Subscriber access provided by Purdue University Libraries
Article
A pH resolved view of the low salinity effect in sandstone reservoirs. Lijuan Shi, Mats H. M. Olsson, Tue Hassenkam, and Susan L. S. Stipp Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.6b00338 • Publication Date (Web): 20 Jun 2016 Downloaded from http://pubs.acs.org on June 27, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 25
Revamping Lijuan’s graphs
600
Force [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
maximum force
400 200
approach
0
adhesion force
-200 0
withdrawal
Tip-surface separation [nm]
ACS Paragon Plus Environment
800
Energy & Fuels
Chicago???
A
B
Height [nm]
2
C
D
4
E
0
0
-4 -2 -8 0.4
0.0
Height [nm]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 25
20
0.8
F
0.4
0.0 150
0.8
1.2
G
100 50
10
0 0 0.0
-50 0.4
0.8
Distance [µm]
1.2
0.0
0.5
1.0
1.5
Distance [µm]
ACS Paragon Plus Environment
2.0
Page 3 of 25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
Lijuan’s figures revamped
ASW
ASW-low
pH 4.5
pH 5.5
pH 7.5
169 pN
114 pN
116 pN
pH 4.5
pH 7.5
pH 5.5
88 pN
55 pN
64 pN
ACS Paragon Plus Environment
Energy & Fuels
Adhesion Force [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 4 of 25
300 ASW
ASW-low
200
100
0
5
6
7
5
pH
ACS Paragon Plus Environment
6
7
Page 5 of 25
Adhesion Force [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Energy & Fuels
600 ASW
ASW-low
400
200
0
5
6
7
8
5
pH
ACS Paragon Plus Environment
6
7
8
Energy & Fuels
–CH3
200 150
Adhesion [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 25
100 50 –COO(H)
400 300 200 100 pH 4.5
pH 5.5
pH 7.5
ACS Paragon Plus Environment
Page 7 of 25
Adhesion force [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Energy & Fuels
400 --CH3 tip in ASW --CH3 tip in ASW-low --COOH tip in ASW --COOH tip in ASW-low
300
200
100
0 4
5
6
pH
ACS Paragon Plus Environment
7
8
Energy & Fuels
300
60
--CH3 terminated tip --COOH terminated tip
200
Relative adhesion [%]
Adhesion difference [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 8 of 25
0
4
pH
8
100
0 4
5
6
pH
ACS Paragon Plus Environment
7
8
Page 9 of 25
50 0 F EDL (pN)
700 mM 30 mM
F EDL [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Energy & Fuels
-50
0
0.4 Ionic strength (M)
0.8
0
-50 4
5
6
pH
ACS Paragon Plus Environment
7
8
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 25
A pH resolved view of the low salinity effect in sandstone reservoirs. L. Shi*, M.H.M. Olsson*, T. Hassenkam, and S.L.S. Stipp Nano-Science Center, Department of Chemistry, University of Copenhagen, Universitetsparken 5, 2100 Copenhagen, Denmark.
Keywords: enhanced oil recovery, EOR, low salinity effect, sandstone, atomic force microscopy, surface titration.
ABSTRACT: Core plug and field tests have shown that significantly more oil can be produced from sandstone reservoirs when the water, that is injected to maintain pressure in the reservoir, has lower salinity than regular sea water. We investigated the adhesion between functional groups known to dominate in crude oil and grains from reservoir sandstone, as a function of pH and salinity. We used atomic force microscopy (AFM) in chemical force mapping (CFM) mode and we functionalized the AFM tips with -CH3, to model adhesion of hydrophobic molecules and with -COO(H), to model polar compounds. As we pH increased, adhesion force decreased. Behaviour was similar for the two types of tip but adhesion was higher for the polar tip. The average adhesion at pH 4.5 for the -COO(H) tip in artificial sea water (ASW) was close to 300 pN, and 200 pN in ASW diluted by a factor of 20. For the -CH3 tip, adhesion was ~200 pN in high salinity and 140 pN in low salinity. Regardless of the salinity, adhesion decreased as pH increased. To gain understanding about the controlling processes, we examined our data in light of DLVO theory and the Henderson-Hasselbalch expression. The commonly used form of the DLVO relationship could not account for the pH dependence. It predicts adhesion to be independent of pH in high salinity solution and strongly dependent on pH in low salinity, opposite to the experimental evidence. Although DLVO theory cannot be expected to quantitatively represent real systems, it is unlikely that DLVO would predict reversed behaviour. Our results indicate that the extended double layer, which is one of the prevailing explanations for the low salinity effect, is not the full explanation. Decrease in surface charge density at low salinity plays an important role. The change in adhesion as pH increases above 5.5, which is the typical pH of sandstone reservoirs, to the neutral range could enhance oil recovery. This is probably a factor in the effectiveness of alkaline flooding and might be an inherent factor in the production that results from flooding with sea water, which is typically pH 8.3. However, the buffering effect in sandstone reservoirs, which maintains pH at ~5.5, suggests that benefits from controlling pH for EOR could be marginal.
ACS Paragon Plus Environment
Page 11 of 25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
Introduction Water flooding is by far the most common method for improving oil recovery (IOR) and has been used since the early days of oil production. Water is injected into the reservoir to maintain pressure, producing oil. It is usually taken from any convenient source, such as a nearby river, the ocean or another rock formation. Often, some of the water produced with the oil is reinjected. Core plug and field tests have demonstrated that from 2 to 40% more oil can be produced from sandstone reservoirs if low salinity water (18 MΩ/cm) and chemicals that were reagent grade or better. pH was adjusted to 4.5, 5.5, 6.5, 7.5 and 8.5 with droplets of 0.3 M HCl or 0.3 M NaOH, depending on the experiment. Table 1. Solution composition in mM Composition ASW ASW-low NaCl 478 20 KCl 12 0.5 MgCl2 57 2 CaCl2 12 0.5 Ionic strength 697 28 Salinity (g/kg) 34.5 1.5 AFM Tips The quartz grain surfaces were scanned with standard AC-mode silicon tips from Olympus (OMCLAC240) that had a resonance frequency of ∼80 kHz and a spring constant of 2 nN/nm for the tapping mode images, which show the topography of the surface. To measure the adhesion force and produce force maps, we used Olympus biolever AFM probes. These are produced with two types of cantilevers, where spring constants range from 20 to 30 pN/nm for the short and from 4 to 8 pN/nm for the long cantilevers. We used only one of them. Just before each experiment, we obtained the actual spring constant for the particular cantilever we chose for the experiment, by fitting to a thermal spectrum18. The tips were functionalized with -CH3 and -COO(H) groups as models for tiny oil droplets. The two terminations represent nonpolar and polar compounds in crude oil. The gold coated biolevers were prepared by rinsing with pure ethanol, dried with a jet of N2, ozone treated for 20 minutes and functionalised by exposure to an ethanol solution of ~5 mM HS(CH2)15CH3 or HS(CH2)10COOH for more than 24 h. This formed a self assembled monolayer that was quite robust, even during long sequences of mapping19. AFM imaging and force mapping. We used an MFP-3D AFM from Asylum Research for all of the experiments. Tapping mode images, showing the height of surface features, were obtained in air on a 3 µm × 3 µm area, with Ångström height resolution. Force maps were obtained by collecting 50 × 50 force curves over a 5 µm × 5 µm area. Each force curve was obtained by moving the tip toward the surface until it senses a force of 500 pN for 0.1 s and then the tip is retracted. Each pixel in each force map is derived from the pull off adhesion from a force curve collected at that (x,y) coordinate. Height maps can also be made using the maximum force data point. The procedure is summarized in Fig. 1. 600
Force [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
maximum force
400 200
approach
0
adhesion force -200 0
withdrawal
Tip-surface separation [nm] 800
ACS Paragon Plus Environment
4
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 25
Fig. 1. A typical force-distance curve. As the tip approaches (red), it feels the surface and is pushed into it until it registers the preset maximum force, in this case, 600 pN. As the tip retracts (blue), adhesion holds it to the surface until it snaps free. Its maximum deflection, in this case 200 pN, is the adhesion force. The small offset (∼10 pN) between the approach and withdraw curves is an artefact that results from drag of the cantilever through the liquid.
For the force mapping, we used a fluid cell with about 3 ml of solution, with a given pH and salinity. We chose an imaging area on the quartz grain and determined average adhesion over the whole area from the full data set. We obtained three to five consecutive force maps in the initial solution for each experiment, to be sure of stable conditions and to minimize uncertainties from the instrument and the mapping procedure itself. After the initial stabilization, we changed the solution to one with a specific pH and salinity. Solution exchange required care so as not to move the sample. It was important to image precisely the same area each time. Two or three force maps were then collected and average adhesion determined. After stepping through the various pH solutions, we returned to the starting solution, to test for reproducibility. The average adhesion, determined from a series of several maps made in the same solution, generally fluctuates, probably because the tip accumulates organic contamination or the organic material associated with the surface changes. Remapping adhesion force with the starting solution provides a measure of reproducibility and uncertainty. To obtain a reliable measure of pH dependent adhesion on inhomogeneous surfaces, it is essential to measure the same area on the sample through the entire pH range. In rocks, different grains and different areas on the same grain often have very different adhesion properties. This has been shown to result from nanometre scale authigenically formed clay particles and inhomogeneous accumulation of organic compounds.20-22 To be sure of having the same area in subsequent scans, we choose a location with a unique set of features, for example, areas with particles of an identifiable shape. DLVO theory DLVO is a continuum theory that describes the force between two macroscopic charged surfaces that interact through a liquid medium, such as an aqueous solution. The surfaces are attracted to each other by van der Waals and hydrophobic interactions and repel each other by electrostatic interactions in proportion to their surface charge. Surface charge gives rise to an electrostatic or electric double layer force, FEDL: =
2
+ +
,
(1)
where R represents the tip radius, λD, the Debye length of the solution, ε, the relative permittivity of water, ε0, the permittivity of a vacuum, σS and σT represent the surface charge densities of the sample and the tip and D, the distance between the tip and the sample surface23. It is assumed that the surface potentials are low (ψS, ψT < 50 mV), surface charge is independent of D, the tip radius is large (R » λD), and the tip-surface distance is large (D ≥ λD) for the expression to be valid. Surface charge density depends on solution pH and can be approximated from an equation with Henderson-Hasselbalch form: =
"#$%&$'( ! ")"#$%&$'(
,
(2)
where σmax represents the maximum surface charge density and the pKa is for the dominant titrating species of the surface. σS is negative for surfaces dominated by acid functional groups and its magnitude increases with increasing pH, and the reverse, for basic groups. We know from previous work that a significant amount of organic material is associated with the surface so we use pKa=4.7 to reflect the organic acids that we know are on the surface, rather than the pKa for pure quartz or clay minerals. Results and discussion ACS Paragon Plus Environment
5
Page 15 of 25
Topography of the sand grain surfaces We examined 5x5 µm2 areas on 16 different quartz grains and explored the surface properties in chemical force mapping experiments. Although the particles came from the same reservoir core plug, their topography and surface composition were very different. The images were consistent with images taken by other researchers on material from the same core plug and from different samples13, 16. Fig. 2 shows the features on three surfaces that represent the range that we observed. Fig. 2A to C show the physical height maps, obtained from AFM imaging in tapping mode. Fig. 2D to G show the height profiles for the cross sections indicated by the coloured arrows. The quartz facets are not always smooth crystal faces, as one might expect but are covered with nanometre scale crystals, probably of clay, and organic material. Terraces and edges are rough. A
B
Height [nm]
2
C
4
D
E
0
0
-4 -2 -8 0.0
Height [nm]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
20
0.4
0.8
F
0.4
0.0
150
0.8
1.2
G
100
10
50
0
-50
0
0.0
0.4
0.8
Distance [µm]
1.2
0.0
0.5
1.0
1.5
Distance [µm]
2.0
Fig. 2. A-C) AFM images of some representative pore surfaces and (D-G) cross sections from lines showing some representative features.
The organic material on the surface is important in this study because the interaction between the tip and surface at various pH and salinity depends predominantly on the surface charge so if the surface is covered by organic material, it would be the charge of this material that is relevant. Organic material is associated with all mineral surfaces exposed to air or water, even where no oil has been in contact24, such as suspended in an alpine stream or in a sedimentary basin. In these rocks from an oil bearing horizon, there is certainly a substantial portion of organic material that has originated from pore surface equilibration with the crude oil. The organic material is not a coating or a continuous layer. Rather, it is inhomogeneous13 in terms of composition and thickness. We can be sure that the pure quartz crystal surface, or the surfaces of the adhering clay nanoparticles, have at least one or a few molecular layers of carbon contamination. Some of the material is probably large aggregates of asphaltene molecules as well as lighter fractions with aromatic, aliphatic, and polar groups.3 The inhomogeneity contributes to the range in height, from a few nm, as in Fig. 2D and E, to more than 200 nm in Fig. 2G. Thus we expect the adhesion to also vary locally, within an AFM image, and between imaging sites on the same grains and different grains or from different cores. We see this range in adhesion, in this work and others13, 16 but we have also seen that the general adhesion properties are consistent across all of our work. Regardless of the initial adhesion force, which is a result of the interaction between the organic material on the surface and the tip, the force changes consistently when the salinity of the solution in contact changes and these changes are reversible through several cycles of solution exchange16. This provides confidence that changes in the adhesion resulting from changing pH would be observable regardless of the absolute adhesion, which 6 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 25
would depend on the character of the adsorbed organic material. Thus, although we only probe small areas of the surface, our results indicate that their behaviour is representative for the pore surfaces through the reservoir, where injected solutions interact with the oil and the pore surfaces. Our approach cannot give a complete picture of the processes inside a reservoir but we can capture the differences resulting from change of salinity and pH. Probing pore surfaces with -CH3 functionalised tips represents the interaction of the hydrophobic parts of crude oil. Adhesion is typically higher on hydrophobic surfaces. The -COO(H) tip represents a tiny oil droplet where polar groups form the interface with the aqueous solution, a model for the acidic components of oil. Adhesion is high for surfaces dominated by polar or charged areas. We relied mostly on the -CH3 tip because it is not affected by pH change itself and our experience shows that it is more stable and the results are more reproducible. Force mapping with -CH3 tips. The interaction between a surface and a -CH3 functionalized tip is shown in Fig. 3. We varied pH (4.5, 5.5, and 7.5) for artificial seawater (top) and low salinity artificial seawater (bottom) to monitor its influence on adhesion. Height maps (yellow), which show topography, are made from the maximum force point from the force-distance curves. Although the resolution is poorer in this type of image than the tapping mode images, they are quite useful because the data are taken simultaneously with the adhesion data so when we see features that remain stable, we know that we can directly compare average adhesion forces for the entire mapped area. The adhesion maps (blue) show the tip-surface interaction and its change with pH. ASW
ASW-low
pH 4.5
pH 5.5
pH 7.5
169 pN
114 pN
116 pN
pH 4.5
pH 7.5
pH 5.5
88 pN
55 pN
64 pN
Fig. 3. AFM height images (yellow) and adhesion force maps (blue) collected over a 5x5 µm2 area collected simultaneously in ASW (top) and low salinity ASW (bottom) with a -CH3 functionalized tip. The images are presented from left to right in order of time. The sequence of solutions was not the same in the two experiments. pH is recorded at the top of each set.
ACS Paragon Plus Environment
7
Page 17 of 25
For the data taken in ASW, there was very little drift so we can average over the entire mapped surface for comparing adhesion behaviour. Often, drift moves the tip with respect to the sample. This is a well known and unavoidable artefact in AFM imaging. An example is the images in the lower row, from experiments in low salinity ASW. In this case, we chose boxes where the position of the features remained constant and determined the average adhesion over a roughly 4 µm × 4 µm area (green rectangle). Average adhesion in the ASW experiments decreased from 169 pN to 114 pN when pH increased from 4.5 to 5.5. This is ~55 pN on an absolute scale and 32% on a relative scale. Further pH increase to 7.5 did not cause further change (116 pN). In low salinity ASW, initial adhesion is significantly lower, consistent with the low salinity response. When pH increased from 4.5 to 7.5, adhesion decreased from 88 pN to 55 pN, i.e. by 33 pN or 37%. Decrease of pH back to 5.5 increased adhesion again, to 65 pN, proving reproducibility. From these experiments, we conclude that adhesion decreases as pH increases, with a larger decrease from 4.5 to 5.5 than from 5.5 to 7.5. The overall adhesion for these experiments, in the range of 100 pN, is relatively low, indicating an overall polar surface. By comparing the features in the topographic image with the force maps (Fig 3), we conclude that for this imaging site, no distinct area of the surface is clearly responsible for the adhesion change, as would be the case if the effect is the result of single large organic macromolecules or nanoparticles on the quartz surface that have their own surface character. Instead, we observe a general shift toward lower adhesion. We made several other, similar adhesion experiments. Four different samples were measured in ASW and three, in low salinity ASW. The results are shown in Fig. 4. The initial adhesion force was different for each area that we measured, as is expected, but the overall trends as pH is altered are similar. On average, adhesion at pH 4.5 is significantly higher in ASW, where the median is ~200 pN, than in low salinity ASW, where the median is ~100 pN. Adhesion force decreases as pH increases and the decrease is highest when pH changes from 4.5 to 5.5. 300 ASW
Adhesion Force [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
ASW-low
200
100
0
5
6
7
5
6
7
pH
Fig. 4. Adhesion force measured with a -CH3 tip as a function of pH. Each set of dots represents an experiment with a new tip on a new surface with a new set of solutions. Error bars are not included because the statistical error is typically very small whereas the variation between points on the surface can be significant because of inhomogeneous surface composition.
The adhesion data in Fig. 4 resembles an acid-base surface titration, where the chemical equilibrium of a surface ionizable group, -COOH + H2O -COO- + H3O+, (3) + is shifted towards the right when the chemical potential of H3O in solution decreases with increasing pH. The magnitude of the surface charge increases as the fraction of deprotonated ionisable surface groups increases with increasing pH. The data collected in Fig. 4 indicate that the effective surface pKa is in the range 4 to 5, which is consistent with reported pKa for surface organic acids. The pKa for a surACS Paragon Plus Environment
8
Energy & Fuels
face bound carboxyl has been estimated to be in the range 4.7 to 5.515 but a value as low as 4 is reasonable, considering that model estimates of 4 have been used successfully to predict surface aspartic and glutamic acid pKa on protein surfaces25. As the surface charge increases, repulsive interactions from double layer and desolvation forces increase and eventually dominate over van der Waals forces. For some of the experiments, the change in adhesion force was linear. These exceptions can be explained by a local difference in surface composition where response to pH is different, such as more hydrophobic character of the adsorbed organic compounds. In summary, adhesion is generally higher in ASW than in low salinity ASW, consistent with the salinity response seen in other experiments and core plug tests. Adhesion in both the ASW and the low salinity solutions decreased by ~30% on average, when pH increased from 4.5 to pH 5.5, which is about the pH where the carboxyl functional group deprotonates. Although the absolute adhesion varies considerably, in line with the inhomogeneity of the adsorbed organic material, the relative change in adhesion as a function of pH is consistent. The -CH3 tip provides clear and reproducible results. Solution pH influences adhesion, even when measured by hydrophobic tips, meaning that attachment of nonpolar oil compounds would weaken as pH increased. Force mapping with -COO(H) tips We made similar experiments with -COO(H) tips to explore the behaviour of oil compounds with polar character. We examined 5 samples with ASW and 4 with the low salinity solutions. The results are summarized in Fig. 5. As expected, the height and adhesion maps were very similar to those obtained with the -CH3 tip because we are probing the interaction between a tiny model oil droplet and the pore surface but maps with -COO(H) tips specifically explore the polar-polar interactions. 600 ASW
Adhesion Force [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 25
ASW-low
400
200
0
5
6
7
8
5
6
7
8
pH
Fig. 5. The average adhesion determined from force maps measured with -COO(H) tips in ASW (left) and low salinity solutions (right). Each series of points represents a separate experiment on different grain surfaces.
Attraction for the -COO(H) and the -CH3 tips are similar but generally higher for the polar tip. In ASW at pH 4.5 (Fig. 5, left), adhesion is generally higher for the polar tip, on average ~330 pN, with a range from 200 to 550 pN, compared with average ~300 pN for the hydrophobic tip. In the low salinity solution, average adhesion was 185 pN, with a range from 130 to 250 pN (Fig. 5, right) compared with ~100 pN with the -CH3 tip. Adhesion decreases in both solutions as the pH increases, again with the most prominant change occurring between pH 4.5 and 5.5. The adhesion in one experiment in ASW (cross symbols) was independent of pH, which is explained by a surface where the adsorbed organic material is more hydrophobic, which fits with a minimal effect of pH and weaker adhesion with the polar tip. ACS Paragon Plus Environment
9
Page 19 of 25
In summary, pH has a significant influence on adhesion of the -COO(H) tip, with the dominant change occurring between pH 4.5 and 5.5. Adhesion decreases as pH increases and the surface charge becomes more negative. The influence of pH on the low salinity effect. The coupled dependence of pH and low salinity is more difficult to describe than the effect of either variable. We conducted a considerable number of experiments using the procedure described above, both in ASW and low salinity solutions. The variation in adhesion resulting from local differences in surface properties complicates interpretation. Increasing the number of experiments provides better statistics but we wanted an approach that would minimize variation by measuring adhesion on the same area for both ASW and low salinity solutions, with all pH values. By measuring on the same area, we minimised the effect of local surface composition differences and we were able to measure adhesion differences and correlate them directly with both pH and salinity. The result of such experiments is presented in Fig. 6. –CH3 200 150
Adhesion [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
100 50 –COO(H)
400 300 200 100 pH 4.5
pH 5.5
pH 7.5
Fig. 6. Example data from two experiments, using a -CH3 and a -COO(H) tip. For each experiment with the same tip, all data were collected on exactly the same surface area, while the solutions were exchanged, to compare the salinity effect over the pH range. Pink represents ASW and blue, low salinity.
The adhesion measurement is reproducible and decreases significantly as pH increases and is lower with lower salinity (Fig. 6). Behaviour is similar with both tips, although the higher adhesion with the carboxyl tip suggests that the surface is dominated by polar molecules. The areas scanned by the two tips are not identical but comparable in terms of composition and sample history. The results presented in Fig. 7 show the same trends that we saw in the previous plots. The low salinity response was clear at all three pH values and with both tips. Adhesion is highest at pH 4.5. It declines with pH increase and becomes flat between pH 5.5 and 7.5. Adhesion is higher for the -COO(H) tip than for the -CH3 tip, completely consistent with the trends in the results collected over the broad range of surfaces.
ACS Paragon Plus Environment
10
Energy & Fuels
Adhesion force [pN]
400
--CH3 tip in ASW --CH3 tip in ASW-low --COOH tip in ASW --COOH tip in ASW-low
300
200
100
0 4
5
7
6
8
pH
Fig. 7. Average adhesion experiments with with data pooled from all of the experiments using either -CH3 or COO(H) tips in ASW and low salinity solutions.
The low salinity response can be described as the change in adhesion in two ways: as the direct adhesion difference, ∆Fadh = FadhASW - FadhASW-low (4) and as a relative change in adhesion, ∆Fºadh = [FadhASW-FadhASW-low]/ FadhASW x 100, (5) which normalises the difference to the initial surface behaviour. The relationship of ∆Fadh and ∆Fºadh to pH is presented in Fig. 8. The trend for ∆Fadh resembles the absolute adhesion data in the previous sections. The highest low salinity response occurs at pH 4.5, whereafter the effect decreases and there is no significant change in the low salinity effect at higher pH. The relative change in adhesion (inset) gives a slightly different picture. The change in adhesion relative to the initial adhesion is essentially the same for the two tips and the change with pH is not significant. The data show that a low salinity effect is expected in a large pH range and is consistent with the interpretation by Suijkerbuijk et al.11, that the pH effect discussed by Austad et al.7 might be the result of the low salinity effect rather than its cause. 300
60
--CH3 terminated tip --COOH terminated tip
200
Relative adhesion [%]
Adhesion difference [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 25
0
4
pH
8
100
0 4
5
6
7
pH
8
Fig. 8. The average difference in the pooled adhesion data, between the high and low salinity solutions , ∆Fadh, and the relative change in adhesion, ∆Fºadh (inset) for the same area with different tips.
Adhesion behaviour and DLVO theory. The double layer force, FEDL, makes the largest contribution to the pH dependence of the total adhesion force, Fadh, that we measure in our AFM experiments. In Figure 9, we depict FEDL calculated from ACS Paragon Plus Environment
11
Page 21 of 25
Equation 1 in the pH range used in this study for the ASW (blue curve) and the low salinity ASW (red curve), as starting point for interpreting our data. We use D=1 nm, ε=80, R=30 nm, σT=0 C/m2, and σS is obtained from Equation 2 with σmax=-0.015 C/m2 and pKa=4.7. With these parameters, FEDL contributes to the total adhesion force with -0.09 pN in ASW and -34 pN in low salinity solution at pH where the surface has reached its maximum surface charge (pH>6). Its contribution at low pH, where the surface is practically uncharged, is 0 pN. For the low salinity case, I=30 mM, the shape of the FEDL titration curve and the corresponding AFM titration curve in Fig. 7 are similar in the pH range 4 to 8. Although DLVO theory represents an ideal case, the overall shape of the force titrations is explained by ionization of charged surface groups. For the low salinity case, based on the DLVO analysis, we could interpret that the tip, our tiny oil droplet, perturbs the double layer. Because the distance between the tip and the surface, D, is smaller than or comparable to the Debye length, λD, the tip has to displace some of the double layer to reach the equilibrium distance, D. Disrupting the double layer results in a repulsive force, which is not insignificant compared with the forces we measured in our experiments. 50 0 F EDL (pN)
700 mM 30 mM
F EDL [pN]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
-50
0
0.4 Ionic strength (M)
0.8
0
-50 4
5
6
7
pH
8
Fig. 9. The double layer force predicted with Equations 1 and 2 for σS = -0.015 C/m2 in 700 mM (blue) and σT = 0.0 C/m2 in 30 mM ionic strength solutions (red).
At the resolution of the plot, the high salinity curve in Fig. 9 (blue) appears to be independent of pH and surface charge density but enlargement shows that its overall form is the same as for the low salinity curve. Because λD in this case is much smaller than D, the last exponential term dominates over the increasing σS2, resulting in a very small force. The adhesion curve is thus compressed at this scale and plateaus at a magnitude < 0.1 pN. The physical reason for its form is easily explained. Because the contact distance between the tip and the surface is significantly longer than the Debye-length, the tip only perturbs the double layer slightly, meaning that minimal energy is needed to disrupt the double layer. From here, the comparison between DLVO predictions and experiments becomes problematic. The force data in Fig. 7 and in many other experiments show that adhesion depends more on pH in high salinity solution than in low salinity solution, whereas the FEDL curves of Fig. 9 show an opposite trend, i.e. larger adhesion change for low ionic strength solutions. It is possible that this contradiction stems from failure of the DLVO treatment but this explanation is insufficient because only pH is changed during the titration. Uncertainties in absolute measurements can annul each other by considering differences so the degree of DLVO failure would have to vary considerably in different pH solutions to explain the inconsistencies. The difference lies in how the tip and substrate responds to a change in pH. The pH dependence of adhesion in solutions where ionic strength is constant suggests, in other words, that double layer expansion is not the only process behind the low salinity response. Double layer expansion is considered by DLVO but the failure of Equation 1 to account for pH dependent adhesion indicates that there must be another electrostatic component. If the expanded double layer alone were responsible for the low salinity effect, the change in surface charge density that accompanies change in pH ACS Paragon Plus Environment
12
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 25
would imply a proportionally higher change in adhesion resulting from pH change in the low salinity solution than in the high salinity solution. However, we see the opposite. This could be explained by lower surface charge density in the lower salinity solution, where fewer divalent ion bridges can keep charged oil components adsorbed to pore walls26. This contrasts with several interpretations presented in the literature, where lower salinity is proposed to increase repulsive electrostatic interactions between the surface and crude oil components. The sandstone σmax and D used to generate Fig. 9 have been taken from Hilner and colleagues16, who reported these values after fitting experimental data to the DLVO model, where pH and solution composition were kept constant, while only the effect of ionic strength on adhesion force was investigated. This is appropriate when σS and D are constant in the range of their experiments, in this case different salinities. A recent study by Olsson et al.26 however, suggests that σS depends strongly on salinity and therefore has to be analyzed as a separate variable rather than as a constant. In that case, such a fit is not warranted. Fitting σS and D to experimental data using Equation 1 also assumes that DLVO theory is quantitatively correct, which is unlikely because the DLVO model is an ideal, macroscopic, mean field model that does not consider chemical structure. The experimental and predicted values for σS and D reported in Hilner et al. are surprisingly similar but the uncertainties are significant. In this study, we do not fit any parameter to Equation 1, but realize that a constant surface charge density, which is assumed in prevailing explanations of the low salinity effect, is inconsistent with observations. This is independent of the values of σMax and D that we chose to generate Fig. 9 but the magnitude of this discrepancy depends on the magnitude of σmax and D. In principle, we could extract σS = -0.015 C/m2 as an estimate for the low salinity solution, because FEDL does not contribute to Fads in high salinity, according to Equation 1, i.e. Fig. 9. This assumes that the tip-surface distance at contact, D, remains the same in the studied ionic strength range, i.e. from 700 to 30 mM, which is feasible but seems unlikely. The inset in Fig. 9 is a plot of the DLVO double layer force calculated from Equation 1 with σS values of -0.010, -0.015 and -0.020 C/m2. The shape of the three curves is identical but their adhesion span corresponds to different values for σS2 in Equation 1. Thus, the ionic strength where the curve decreases exponentially is determined by the fit to the experimental data collected from the low salinity experiments so reasonable values for σS and D derived from the fit, that correspond also to the high ionic strength solutions, might be somewhat fortuitous. Conclusions We used chemical force mapping to explore the interactions between a functionalized tip that is made to model polar and nonpolar oil compounds and quartz grains plucked from a reservoir sandstone. The purpose was to discover the processes that are responsible for the low salinity effect. We measured the adhesion in high and low salinity solutions, with pH spanning the range from 4.5 to 7.5. For experiments with -CH3 and with -COO(H) tips, adhesion decreased as pH increased, regardless of salinity. The adhesion curves resemble titration curves. Absolute adhesion was highest for experiments with the -COO(H) tip in high salinity solution, then -COO(H) tip in low salinity, the -CH3 tip in high salinity and -CH3 tip in low salinity solution (FadhASW, COOH > FadhASW-low, COOH ≈ FadhASW, CH3 > FadhASW-low, CH3). For all, adhesion decreased with increasing pH, especially for the transition from 4.5 to 5.5, near the pKa value of carboxyl deprotonation. The decrease in adhesion became smaller with increasing pH so there was very little difference between pH 5.5 and 7.5. The DLVO relationship is not effective for explaining the data. It predicts large pH dependence in low salinity and small to none in high salinity solution, opposite to results from experiments. Failure to describe our results is understandable however because DLVO theory was developed to describe interactions between particles with a certain surface charge density, in an uncomplicated electrolyte with specific ionic strength. Our results make it clear that the surface charge density and equilibrium tip-surface distance depend on the conditions of the system and cannot be assumed to be constant such as when pH and ionic strength are changed. ACS Paragon Plus Environment
13
Page 23 of 25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
Our result therefore confirms that electric double layer behaviour, while an integral part of the low salinity effect, is not a simple mechanism. Change in surface charge density, as would occur when low salinity solutions are injected, also changes the ion environment in the solution near the pore surfaces, whether the fluid is in contact with the bare mineral surface or with adsorbed organic compounds. The results of these experiments have provided new insight into the fundamental processes that control the low salinity response but do they have implications for enhancing oil recovery? No matter what the salinity, the surfaces were more water wet at pH 5.5 than at 4.5. This is fortunate for oil production because there is evidence to indicate a buffering effect in sandstone reservoirs that holds pH at about 5.5. For solutions where pH>5.5, adhesion continues to decrease but the extent of decrease is smaller. This suggests that sea water, with its pH of 8.3, or low salinity water in the neutral range, would be quite effective at decreasing adhesion. This change of surface charge is one of the reasons for the effectiveness of alkali flooding. Sea water pH drops to ~5.5 very soon after injection, mainly because of reaction with gases dissolved in the pore fluids, but our results imply that keeping pH in the neutral range would have a slight positive effect on oil production. However, the effectiveness of the low salinity response at pH 5.5, shown by our study, suggests that the benefit of enhancing oil recovery by raising or controlling water flood pH would be marginal. Acknowledgements We thank the members of the NanoGeoScience Research Section for discussion and Maria Bruun for help in the laboratory. This work was part of the Nano-Sand Project, funded by BP Exploration Operating Company Limited. We also acknowledge the Danish e-Infrastructure Cooperation, DeIC, for computer facilities. Corresponding authors *
[email protected],
[email protected] Abbreviations ASW, artificial seawater; ASW-low, low-salinity artificial seawater; AFM, atomic force microscopy; CFM, chemical force microscopy; DLVO theory, Dejarguin, Landau, Verwey, Overbeek theory. References
1. Morrow, N. R., Tang, G. Q., Valat, M., and Xie, X. (1998) Prospects of improved oil recovery related to wettability and brine composition, J. Petrol. Sci. Eng. 20, 267-276. 2. Lager, A. W., Collins, I. R., and Richmond, D. M. (2008) LoSal(TM) Enhanced Oil Recovery: Evidence of Enhanced Oil Recovery at the Reservoir Scale, SPE, 113976, In Improved Oil Recovery Symposium. 3. McGuire, P. L., Paskvan, F. K., Sommer, D. M., and Carini, F. H. (2005) Low Salinity Oil Recovery: An Exciting New EOR Opportunity for Alaska’s North Slope., SPE, 93903, In Improved Oil Recovery Symposium, Irvine, California. 4. Tang, G. Q., and Morrow, N. R. (1999) Influence of brine composition and fines migration on crude oil/brine/rock interactions and oil recovery, J. Petrol. Sci. Eng. 24, 99-111. 5. Sharma, M. M., and Filoco, P. R. (2000) Effect of brine salinity and crude-oil properties on oil recovery and residual saturations, In SPE Journal, 5, 293-300. 6. Lager, A., Webb, K. J., Black, C. J. J., Singleton, M., and Sorbie, K. (2006) Low salinity oil recovery: An experimental investigation, Petrophysics 49, 12. ACS Paragon Plus Environment
14
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 25
7. Austad, T., RezaeiDouhst, A., and Puntervold, T. (2010) Chemical Mechanisms of Low Salinity Water Flooding in Sandstone Reservoirs., SPE, 129767-PP, In Improved Oil Recovery Symposium. 8. Lager, A., Webb, K. J., and Black, C. J. J. (2007) Impact of brine chemistry on oil recovery, In 14th European Symposium on Improved Oil Recovery, Cairo, Egypt. 9. Nasralla, R. A., and Nasr-El-Din, H. A. (2012) Double-Layer Expansion: Is It a Primary Mechanism of Improved Oil Recovery by Low-Salinity Waterflooding, SPE, 154334-PA, In Improved Oil Recovery Symposium, Tulsa, Oklahoma, USA. 10. Ligthelm, D. J., Gronsveld, J., Hofman, J. P., Brussee, N. J., Marcelis, F., and van der Linde, H. A. (2009) Novel waterflooding strategy by manipulation of injection brine composition. Paper SPE 119835, In EUROPEC/EAGE Conference and Exhibition Amsterdam, The Netherlands. 11. Suijkerbuijk, B. M. J. M., Hofman, J. P., Ligthelm, D. J., Romanuka, J., Brussee, N. J., van der Linde, H. A., and Marcelis, A. H. M. (2012) Fundamental investigations into wettability and low salinity flooding by parameter isolation, SPE, 154204-MS, In Improved oil recovery symposium, Tulsa, Oklahoma, USA. 12. Lee, S. Y., Webb, K. J., and Collins, I. R. (2010) Low salinity oil recovery - increasing understanding of the underlying mechanisms, SPE, 129722-MS, In Improved oil recovery symposium, Tulsa Oklahoma. 13. Matthiesen, J., Bovet, N., Hilner, E., Andersson, M. P., Schmidt, D. A., Webb, K. J., Dalby, K. N., Hassenkam, T., Crouch, J., Collins, I. R., and Stipp, S. L. S. (2014) How Naturally Adsorbed Material on Minerals Affects Low Salinity Enhanced Oil Recovery, Energy Fuels 28, 48494858. 14. Pedersen, N. R., Hassenkam, T., Ceccato, M., Dalby, K. N., Mogensen, K., and Stipp, S. L. S. (2016) The low salinity effect at pore scale: Probing wettability in Middle East limestone, Energy Fuels, submitted 15. Andersson, M. P., Olsson, M. H. M., and Stipp, S. L. S. (2014) Predicting the pK(a) and Stability of Organic Acids and Bases at an Oil-Water Interface, Langmuir 30, 6437-6445. 16. Hilner, E., Andersson, M. P., Hassenkam, T., Matthiesen, J., Salino, P. A., and Stipp, S. L. S. (2015) The effect of ionic strength on oil adhesion in sandstone - the search for the low salinity mechanism, Scientific Reports 5. 17. Isrelachvili, J. N. (2010) Intermolecular and surface forces, 3 ed., Academic press. 18. Hutter, J. L., and Bechhoefer, J. (1993) Calibration of Atomic Force Microscope Tips, Rev. Sci. Instrum. 64, 1868-1873. 19. Love, J. C., Estroff, L. A., Kriebel, J. K., Nuzzo, R. G., and Whitesides, G. M. (2005) Selfassembled monolayers of thiolates on metals as a form of nanotechnology, Chem. Rev. 105, 1103-1169. 20. Hassenkam, T., Skovbjerg, L. L., and Stipp, S. L. S. (2009) Probing the intrinsically oil-wet surfaces of pores in North Sea chalk at subpore resolution, P. Natl. Acad. Sci. USA 106, 6071-6076. 21. Skovbjerg, L. L., Hassenkam, T., Makovicky, E., Hem, C. P., Yang, M., Bovet, N., and Stipp, S. L. S. (2012) Nano sized clay detected on chalk particle surfaces, Geochim. Cosmochim. A. 99, 5770. 22. Matthiesen, J., Bovet, N., Hilner, E., Andersson, M. P., Schmidt, D., Dalby, K. D., Hassenkam, T., Crouch, J., Collins, I. R., and Stipp, S. L. S. (2014) How organic material adsorbed on minerals affect low salinity enhenced oil recovery in sandstone reservoir, Energy Fuels 28, 4849-4858. 23. Butt, H. J., Cappella, B., and Kappl, M. (2005) Force measurements with the atomic force microscope: Technique, interpretation and applications, Surf. Sci. Rep. 59, 1-152. 24. Stipp, S. L., and Hochella, M. F. (1991) Structure and Bonding Environments at the Calcite Surface as Observed with X-Ray Photoelectron-Spectroscopy (XPS) and Low-Energy ElectronDiffraction (LEED), Geochim. Cosmochim. A. 55, 1723-1736.
ACS Paragon Plus Environment
15
Page 25 of 25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
25. Olsson, M. H. M., Sondergaard, C. R., Rostkowski, M., and Jensen, J. H. (2011) PROPKA3: Consistent Treatment of Internal and Surface Residues in Empirical pK(a) Predictions, J. Chem. Theor. Comp. 7, 525-537. 26. Olsson, M. H. M., Matthiesen, J., Dobberschütz, S., Andersson, M. P., Pedersen, N., Hassenkam, T., and Stipp, S. L. S. (2014) The protonation state of ionizable self assembled monolayers at various pH and ionic strengths; predictions and experiments, J. Am. Chem. Soc., submitted.
ACS Paragon Plus Environment
16
