On unexpected behavior of viscosity of diethylene glycol-based MgAl₂O₄ nanofluids

Abstract

The article is a continuation of research on the rheological properties of diethylene glycol-based MgAl₂O₄ nanofluids. Previous article [Żyła et al., RSC Adv., 2013, 3, 6429–6434] presents the results of research on the rheological properties of the material. The results have shown an unexpected behavior of the viscosity curves for this material provided that the dynamic viscosity of the nanofluid in a certain area of the shear rate decreases, then increases, and beyond a certain shear rate declines. Additionally, it has shown that this effect can be observed only with the use of freshly prepared samples. If the samples are subjected to long-term shear, the experiment results with pseudoplastic flow, with greater dynamic viscosity than in the case of fresh samples. This paper is an attempt to explain this phenomenon. The authors have conducted a series of experiments using RheoScope in the range of shear rates from 1 s⁻¹ to 1000 s⁻¹ and Rheo-NMR. Studies have shown that during the time of rotational measurements nanoparticles agglomerate in suspension, which causes the unexpected shape of the viscosity curves of the material.

---

1 Background

Nanofluids are a new group of materials which are the suspensions of nanoparticles in a liquid base. Despite the fact that this group of materials is relatively new, it are more and more advanced in its' application both in industry and medicine.^1,2

The addition of nanoparticles to the liquid results in an increase in the thermal conductivity of the resulting material,^3,4 which rationalises nanofluids applications. Due to the increased thermal conductivity, nanofluids are more convenient in heat exchanging processes while minimising systems. One of the most promising industrial uses of nanofluids is cooling modern electronics as described by Ijam et al.⁵ Senthilraja et al.⁶ and Leong et al.⁷ presented papers in which the possibility of using nanofluids in the automotive industry appears. Kulkarni et al.⁸ describe possibilities of application of nanofluids in heating buildings. He et al.⁹ presented the possibility to use nanofluids in the process of electricity generation.

The wide potential in applications of nanofluids cause acceleration of the experimental work on the nature of their thermal conductivity.^10–14 In practice however, the use of these materials must be well acquainted with their mechanical properties and rheological properties in particular. The rheological properties of nanofluids are intensively studied.^15–21 Unfortunately, classic theoretical models of rheological properties of suspensions do not apply to the case of nanofluids. Mackay et al. showed that the rheological properties of nanofluids depend on nanoscale effects.²² Due to the creation of experimental results database, there is a need of more accurate experiments, which, in future, might be use to create a theoretical model.

There are two basic methods for preparing nanofluids: (a) one-step method and (b) two-step method.²³ The one-step method of preparing nanofluids is the production of nanoparticles directly in the base fluid. The second method involves the preparation of dry nanoparticles (first step), and then distribution of the resulted nanoparticles to a liquid base (second step).

Nanofluids are produced based on both the metals^24–26 and the oxides^27–35 nanoparticles. Thanks to bigger and bigger number of methods of obtaining an improved graphene, its availability has increased in recent years, resulting in more frequent use of this material in the production of nanofluids.^36–38 A completely different group of nanofluids are magnetic nanoparticles suspensions, which are being intensively studied recently.^39,40

2 Experimental methods

2.1 MgAl₂O₄ nanopowder

MgAl₂O₄ ceramic nanopowder is commercially available as magnesium–aluminum spinel produced by Baikowski (Annecy, France), ID LOT: 101488. The average size of the nanoparticles is 40 nm, and it was measured with an X-ray diffraction technique (XRD) and confirmed on scanning electron microscope (SEM).

Dry nanoparticles can be used for the production of transparent ceramics^41,42 by hot isostatic pressing (HIP) method. Detailed information on the characteristics of dry nanoparticles and its' applications can be found in ref. 43.

2.2 Sample preparation

The sample was prepared exactly as described in ref. 44. This allows us to compare the results of experiments carried out in this study.

Nanofluids were prepared by two-step method based on the dry nanopowders and diethylene glycol. On an analytical balance, WAS 220/X (Radwag, Radom, Poland), nanopowder was placed and adequately dosed, then diethylene glycol (Chempur, CAS: 111–46–6) was added to give the required concentration. The sample was mixed mechanically for 30 minutes in Genius 3 Vortex (IKA, Staufen, Germany) and sonicated for 200 minutes inside ultrasound wave machine Emmi 60 HC (EMAG, Moerfelden-Walldorf, Germany). All samples were prepared in temperatures below 25 °C.

2.3 Rheo-NMR measurement

Rheo-NMR (Bruker BioSpin, Rheinstetten, Germany) system can use nuclear magnetic resonance (NMR) in studies of rheological properties of fluids. The measurement system was designed and described in details by Callahan.⁴⁵

The Rheo-NMR measurements of MgAl₂O₄-DG nanofluids were performed on a Bruker AV III spectrometer with a 300 MHz wide bore magnet. Spectrometer with a NMR Microscopy accessory, a Micro 2.5 gradient system (25 mT mA⁻¹) and a 25 mm 1H birdcage resonator was used. Rheo-NMR cell was inserted in the magnet bore and rotated using stepper motor gear assembly mounted over magnet bore. The stepper motor was controlled by NMR spectrometer. Cone and plate cell was placed into the 25 mm birdcage resonator, which allowed to get NMR images of the sample material in the measuring cell. The whole provided informations about the changes in chemical structure of a fluid during shearing.

Standard Theo-NMR imaging method allows to get resolved velocity maps of the fluid in the cell.⁴⁶ Considering the fact that the lower plate in measuring geometry is static, velocity distribution in the sample subjected to shear provides us with information on whether the tested substance is Newtonian fluid or not. In Newtonian fluids velocity value depends linearly on the distance from the rotor. The velocity distribution in non-Newtonian fluids is non-linear, because the viscosity depends on the shear rate.

2.4 RheoScope measurement

RheoScope (Thermo Electron Corporation, Karlsruhe, Germany) system is coupled with a rheometer, which enables visual observation of samples during the rheological tests. Microscope lens installed in the bottom plate allows us to measure with 20× magnification.

RheoScope system is mounted on HAAKE MARS 3 rheometer (Thermo Electron Corporation, Karlsruhe, Germany) which enabled to control torques from 0.05 μNm to 200 mNm. Plate–plate measurement geometry (diameter 60 mm) was used.

The study with the use of RheoScope equipment was conducted on two samples with 10 wt% and 20 wt% concentrations of nanoparticles. The study was conducted in a manner analogous to thixotropy measurements that was described in ref. 44.

In the first stage of experiment, viscosity of the samples was measured with increasing shear rate from 1 s⁻¹ to 1000 s⁻¹ in time of 600 s, then constant shear rate 1000 s⁻¹ was used for 600 s. The third stage of measurement was conducted with shear rate decreasing from 1000 s⁻¹ to 1 s⁻¹ during 600 s.

Both Rheo-NMR and RheoScope devices had already been applied to the study of the rheological properties of nanofluids.⁴⁷

3 Results and discussion

3.1 Rheo-NMR measurement

This method allows the observation of chemical changes within the sample subjected to shear. Due to the fact that the tested materials do not have complex chemical composition, this possibility have no use in presented measurements. More interesting subject were the velocities of each individual layer within the sample subjected to shear.

A single study lasted 8 minutes, in which the sample was subjected to shear at a constant rate. Fig. 1 summarises the visualization of the sample measured inside the cell Rheo-NMR. Different linear velocity of the liquid layer is represented by different colour.

Fig. 1 Velocity distribution inside the samples of MgAl₂O₄-DG nanofluids at various shear rates. Different colours represent different velocities. (A) 10 wt% MgAl₂O₄ [small gamma, Greek, dot above] 20 s⁻¹, (B) 10 wt% MgAl₂O₄ [small gamma, Greek, dot above] 100 s⁻¹, (C) 20 wt% MgAl₂O₄ [small gamma, Greek, dot above] 20 s⁻¹, (D) 20 wt% MgAl₂O₄ [small gamma, Greek, dot above] 100 s⁻¹, (E) 30 wt% MgAl₂O₄ [small gamma, Greek, dot above] 20 s⁻¹, (F) 30 wt% MgAl₂O₄ [small gamma, Greek, dot above] 100 s⁻¹.

Based on Fig. 1 it is not clearly specified whether the test fluid is Newtonian or non-Newtonian. The processing of raw data obtained in the measurement process provides much clearer view of the issue. Measurement results can be exported in the form of a matrix in which each pixel of the resulted image is assigned with a value of velocity of fluid flow at a certain point. This matrix contains columns with informations required for drawing a graph of velocity of flow depending on distance from rotor. It is important not choose the column matrix describing the flow near the wall of the measurement geometry, provided that there should be expected effects from the friction between the fluid and the static wall of the container.

Each matrix contained thirty-two columns with measuring points. To be able to compare each individual series of experimental results, the same area inside geometry should be consider. For the purpose of analyzing the results, it was decided to compare each sixth column of the matrix of. Such a choice was dictated by the lack of friction between the liquid and the wall of the measuring chamber impact in this area. Based on these results, plotted according to the speed of the individual layers of the distance from the rotor presented on Fig. 2. Additionally Fig. 2 presented linear function fitting measurements data.

Fig. 2 Velocity distribution of fluid layers inside the measured sample depending on distance from the rotor in Rheo-NMR chamber.

The graph shows that the tested samples can not be considered as Newtonian fluids. While in the case of concentration 10 wt% and 20 wt%, non-linearity of the velocity distribution is not as clear as for the concentration of 30 wt%, where this phenomenon is clearly noticeable.

Based on a Rheo-NMR measurements, MgAl₂O₄-DG nanofluid can be considered as non-Newtonian fluid, confirming all the experimental results presented in the previous papers.⁴⁴

3.2 RheoScope measurement

Experiments with the use of RheoScope module installed on the HAAKE MARS 3 rheometer allowed optical observation of the sample during rheological measurements. The measurement was performed in accordance with the measurement procedure used to determine the thixotropy. The aim was to test whether the differences in the trends of viscosity curves for increasing and decreasing shear rate will cause an optically visible changes within the sample. Samples were examined with two mass concentrations of nanoparticles: 10 wt% and 20 wt%. The measurement results are shown in Fig. 3 and the pictures taken during the study are summarized in Table 1. In both measured samples settings were identical (including the measurement time, the number of measurement points, the intensity of light and the focal length of the microscope). The differences between brightness of pictures of 10 wt% and 20 wt% result from the concentration of nanoparticles in the suspension and thus different degree of transparency of test material.

Fig. 3 Dynamic viscosity of two MgAl₂O₄-DG nanofluids with different mass concentrations of nanoparticles at 15 °C measured with increasing and decreasing shear rates. Arrows present directions of shear rate changes.

Table 1 Photos of the bottom of geometry measurement taken using RheoScope with ×20 magnification lens for different shear rates during the measurements of dynamic viscosity of MgAl₂O₄-DG nanofluid presented in Fig. 3

---

The conducted experiments show that increasing shear rate results with (a) the agglomeration of nanoparticles, or (b) organization of their movements inside the sample. Observed area of the bottom of measuring geometry was about 0.5 mm². In the initial phase of the study there was a small amount of particles agglomerations in the observed area and it sizes were slight. Along with the increasing shear rate a creation or accumulation of aggloerations happened in the monitored area in the sample.

It is clear that the agglomerates are much bigger in the third stage of measurement (when the shear rate decreases). On the basis of presented studies it can be concluded that the shear causes (a) agglomeration of nanoparticles, or (b) stabilization of the sample flow, organization of direction of nanoparticles and agglomerates flow.

Measurements has shown that after shearing samples for a long time with a shear rate of 1000 s⁻¹, MgAl₂O₄-DG nanofluids present a well-known pseudoplastic flow.

Research has shown that the effect on the viscosity curves is caused by the formation of agglomerations of nanoparticles, not by the organized flow inside the sample. This is because flow organize the nanosuspension would result in a decrease of dynamic viscosity, however, presented results of research of dynamic viscosity of the MgAl₂O₄-DG nanofluid show that it increases during long shear. In addition, the explanation for this phenomenon is the fact that MgAl₂O₄ nanoparticles tend to form agglomerates.

It should be noted that the formation of agglomerations of nanoparticles does not change their volume or mass concentration in the suspension. But total area between the nanoparticles and the liquid change. The study shows that it is the value of the area, not the mass or volume concentration that has a decisive influence on the rheological parameters nanofluid.

Duan et al.⁴⁸ described that nanoparticle aggregation affected viscosity in Al₂O₃–water nanofluids. However their researches were focused on aggregation occurring two weeks after the sample preparation. The influence of agglomeration on the rheological properties of nanofluids was also investigated by Anoop et al.,⁴⁹ but they also did not examine the impact of agglomeration formed during the shearing of the sample – the study of rheological properties.

The dependence of the rheological properties of nanofluids on particle size was previously presented by Pastoriza-Gallego et al.²⁰ The radius of the particles is also included in the theoretical model of heat conduction in nanosuspensions.⁵⁰ The influence of particle size on the thermal conductivity has also been investigated by Beck et al.⁵¹ and Kim et al.⁵² However, the data presented in this paper shows that the rotational viscosity measurement disturbs the state of the sample leading to the formation of agglomerates.

4 Conclusions

Paper⁴⁴ presented unexpected form of MgAl₂O₄-DG nanofluids viscosity curves. Paper⁵³ presented that for MgAl₂O₄-DG nanofluids type of geometry used for measuring rheological properties does not affect the unexpected shape of its dynamic viscosity curves. It also presents that this behavior is not caused by any occurrences of uncontrolled electric fields during the experiment.

This paper shows the results of using Rheo-NMR and RheoScope in experiments on viscosity of MgAl₂O₄-DG nanofluids. From the use of the NMR technique it is possible to confirm the non-Newtonian nature of the material. Second device has enabled the optical observation of the samples at the time of the rheological measurements.

It was confirmed that the viscosity curves measured with the increasing shear rate and with decreasing shear rate are affected from the fact that of forming the agglomerates of nanoparticles during the time of shearing. After establishing a stable state within the sample the material exhibits well-known pseudoplastic flow.

Acknowledgements

The authors would like to thank the companies: (a) ThermoFisher Scientific (Karlsruhe, Germany), in particular Dr Fabian Meyer, Dr Fritz Soergel and (b) Bruker BioSpin (Rheistetten, Germany), in particular Dr Volker Lehmann, Dr Thomas Oerther and Dr Dieter Gross for the opportunity to make measurements in their research laboratories.

References

1. R. Saidur, K. Leong and H. Mohammad, Renewable Sustainable Energy Rev., 2011, 15, 1646–1668.
2. R. Taylor, S. Coulombe, T. Otanicar, P. Phelan, A. Gunawan, W. Lv, G. Rosengarten, R. Prasher and H. Tyagi, J. Appl. Phys., 2013, 113, 011301.
3. G. Ramesh and N. Prabhu, Nanoscale Res. Lett., 2011, 6, 334.
4. P. Keblinski, J. A. Eastman and D. G. Cahill, Mater. Today, 2005, 8, 36–44.
5. A. Ijam and R. Saidur, Appl. Therm. Eng., 2012, 32, 76–82.
6. S. Senthilraja, M. Karthikeyan and R. Gangadevi, Nano-Micro Lett., 2011, 2, 306–310.
7. K. Leong, R. Saidur, S. Kazi and A. Mamun, Appl. Therm. Eng., 2010, 30, 2685–2692.
8. D. P. Kulkarni, D. K. Das and R. S. Vajjha, Appl. Energy, 2009, 86, 2566–2573.
9. Y. He, S. Vasiraju and L. Que, RSC Adv., 2014, 4, 2433–2439.
10. T.-P. Teng, Y.-H. Hung, T.-C. Teng, H.-E. Mo and H.-G. Hsu, Appl. Therm. Eng., 2010, 30, 2213–2218.
11. M. J. Pastoriza-Gallego, L. Lugo, J. L. Legido and M. M. Piñeiro, J. Appl. Phys., 2011, 110, 014309.
12. M. J. Pastoriza-Gallego, L. Lugo, D. Cabaleiro, J. L. Legido and M. M. Piñeiro, J. Chem. Thermodyn., 2014, 73, 23–30.
13. C. Kleinstreuer and Y. Feng, Nanoscale Res. Lett., 2011, 6, 229.
14. G. Paul, M. Chopkar, I. Manna and P. Das, Renewable Sustainable Energy Rev., 2010, 14, 1913–1924.
15. D. R. Heine, M. K. Petersen and G. S. Grest, J. Chem. Phys., 2010, 132, 184509.
16. H. Chen, Y. Ding, A. Lapkin and X. Fan, J. Nanopart. Res., 2009, 11, 1513–1520.
17. M. Chadwick, J. Goodwin, B. Vincent, E. Lawson and P. Mills, Colloids Surf., A, 2002, 196, 235–245.
18. M. Pastoriza-Gallego, L. Lugo, J. Legido and M. Piñeiro, Nanoscale Res. Lett., 2011, 6, 560.
19. S.-Q. Zhou, R. Ni and D. Funfschilling, J. Appl. Phys., 2010, 107, 054317.
20. M. Pastoriza-Gallego, C. Casanova, J. Legido and M. M. Piñeiro, Fluid Phase Equilib., 2011, 300, 188–196.
21. M. Hojjat, S. Etemad, R. Bagheri and J. Thibault, Int. Commun. Heat Mass Transfer, 2011, 38, 144–148.
22. M. E. Mackay, T. T. Dao, A. Tuteja, D. L. Ho, B. Van Horn, H.-C. Kim and C. J. Hawker, Nat. Mater., 2003, 2, 762–766.
23. Y. Li, J. Zhou, S. Tung, E. Schneider and S. Xi, Powder Technol., 2009, 196, 89–101.
24. C.-N. Chen, C.-T. Huang, W. J. Tseng and M.-H. Wei, Appl. Surf. Sci., 2010, 257, 650–655.
25. H.-J. Chen and D. Wen, Nanoscale Res. Lett., 2011, 6, 198.
26. H. Zhang, H. Cui, S. Yao, K. Zhang, H. Tao and H. Meng, Nanoscale Res. Lett., 2012, 7, 583.
27. W. J. Tseng and K.-C. Lin, J. Mater. Sci. Eng. A, 2003, 355, 186–192.
28. D. K. Singh, D. K. Pandey, R. R. Yadav and D. Singh, J. Exp. Nanosci., 2013, 8, 731–741.
29. H. Zhu, D. Han, Z. Meng, D. Wu and C. Zhang, Nanoscale Res. Lett., 2011, 6, 181.
30. M. Moosavi, E. K. Goharshadi and A. Youssefi, Int. J. Heat Fluid Flow, 2010, 31, 599–605.
31. X. Wei, H. Zhu, T. Kong and L. Wang, Int. J. Heat Mass Transfer, 2009, 52, 4371–4374.
32. B. Wang, B. Wang, P. Wei, X. Wang and W. Lou, Dalton Trans., 2012, 41, 896–899.
33. S. Chung, J. Leonard, I. Nettleship, J. Lee, Y. Soong, D. Martello and M. Chyu, Powder Technol., 2009, 194, 75–80.
34. C.-Y. Lin, J.-C. Wang and T.-C. Chen, Appl. Energy, 2011, 88, 4527–4533.
35. D. Zhu, X. Li, N. Wang, X. Wang, J. Gao and H. Li, Curr. Appl. Phys., 2009, 9, 131–139.
36. B. Wang, J. Hao and H. Li, Dalton Trans., 2013, 42, 5866–5873.
37. T. T. Baby and S. Ramaprabhu, J. Mater. Chem., 2011, 21, 9702–9709.
38. M. Martin-Gallego, R. Verdejo, M. Khayet, J. M. Ortiz de Zarate, M. Essalhi and M. A. Lopez-Manchado, Nanoscale Res. Lett., 2011, 6, 610.
39. E. Ghasemi, A. Mirhabibi and M. Edrissi, J. Magn. Magn. Mater., 2008, 320, 2635–2639.
40. M. J. Pastoriza-Gallego, M. Pérez-Rodríguez, C. Gracia-Fernándeza and M. M. Piñeiro, Soft Matter, 2013, 9, 11690–11698.
41. J.-G. Li, T. Ikegami, J.-H. Lee and T. Mori, J. Am. Ceram. Soc., 2000, 83, 2866–2868.
42. J. Zhang, T. Lu, X. Chang, N. Wei and W. Xu, J. Phys. D: Appl. Phys., 2009, 42, 052002.
43. G. Żyła, M. Cholewa and A. Witek, Nanoscale Res. Lett., 2012, 7, 412.
44. G. Żyła, M. Cholewa and A. Witek, RSC Adv., 2013, 3, 6429–6434.
45. P. T. Callaghan, Rep. Prog. Phys., 1999, 62, 599.
46. P. T. Callaghan, Curr. Opin. Colloid Interface Sci., 2006, 11, 13–18.
47. G. Żyła, M. Cholewa, A. Witek, J. P. Plog, V. Lehmann, T. Oerther and G. Dieter, J. Nanosci. Nanotechnol., 2012, 12, 8920–8928.
48. F. Duan, D. Kwek and A. Crivoi, Nanoscale Res. Lett., 2011, 6, 248.
49. K. B. Anoop, S. Kabelac, T. Sundararajan and S. K. Das, J. Appl. Phys., 2009, 106, 034909.
50. D. H. Kumar, H. E. Patel, V. R. R. Kumar, T. Sundararajan, T. Pradeep and S. K. Das, Phys. Rev. Lett., 2004, 93, 144301.
51. M. P. Beck, Y. Yuan, P. Warrier and A. S. Teja, J. Nanopart. Res., 2009, 11, 1129–1136.
52. S. Kim, S. Choi and D. Kim, J. Heat Transfer, 2007, 129, 298–307.
53. G. Żyła, J. Grzywa, M. Cholewa and A. Witek, Nanoscale Res. Lett., 2014, 9, 170.

---