. G. RobustOrganogels from Nitrogen-Containing Derivatives of (R)-12-Hydrox-ystearic Acid as Gelators: Comparisons with Gels from Stearic AcidDerivatives. Langmuir 2009, 25, 8615.(10) Abdallah, D. J.; Lu, L.; Weiss, R. G. Thermoreversible Organogelsfrom Alkane Gelators with One Heteroatom. Chem. Mater. 1999, 11,2907.(11) Serck-Hanssen, K. Absolute Configuration of Ricinoleic Acid.Chem. Ind. (London) 1958, 1554.(12) (a) Melo, F.; Umbanhowar, P. B.; Swinney, H. L. Transition toParametric Wave Patterns in a Vertically Oscillated Granular Layer.Phys. Rev. Lett. 1994, 72, 172. (b) Melo, F.; Umbanhowar, P. B.;Swinney, H. L. Hexagons, Kinks, and Disorder in Oscillated GranularLayers. Phys. Rev. Lett. 1995, 75, 3838. (c) Goodridge, C. L.; Hentschel,H.; Lathrop, D. P. Breaking Faraday Waves: Critical Slowing of DropletEjection Rates. Phys. Rev. Lett. 1999, 82, 3062. (d) Aranson, I. S.; Blair,D. L.; Kalatsky, V. A.; Crabtree, G. W.; Kwok, W.; Vinokur, V. M.; Welp,

U. Electrostatically Driven Granular Media: Phase Transitions andCoarsening. Phys. Rev. Lett. 2000, 84, 3306.(13) Stone, H. A. Dynamics of Drop Deformation and Breakup inViscous Fluids. Annu. Rev. Fluid Mech. 1994, 26, 65.(14) Purcell, E. M. Life at Low Reynolds Number. Am. J. Phys. 1977,45, 3.(15) Sauret, A.; Spandagos, C.; Shum, H. C. Fluctuation-InducedDynamics of Multiphase Liquid Jets with Ultra-Low Interfacial Tension.Lab Chip 2012, 12, 3380.(16) Mallia, V. A.; Weiss, R. G. Structural Bases for Mechano-Responsive Properties in Molecular Gels of (R)-12-Hydroxy-N-(ω-Hydroxyalkyl)octadecanamides. Rates of Formation and Responses toDestructive Strain. Soft Matter 2015, 11, 5010.(17) Barnes, H. A. Thixotropy-a Review. J. Non-Newtonian Fluid Mech.1997, 70, 1.(18) Avrami, M. Kinetics of Phase Change. I. General Theory. J. Chem.Phys. 1939, 7, 1103.(19) Avrami, M. Kinetics of Phase Change. II. Transformation-TimeRelations for Random Distribution of Nuclei. J. Chem. Phys. 1940, 8,212.

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.5b04267Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

G