Investigation on the adsorption behavior of polyacrylamide on resin by dual polarization interferometry

Abstract

Polyacrylamide (PAM) is widely used in the petroleum industry to enhance oil recovery all over the world. However, introducing PAM into the liquid changes the property of the oil–water interface and, finally, makes the wastewater treatment more difficult. It is ambiguous whether PAM adsorbs onto the oil–water interface and influences the stability of the interface. In this study, resin, one of the most important components in crude oil, was immobilized onto the silicon oxynitride chip surface. The behavior of PAM on resin was investigated by dual polarization interferometry (DPI). DPI is a novel instrument to real-time monitor the adsorption behavior and structure changes of a polymer on resin. Different concentrations of PAM solutions were respectively injected onto the immobilized chips. The real-time mass, thickness and density changes of PAM on resin were accordingly recorded and calculated by DPI. In conclusion, when a low-concentration polymer solution was injected, polymer molecules were adsorbed on the resin in disorder; when a high-concentration polymer solution was injected, pre-adsorbed polymer molecules were rearranged to take in more molecules and an isotropic adsorption layer will be formed.

---

Introduction

Polyacrylamide (PAM) has been widely used in many fields, such as flocculants, dehydrating agents, retention agents, filter aids and enhanced oil recovery.^1,2 As one of the most utilized polymers, it has attracted much attention due it has properties,^3–5 especially in the petroleum industry. Although polymer flooding is an important and highly effective method in enhanced oil recovery, some technical problems still exist. For example, it is hard to treat the produced water from polymer flooding because of its high viscosity and emulsification degree. Recently, some researchers suggest that PAM remaining on the oil–water interface is the main cause for this problem.^6–8 Research conducted by Zhifeng Lv reveals that polymers could enhance the stability of emulsions.⁹ Wei Zhang thinks that polymer molecules diffuse towards the interface of oil and water, leading to the change of interfacial properties.¹⁰ Xiangchun Meng thinks that polymer enhance the viscosity of the water phase and reduce the collision of oil droplets.¹¹ However, whether there are interactions between polymer and crude oil is still ambiguous. Resin, as one of the main components of crude oil, is important for the stabilization of oil–water interface for its polarity. Therefore, investigating of the adsorption behavior of PAM on resin will elucidate the interaction at the oil–water interface. And then we can provide more useful measures to oil and water separation.

It is possible to acquire the adsorption behavior of polymer on the oil–water interface with technologies of ellipsometry,^12,13 neutron reflection,¹¹ fluorescence spectroscopy,^14–16 radiochemistry,^17,18 interfacial tensiometry^19,20 and so on. All of them, however, have their limitations. Ellipsometry has been emerged as an increasingly pivotal technology in the past decades. But it can only provide relative thickness of adsorbed layer. Although neutron reflection could obtain static structural information in detail, it is not suitable for real-time measuring. Methods of fluorescence spectroscopy and radiochemistry method, in which labeling must be involved, can achieve high detection sensitivity. While labeling process may introduce chemical modification of reactants. Interfacial tensiometry is the most widely used method, but the adsorbed mass measured is lower than the actual mass.

These days, a newly developed technique was introduced into the research of surface and interface science. Dual polarization interferometry (DPI) is a label-free and surface-sensitive technique which is widely used in the solid–liquid interface to record the real-time changes of layer mass, thickness and density.^21,22 Due to the advantages of DPI, it has been applied to a variety of research areas including surfactants, protein structural changes, polyelectrolyte assemblies, biomimic biomembranes and crystallization conditions.^23–29 It is proved that DPI is a powerful and accurate method for monitoring the adsorption behaviors on the solid–liquid interface in real-time.

In the present paper, DPI was introduced to investigate the interaction between PAM and resin. Different concentrations of PAM solution were prepared by dissolving in Milli-Q water. Analyzing as isotropic and anisotropic adsorbed layers, the real-time mass, thickness and density changes on the adsorbed layers were recorded by DPI. Based on the above data, the adsorption behavior of PAM on resin was discussed in detail. It is shown that the mass transfer and rearrangement of adsorbed molecules are the mainly forces for the adsorption. However, the interaction between polymer and resin was so weak that little polymer adsorbed on resin and the adsorbed polymer molecules could be rinsed by flow phase (Milli-Q water in this case).

Experimentals

Polyacrylamide was synthesized in laboratory. The molecular weight of polymer is 2.1 × 10⁵. Resin was extracted from crude oil of SZ 36-1 offshore well and dissolved in toluene–heptane mixtures (1 : 1, in volume) with the concentration of 0.3% (wt). Filter membranes (pore size = 0.45 μm) were purchased from Kelong Chemical Reagent Company of China.

All the solutions used in the present study were prepared with Milli-Q water. 0.8 g L⁻¹ PAM stock solution was prepared and filtered. Then the filtered solution was diluted to concentrations as needed. Resin solution used for immobilization was obtained by following steps. Firstly, resin was dissolved by toluene–heptane mixtures and then treated by ultrasonic vibration (KH-50B, Kunshan Hechuang Ltd, China) for 30 minutes. In the end, the vacuum filtration and filter membranes were also used. The silicon oxynitride sensor chips (FB 100 Unmod chips, Farfield Scientific Ltd, UK) were used for DPI experiments. The resin was immobilized on the chip by spin coating with the spinning speed of 1000 rpm.

An Analight® DPI (Farfield Group Ltd, Crewe, UK) was used to record the adsorption of PAM on resin. A detailed description (chip structure, operation procedures, calculation theory) of the instrument has been reported in previously papers.^29–31 Scheme of the structure of DPI chip is present in the Scheme 1. In brief, the senor chip comprises dual slab waveguides (sensing waveguide and reference waveguide) and three channels (one reference channel and two working channels). The glass clad channel 2 acts as reference channel. Meanwhile, the fluidic channels, channel 1 and channel 3, exposed to the test solution directly. The helium–neon laser is used as light source. The light travels through the chip from one end to another, with two separate orthogonal polarized lights, transverse electric (TE) and transverse magnetic (TM). Total reflection happens within the reference and sensing channel. Whenever adsorption occurs at the water/solid interface, the phase angles of the two polarized lights will be influenced. The emergent light at the end of the chip will be different with the reference lights. Then the interferometry fringes will shift quantitatively. By real-time recording the shifts and calculating, the thickness and refractive index (RI) of adsorbed layer can be obtained. In addition, DPI enables analysis for both isotropic and anisotropic layers. For isotropic layer, the RI, thickness, mass and density of adsorbed layer can be resolved. And the birefringence is regarded as zero. For anisotropic layer, the mass and birefringence of the adsorbed layer can be resolved with fixed RI and RII (refractive index increment). If the layer is anisotropic, it cannot be resolved as an isotropic layer. If using isotropic analysis, the calculation gives errors or unreasonable values. The adsorbed mass, density and molar coverage can also be quantitatively calculated according to De Feijter's equations as follows.³²

m_L = ρ_L × τ_L (2)

where ρ_L is the layer density (ng mm⁻³), n_L and n_buffer are RIs of adsorbed layer and buffer solution respectively. m_L is the layer mass per unit area (ng mm⁻²), τ_L is the layer thickness (nm) and d_RI/d_c is the RI increment of PAM. A d_RI/d_c value of 0.3874 is used for isotropy analysis and a RI value of 1.45 is used for anisotropy analysis in this paper.

Scheme 1 Schematic representation of the structure of DPI sensor chip.

All the experiments were conducted at 20 °C ± 0.1 °C and Unmod chips immobilized with resin were used. Before the experiment, the buffer (Milli-Q water in this case) and sample solutions were degassed to avoid errors caused by air bubbles. Then sample solution was injected and flowed with the rate of 5 μL min⁻¹ and total PAM volume of 200 μL. Afterwards, running buffer was flowed over channels with the rate of 10 μL min⁻¹ until the balance was reached. Then, toluene was injected at the rate of 10 μL min⁻¹ for 40 minutes for further chip rinsing. In this rinse step, resin and polymer were totally washed. Finally, chip and buffer calibration was conducted by using 80% ethanol–water (v/v) mixture solution and water.

Results and discussion

The conformational changes of adsorbed layer during the adsorption and desorption process

In the adsorption experiments, DPI was used to study the adsorption behavior of PAM with resin. Herein, a resin layer was immobilized on the surface of oxynitride sensor chip. Through the experiments, the real-time data of TM and TE polarizations were recorded and analyzed by the software Analight Explorer. Finally, the real-time adsorbed mass, thickness and density data could be obtained.

In the blank experiment, PAM of 400 ppm was injected into DPI and the resin was not immobilized on the chip. The adsorbed data were resolved as an anisotropic layer. As can be seen in the Fig. 1, the plots of adsorbed mass and thickness increase immediately at the start of injection and a change of slope occurs at around 300 s. At the beginning of buffer rinsing, adsorbed mass and thickness decrease slowly and almost reach a plateau at around 2500 s. Adsorbed mass and thickness after buffer rinsing are around 0.01 ng mm⁻² and 0.035 nm, respectively. Obviously, the variation range of mass and thickness are 0.02 ng mm⁻² and 0.08 nm, respectively. Compared with the adsorbed data in the Fig. 2c, adsorbed mass and thickness changes with 400 ppm of PAM are slight. All the differences indicate that PAM does not simply replace all or part of the pre-bound resin.

Fig. 1 PAM was injected onto Unmod sensor chip as a blank experiment.

Fig. 2 PAM in 100 ppm was injected onto resin surface in Fig. 1a; PAM in 200 ppm was injected onto resin surface in Fig. 1b; PAM in 400 ppm was injected onto resin surface in Fig. 1c; PAM in 800 ppm was injected onto resin surface in Fig. 1d.

In the follow-up experiments, resin was immobilized on the chip and these chips acted as hydrophobic surface. Some literatures indicated that resin is a complex mixture of saturated and aromatic compounds. And resin is rich in nitrogen and oxygenated compounds with short carbon chains.^33,34 However, its accurate molecular structure is still unrevealed. The plots of adsorbed mass, thickness and density changes of PAM are shown in Fig. 2. It is clearly that all the adsorbed parameters are functions of time with polymer concentration from 100 ppm to 800 ppm. Anisotropic analysis method was used to calculate the adsorbed data of 100 ppm polymer while isotropic analytic method was used for the other concentrations. As shown in the Fig. 2, there is certain amount of mass and thickness before sample injection due to resin-coated on the sensor chip. Then, adsorbed mass and thickness are sharply increased and the density is rapidly decreased at the start of polymer injection, indicating the structure changes of the adsorbed layer at the initial adsorption. As shown in Scheme 2, resin is immobilized on the sensor chip tightly in the first step and polymer molecule bonds to it loosely (compared with resin) at the initial adsorption, which results in the increase of mass and thickness along with the decline of density. In the subsequent adsorption, polymer on the resin surface changes from diffused to dense, which will affect the layer structure and adsorption behavior. With continuous adsorption, adsorbed mass and thickness increase slowly and reach a short-lived semi-plateau. While all the curves increase rapidly again several seconds later until the peaks. In the end, all the mass and thickness plots begin to decrease because of water rinsing and adsorbed plateau is observed later. Due to the water solubility of polymer, the adsorbed plateau is an almost-plateau. Compared all the curves, it is interesting to find that the changes of mass were less evident than thickness changes. Besides, the variation region of density curves is smaller than variation region of mass and thickness (Fig. 2). Obviously, different conformational changes occurred at the periods. Adsorbed polymer molecules extend to liquid phase because the interaction with polymer molecules existed in the liquid phase.

Scheme 2 Schematic representation of preparation steps and adsorbed behaviors on DPI sensor chips during experiment process.

Analysis on adsorption dynamics of PAM with resin

In order to interpret the adsorption dynamics, it is necessary to introduce the knowledge of chemical engineering and physical chemistry such as mass transfer and the fluid type.

The calculation method used to obtain the fluid type is listed as follow,

where Re represents the Reynolds number and d is the diameter of fluid in the channels. ρ and μ are density and viscosity of fluid. u is the flow rate of fluid. The flow type is named as laminar flow region when the value of Re is less than 2000, whereas the flow type is called as turbulence region when the value of Re is more than 4000 and flow type is transition region in the other condition.

In this study, the volume of fluid flowed over the DPI chip is 1 μL (the width, length and height are 1 mm, 5 mm and 0.2 mm, respectively), with a flow rate of 5 μL min⁻¹. It takes about 12 s for the sample solution flow over the channels. Herein, it is reasonable to consider the influence of flow type on the experiments. Hence, eqn (3) is used and a laminar flow is ensured.

Based on the theory of laminar flow, the fluid on the chip can be equally divided into countless layers. If the polymer solution which flows through the sensor chip is equally divided into n parts, the equation of mass transfer can be expressed as follow,^35–37

J = K_m × (C_n+1 − C_n) (4)

where J is the flux determined by mass transport coefficient K_m and concentration differences. C₀ is the initial concentration of the polymer on resin surface, C₁ is the concentration of the first polymer solution layer adjacent to the resin surface while C₂ is the concentration of second polymer solution layer adjacent to the first, and so on. D_app is the apparent diffusion constant, F is volumetric flow rate. h, w and l are height, width and length of flow channel, respectively. From the above equations, it is obvious that K_m, D_app, V, l and w are constants and J is determined by concentration differences only.

As all the behaviors of PAM adsorption with resin are similar, here we just use the PAM concentration of 800 ppm as the example for discussing the adsorption dynamics.

The adsorption rate dm/dt is plotted against mass m to explain the whole isotropic adsorption process (Fig. 3a). Herein, the complete process of adsorption is separated into several segments and five points are labeled on the plot. At the beginning of each experiment, adsorption sites of resin are occupied by flow phase (Milli-Q water). After injection, polymer solution is influxed and last for 12 seconds. Consequently, the adsorption rate increases rapidly in the first 12 seconds (OA segment). But the tendency driven by mass transfer plays little role in this process. During the AB segment, water molecules which bonded to the resin surface are replaced gradually by polymer molecules due to mass transfer. In this process, mass transfer at the interface of resin and polymer solution is driven by concentration gradient. Meantime, C₀ is rather small and can be regarded as zero, while C₁ is the bulk concentration. In addition, the concentration differences between polymer solution layers results in mass transfer within polymer solution and can't reach equilibrium during the experiments, because fluid is pumped in continuously, whereas the concentration differences was reduced gradually. Then, the adsorption rate increases slightly again for a short time and decreases later (BC and CD segments, respectively). The raise of the BC segment is resulted from the rearrangement of polymer molecules as the adsorbed layer initially formed was not stable.^37,38 At the beginning of adsorption, polymer molecules transfer and diffuse onto the adjacent layer. With continuous adsorption, polymer bonded on the resin surface changed from loose to dense. Therefore, those adsorbed polymer rearranges to make the surface more available for subsequent adsorption, resulting in increased adsorption rate.³⁵ At the point D, numerous mass points were appeared, indicating that the adsorbed rate has been reached balance. As shown in the Fig. 3b, two different slopes were observed after we plotted the layer thickness change against the mass change. This result indicates that a steep increase of thickness occurs after initial adsorption. It is obviously that the break of the slope corresponds to the point D in the Fig. 3a. After the point D, although the adsorbed polymer molecules rearrange themselves to take more complexes from the solution, the chip surface was so crowded that the following polymer molecules could only attach partly onto the surface, resulting in the steep increase in layer thickness and dramatic decrease in layer density.

Fig. 3 Changes of adsorption rate as a function of layer mass was shown in (a); changes in the layer thickness as a function of layer mass was shown in (b).

In order to discuss the adsorption dynamics comprehensively, it is necessary to consider the anisotropic adsorption. In this study, the adsorption behavior in 100 ppm PAM is used as an example to discuss the anisotropic adsorption dynamics.

Adsorption rate dm/dt is plotted against mass m for better understanding of the whole isotropic adsorption process (Fig. 4a). Herein, it is obviously that the dm/dt-mass changes in 100 ppm are different from 800 ppm. As can be seen in Fig. 4, there is only single peak in the plot. To analyze anisotropic adsorption dynamics clearly, the plot in Fig. 4 is divided into four segments and three points. As same as the isotropic adsorption, water molecules occupy the resin surface before sample injection. During injecting, PAM molecules substitute the surface and adsorb with resin. It costs about 12 s to make polymer solution to flood over the chip surface. Adsorption rate increases sharply in this process (OA segment). After reaching peak point A, adsorption rate decreases gradually (AB segment). In this process, the adsorption is driven by mass transfer. But the adsorption rate declines slowly with the decrease of concentration differences. At the point B, adsorption rate reaches a balance. As shown in the Fig. 4b, the change of thickness was plotted against the mass change. It is obvious that thickness increases linearly with the increase of mass, indicating that there is no structural change of adsorbed layer.

Fig. 4 Changes of adsorption rate as a function of layer mass was shown in (a); changes in the layer thickness as a function of layer mass was shown in (b).

In order to demonstrate the adsorbed behavior and dynamics of polymer on resin in detail, preparation step and adsorption process on DPI silicon oxynitride chip surface are schematically depicted in Scheme 2. Before each experiment, a uniform resin layer should be constructed on the DPI sensor chip (A). At the initial stage of the experiment, Milli-Q water is used as flow phase, which flows onto the surfaces of both channels continuously until a steady baseline comes into being. For the high-concentration polymer solution, polymer molecules adsorb on the resin surface in disorder at the beginning of injection (B). With the continuous injection, the approaching polymer molecules lead to the rearrangement of adsorbed polymer and more polymer molecules could be bonded to, resulting in the formation of isotropic adsorption layer (C). However, little polymer molecules adsorb on the resin surface in the injection of low-concentration polymer solution. In this case, PAM adsorbs on the resin surface in disorder and all the orientations are just possible, such as end-on, side-on and lying-on (D). This disordered adsorption phenomenon can be explained by Gibbs function. Based upon the thermodynamics theory, the surface Gibbs function of adsorbents will be reduced in the process of adsorption. In order to minimize the Gibbaloney energy of resin and obtain a steady adsorption system, more adsorption site should be occupied.³⁹ Therefore, disordered adsorption is more possible when low-concentration polymer solution is injected onto DPI chip.

Conclusion

Through these experiments, we confirm that DPI is a useful and easily operated instrument to real-time monitor the adsorption behavior and structure changes of adsorbate, which could be applied into surface and interface science. On the other hand, we ensure that there are some interaction between PAM and resin indeed, which resulted in adsorption behavior of polymer on resin. In addition, two adsorption types of PAM named as isotropic adsorption and anisotropic adsorption were observed. The real-time adsorbed mass, thickness, density recorded by DPI reveal the adsorption dynamics of polymer on the resin. Different adsorption types are caused by different concentration of PAM solution. The introduction of polymer solution and mass transfer are two principal forces for those two adsorption types. Furthermore, isotropic adsorption layer will be formed in the high-concentration polymer solution while anisotropic adsorption layer will be formed in the low-concentration solution. Therefore, the treatment of wastewater should aim at the desorption of polymer on resin.

Acknowledgements

The work was supported by the Open Founds of State Key Laboratory of Offshore Oil Exploitation (CRI2012RCPS0151CNN), the New Century Talent Supporting Project (NCTET-13-0983). Sichuan Youth Science & Technology Foundation for Distinguished Young Scholars (2013JQ0009) and Applied Basic Research Programs of Science and Technology Commission Department of Sichuan Province (2014JY0120). The authors were grateful to Dr Marcus Swann for useful discussion in data analysis.

References

1. Y. Li, M. G. Shao and R. Horton, Procedia Environ. Sci., 2011, 11, 763.
2. R. E. Sojka and J. A. Entry, Environ. Pollut., 2000, 108, 405.
3. M. Ochowiak, L. Broniarz-Press, S. Rozanska, J. Rozanski and S. Woziwodzki, Procedia Eng., 2012, 42, 1849.
4. S. B. Deng, R. B. Bai, J. P. Chen, G. Yu, Z. P. Jiang and F. S. Zhou, Colloids Surf., A, 2002, 211, 275.
5. R. Yuan, Y. Li, H. B. Fang and W. Wang, Acta Phys.-Chim. Sin., 2013, 29, 706.
6. H. Zong, M. Q. Lin, P. X. Gu, H. Wu and M. Y. Li, Appl. Chem. Ind., 2005, 34, 86.
7. S. B. Deng, F. S. Zhou, Z. X. Chen, F. J. Xia, G. Yu and Z. P. Jiang, Environ. Sci. Technol., 2002, 23, 69.
8. X. L. Cao, L. Xu, Y. Li, S. X. Jiang, H. Q. Sun and L. X. Dou, J. Univ. Pet., 2005, 29, 70.
9. Z. F. Lv, Z. X. Wang, F. He, J. H. Xiao and X. B. Zhang, Chem. Eng. Oil Gas, 2008, 4, 333.
10. W. Zhang, M. Y. Li, M. Q. Lin, Z. X. Dong and J. X. Guo, Daqing Petroleum Geology and Development, 2007, 26, 110.
11. X. C. Meng, Environ. Pollut. Control, 2008, 7, 28.
12. S. C. Russev, T. V. Avguriov and T. D. Gurkov, Colloids Surf., B, 2000, 19, 89.
13. R. Teppner and H. Motschmann, Macromolecules, 1998, 31, 7467.
14. J. Bowers, A. Zarbakhsh, J. R. P. Webster, L. R. Hutchings and R. W. Richards, Langmuir, 2001, 17, 140.
15. M. J. Tupy, W. H. Blanch and J. C. Radke, Ind. Eng. Chem. Res., 1998, 37, 3159.
16. A. Gajraj and R. Y. Ofoli, Langmuir, 2000, 16, 8085.
17. A. Gajraj and R. Y. Ofoli, Langmuir, 2000, 16, 4279.
18. R. M. Wang and J. B. Schlenoff, Macromolecules, 1998, 31, 494.
19. C. J. Beverung, C. J. Radke and H. W. Blanch, Biophys. Chem., 1999, 81, 59.
20. T. Sengupta and S. Damodaran, Langmuir, 1998, 14, 6457.
21. F. Millet, M. Nedyalkov, B. Renard, P. Perrin, F. Lafuma and J. J. Benattar, Langmuir, 1999, 15, 2112.
22. F. J. Huang and H. J. Liang, ACS Appl. Mater. Interfaces, 2013, 5, 5025.
23. M. Duan, H. Wang, S. W. Fang and Y. Liang, J. Colloid Interface Sci., 2014, 417, 285.
24. M. J. Swann, L. L. Peel, S. Carrington and N. J. Freeman, Anal. Biochem., 2004, 329, 190.
25. K. Karim, J. D. Taylor, D. C. Cullen, M. J. Swann and N. J. Freeman, Anal. Chem., 2007, 79, 3023.
26. A. Mashaghi, M. Swann, J. Popplewell, M. Textor and E. Reimhult, Anal. Chem., 2008, 80, 3666.
27. B. Lillis, M. Manning, H. Berney, E. Hurley, A. Mathewson and M. M. Sheehan, Biosens. Bioelectron., 2006, 21, 1459.
28. A. Boudjemline, E. Saridakis, M. J. Swann, L. Govada, I. M. Mavridis and N. E. Chayen, Anal. Chem., 2011, 83, 7881.
29. H. Berney and K. Oliver, Biosens. Bioelectron., 2005, 21, 618.
30. P. D. Coffey, M. J. Swann and T. A. Waigh, J. Opt. Soc. Am., 2009, 17, 10959.
31. O. Tan and G. H. Cross, Phys. Rev. E: Stat., Nonlinear, Soft Matter Phys., 2009, 79, 021703.
32. J. A. De Feijter, J. Benjamins and F. A. Veer, Biopolymers, 1978, 17, 1759.
33. A. Gaspar, E. Zellermann, S. Lababidi, J. Reece and W. Schrader, Energy Fuels, 2012, 26, 3481.
34. Y. J. Cho, J.-G. Na, N.-S. Nho, S. H. Kim and S. Kim, Energy Fuels, 2012, 26, 2558.
35. S. Sjölander and C. Urbaniczky, Anal. Chem., 1991, 63, 2338.
36. R. Glaser, Anal. Biochem., 1993, 213, 152.
37. G. Bijelic, A. Shovsky, I. Varga, R. Makuska and P. M. Claesson, J. Colloid Interface Sci., 2010, 348, 189.
38. K. Xu, M. M. Ouberai and M. E. Welland, Biomaterials, 2013, 34, 146.
39. X. C. Yan and M. D. Luo, Chemical Industry Press, 2005, 7, 1111.

---