CC BY
11Journal of Siberian Federal University. Engineering & Technologies 3 (2012 5) 275-284
УДК 537.9:543.429.23
NMR Imaging Application for Study
of the Polymethylmethacrylate
Colloidal Crystals Infiltration
in the Inverse Opals Preparation Technology
Evgeny V. Morozova,b*, Olga V. Shabanovab and Oleg V. Falaleeva
a Kirensky Institute of physics, Krasnoyarsk Scientific Centre SB RAS, 50 Akademgorodok, Krasnoyarsk, 660036 Russia b Special Designing and Technological Bureau
«Nauka» KSC SB RAS, 53 Mira, Krasnoyarsk, 660049 Russia 1
Received 13.08.2012, received in revised form 20.08.2012, accepted 27.08.2012
NMR Imaging results of polymethylmethacrylate colloidal crystals infiltration are presented. For infiltration study the colloidal silica andfluids with different relation of surface tension and viscosity were utilized. Wetting front has a trend to be steady while propagated into bulk sample as revealed by means of NMR Imaging. As regards ofpowder sample, the wetting front instability was observed. Both the wetting front dynamic and structural properties of samples namely porous media permeability, fractal dimension of porous space and tortuosity were successfully measured. NMR Imaging techniques was demonstrated to be suitable for noninvasive study of infiltration processes occurred in the course of inverse opals preparation.
Keywords: NMR imaging, inverse opals, PMMA colloidal crystals, infiltration, wetting front instability.
Introduction
Last two decades can be marked by vigorous investigation of new challenging nanomaterials -inverse opals. Investigations involved development, synthesis and properties characterization of these materials [1-3]. The inverse opal is well known to be a three-dimensional highly ordered structure consisted of spherical voids surrounded by solid walls. Due to the dielectric constant difference between void space and solid walls medium the inverse opals structure can be regarded as a photonic crystal [4]. Photonic crystals are interesting for both fundamental and applied points of view. There are optical filters, wave guides, different types of sensors, antireflecting coating etc. which were developed on the base of photonic crystals (for instance, [5]).
* Corresponding author E-mail address: [email protected]
1 © Siberian Federal University. All rights reserved
There are some methods to fabricate inverse opals but the most convenient is template approach [3, 6]. This approach involves the crystallization of a colloidal dispersion of monodisperse spheres (the most widely used is polymethylmethacrylate (PMMA)) to form a material with a three-dimensional periodic structure in which 24% of the volume is air (a colloidal crystal or opal). Inverse opal structures can readily be fabricated from colloidal crystal (opal) templates, by filling the voids in the opal with a dielectric material, and then removing the original opal template by calcination (thermal treating at t > 500 °C). In this sequential process the stage of PMMA colloidal crystals infiltration by sol-gel precursor (colloidal SiO2) is the most important. The size of samples, their structure, both cracks and heterogeneities formation is strongly influenced by features of preparation technology. Thus the study and nondestructive monitoring of colloidal crystals infiltration seems to be an urgent problem.
Among the experimental approaches to study transport to porous media there are some advantageous methods based on NMR phenomenon and particularly the NMR Imaging [7]. Nuclear Magnetic Resonance Imaging (NMR Imaging or MRI) is well known noninvasive and sufficiently informative technique first of all in medicine. Apart from medical and biomedical application NMR Imaging nowadays actively becomes integrated into chemical engineering and materials science [8]. So, the polymerization processes, cementation, chemical waves propagation, catalysts materials etc. were successfully studied by MRI [9-11]. Some NMR Imaging works were devoted to study of wetting front propagation into soil (natural porous media) [12-13]. At present there is no any information concerned about MRI application for study the PMMA colloidal crystals infiltration during the inverse opals preparation. The detailed visualization of wetting front structure and dynamic as well as liquid phase distribution inside of colloidal crystal seems to be successfully addressed to NMR Imaging.
NMR imaging techniques
The principles of NMR Imaging method are well described in literature [14]. It is necessary to make mention only on specific techniques employed in the experiments. For image acquisition the spin echo based technique (Multi Slice Multi Echo [15]) was made use of. Basic timing diagram of spin echo pulse sequence is presented in Fig.1. In such type of techniques two RF pulses are applied: 90° (excited) and 180° (generated spin echo signal). In comparison with other imaging techniques the spin echo one possessed by some substantial advantages, first of all this technique allows us to acquire Tb T2 and proton density weighted images.
The dehcribed spin echo pulse sequence ih defined by set of parameters: TE (Time Echo) - time interval between 90° pulse and echo srgnal; TR (Time Repetition) - timeinterval between every applied pulse )rrin (from first exciting 9)0° pulse to second one). A pulse sequence is repeated defined number of times. Tire signal intensity I after 180° ¡pulse can tie expresse d as fol lows [14]:
where T - spin-lattice (longitudinal) relaxation time, T2 - spin-spin (transverse) relaxation time, M0 -initial magnetization (depended from proton density in object under study).
Experimental part
(1)
.. Il 1 A ■
1 rn „ill.
u L
i i m ■ i „
—* m
»_JL_ -JL _ | . ., fl .
- m^ kb. J ..1 At 1 I
fr ^^ _fl_L *
•al i i n
ft ■ *
Fig.1. Timing diagrams of pulse sequences for the spin echo (left) and pulsed gradient spin echo (right) techniques
For self-diffusion measurements ihe pulsed gradient epin echo tecfmc[ue was emplogfd f15] (Fig. 1). in this casr o pulsed gradient is introduced into a spin echo sec-eecee. The signal intensity I can "be expressed as foilows [14]:
I = Mi exp(q bD), at. b = -^GsRa-^ , (2)
wlierr G is the diffusion gradfent strengih, 5 is the durasion of -he diffusion gradient, A "s the time be-toeon the gradient pulsns, it is the gfromagnetic oatio, end D is ^li^ relf-diffus-an coefficieno oor a given substence. Tfereby the diffusion experiment atlowi the -elr-diftusion coefficient D to be extracted front dependence oS-lnf vs. 6.
All imaging egperiment were carried out using NMR microimaging installation based on Bruker AVANCE DPX 200 with 4.7 T magnetic field. probe PH MICRO 2.5 (radio etequency coil dimeter 25 mm)), ft 25 °C.
Samples preparation
Some aspects nf the tample pneparation are iilusirnfed it Ftg.2. Monoditpecso PMMA spheres wifh diemeter 300 nm (^I^iigs,. 22ai.) were ptepoced bythe method on Schrodeis et al. t6, 16. PMMA collo^dit^ suspensions wece left in open glass ampoulet (~20 nrm) with hemispheiical bottom at 25 °C. After 2-3 weeks, water in the coUoidaf suspenses evaporated to leave wcll-ordered PMMA photonig crysials of ihe round shape. Then the colloidal crystals were left to dry in air at 25 °C for a further i-2 weeks. After drying the large.sized pieces (about 7e7xi0 mm. were eut out and iaken doh the first experiment (bulk sample, Fig. 2b). Residual part of colloidal cryctrlc wits siightly rrushed (rccording to technology practice) to obtain a powder for second experiment (powder sample, Fig.2c). Bulk sample was then placed into a glass tube with the infiltrating liquid. Noncompacted powder sample was filled up to glass pipe with open ends in order to provide air free yield. The glass pipe with powder sample was then mounted into MRI probe and infiltration occurred top-down.
The coAoidol SKi:, obtained by olkoli silicates acidic reaehing and set of liquifo possessed by different values of both surface tension and viscosity (Table 1) were utilized for infiltration experiment.
Table 1. Parameters of surface tension a and viscosity ^ for different types of liquid media utilized for PMMA colloidal crystalinfiltratkm.
H2O Glycerol Isobutyl alcohol Ethanol Hexane
a (N/m) 0.013 0.065 0.023 0.023 0.0181
H, 10-4 (Pa-c) 8.9 94-50 39.5 10.96 2.94
abc
Fig.2. Electron microphotography of PMMA colloidal spheres deposited on aluminum substrate (a) and bulk sample (b), powder sample (c) photos
For PMMA colloidal crystal infiltration by SiO2 sol the dwell time Td has a decisive importance. Dwell time is defined by period between of SiO2 sol preparation and the infiltration of sample being started. It is concerned with intensive sol-gel transfer processes occurred in the colloidal silica (particularly at the very first time after preparing) to result in spatial gel structures growth accompanied by substantial viscosity increasing.
Results and discussion
Bulk sample infiltration
Because the colloidal silica is commonly used for inverse opals preparation, that is the SiO2 precursor which was chosen for PMMA colloidal crystals infiltration. Besides of colloidal SiO2 the liquid media listed in Table 1 and liquid glass as well were utilized additionally for infiltration experiments. It enables to reveal the more suitable media for comparative study of wetting front propagation dynamic. So, the preliminary results have shown that the glycerol is the most convenient liquid (large value of viscosity is provided), due to the optimal infiltration time accompanied with sample structural integrity retention; hexane is imbibed too fast to be detected by imaging technique, and sample infiltration by ethanol (as well as by water) leads to violent cracks formation with subsequent sample fracturing (as can be seen in Fig.3a).
The most informative results were obtained in case of glycerol infiltration. In particular, the internal defects in form of microcracks were also successfully revealed in different samples by means of MRI. These microcracks lead to wetting front distortion in appropriated parts of sample as it is shown in Fig.3b,c. As soon as the liquid media appears inside of microcrack it becomes a secondary source of wetting front propagation.
¿J&-
abc
Fig.:?. Images of PMMA bulk; samples transverse slice in the course of infiltration by water (a) and glycerol (b, c). Image parameters: TR/TE 160 (a), 68 (b), 48/ (c); slice thickness 0.5 mm (a, c) and 0.4 mm (b)
Fig.4. Image of sample transverse slice in the course of glycerol infiltration (left), cross section time-spaced profiles (right). Image parameters: TR/TE 48; slice thickness 0.5 mm
If there is no any internal microcrack-scale defect, the wetting front has a trend to be steady over the time of whole process occurred. For a high quality sample the NMR image of sample transverse slice in process of glycerol infiltration is presented in Fig.4. For this process the cross section timespace d profiles (I/It vs. L) were plotted. The i ntensity I0 corresponds to NMR signal Wrom 1 H located outsihe or porous media while I be)ng variable cosrespond to signal intensity inside of sample (L corresponds to wetting front lengths). It is worth noticing that the normalized signal intensity I/I0 fnside 01s sample is sebstantially less as compared with those co/responded to external part. It is caused try (firstly) absolute proton density decreasing (1) while wetting Oront having moved from interface into deep og sample rnd (secondly.) surface relaxation effects. The wetting front interface keeps being sharp with wetted part intensity having the same value over the all time of process.
Fresh prepared SiO2 sol in the most cases of infiltration give rise to bulk PMMA sample fracturing and there is no fracturing when utilized SiO2 sol with Td more than 20 min. For this reason the SiO2 sols set with equal concentration but different dwell times 20 min, 90 min, 180 min were used for study. In all cases if the sample did not have a crack the wetting front was a steady. The NMR image of sample
Fig.5. Image of sample transverse slice in the course of SiO2 sol (Td 180 min) infiltration (left), cross section time-spaced profiles (right). Image parameters: TR/TE 48; slice thickness 0.5 mm
transverse slice in process of SiO2 sol infiltration with Td 180 min is presented in Fig.5; the cross section time-spaced profiles also were plotted. The sharp form of wetting front interface can be found out in the all NMR images, making it possible to measure the dynamic of front propagation.
Infiltration is modeled using a variety of metho ds, depending o n scale and scope of the problem [17]. Among of them the Green and Ampt approach accounts for a numben of variables and is widely usad gos diSferent applications. This approach was to recognfze that infiltration is driven by capillary and gravitational forces but hindered by fluid miscosity. It is assumed:
- the hydraulic gradient to be linearly distributed between sample surface and position of the wetting front,
- the medium behind the wetting front to be fully saturated,
- the medium ahead of the wetting front to be at its initial water content.
Based on this approach Kao and Hunf presented a modfl for wetting front propagation in porous media [18]. Horizontal wetting front lengrhs weoe predicted to propagate as the square root of time as:
1
L = 2 k 412, (3)
where L - distcnce from the source to the wetting front, a - fluid-air interfacial tension, ^ - fluid viscosity, k - porous media permeability, B - a dimensionless geometric factor determined to be 0.5. In Fig.6 the watting front length, L vs. scjuare root of time t1/2 are jmesented for some fluids which were used. For colloidal silica the dafa presented for different dwell t/mes. In terms of data from Table r aad Fig.6 the bulk PMMA colloidal crystal petmeability was calculated according to (3) to be d = 0.3-1.7 mD (~f0-15 m2). In conclusion the porous media described con be referred to as weakly permeable (corresponded fo a range of 1-10 mD).
Powder sample infiltration
For infiltration in powder sample only glycerol and colloidal silica were utilized. The NMR images of sample longitudinal slice in process of infiltration are presented in Fig.7 (images are pointed out for
Fig.6. Time dependence of wetting front length for different fluid (left) and colloidal silica with different dwell times (right)
a b c d
Fig.7. Images of sample longitudinal slice in the course of glycerol infiltration, 15 min (a), 35 min (b) and SiO2 sol (xd 180 min), 5 min (c), 10 min (d). Image parameters: TR/TE 24 (a, b), 48 (c, d); slice thickness 0.6 mm (a, b), 0.5 mm (c, d)
different time after infiltration being started). The inhomogeneous pattern of powder can tie observed: there are areas in the image with relatively large-scale size grains as well as with small ones. The first mentioned coeresponds to low signal intensity while the second one; corresponds ta high signal intensity. The topmost part in tine images corresponds tofluid-powder inferface. The same nature of infiltration was observed when used both glycerol (Fig.7a,b) and colloidal silica (Fto.7c,d). Since the SiO2 sols with dwell time Td < 2 h practically does not differ on their properties, it was decided to use sois with Td 1 h, 3 h ond 5 h. It was not found out the alteration of infiltration law against of dwell time, but as Td increased the rate of wetting front displacement appropriately decreased.
By means o( NMR Imaging tire wetting froyt instability wars dearly visualized in procers of PMMA powder infiliration. It is welliknown to Oe o "fingeging" (likr the "viscous fingef formatron" in Hele-rhaw cell), whifh usually having bten observed when studied of transport to porous media [19]. It doer not allow the Gteen and Ampt approacO to beapplied for wettmg fronr propagation dynamic. Howeyer the fact of wetting front instability could testify to Crastal oature of porous space. The area S of figure is known correlate to its own perimeter L as follows [20]:
S x Ld,
where n is a topological dimension (tc tie 2 for two-dimensional space). For fractals d would be a fractional number ond always will vary from 2. en Fig.8 the dependencies of lnS vs. lnL are presented
Fig.8. Dependencies of lnS vs. lnL for samples infiltrated by glycerol (left) and colloidal silica with different Td (right)
Fig.9. Dependencies of D/D0 vs. A for experiments with different SiO4 sol dwell times
(in this case L corresponds to contour length of wetted part in image). One can see that parameter d ~ 1.3 and it does not depend on fluid utilized, being a property of powder itself.
Pore space oh powder sample can also be characterized to measure of water self-diffusion cnefficient behavior after infiltration having been finished. The self-diffuston coefficient D in common case is knowndepends on At i.e. D=f(A). It was determined tfrt [7]:
DiA)= 1s
D„
4 S 9-jn V
jdd+o(d0 a )
(4)
where Df is a wfter self-diffusiof croeufiiciiieiil; outside of poeous meOia (D0 = 2.1-10-3 mm2/s for pure watef) »S'iii; fi surface afea of porous cpf ce, V is a total po res volumn. At A—rco (4) can be written:
lim
Ata
Ddh
D„
d_
T
(5)
where T is fortuosity ey porous space. Thuo the asympeot— behavior cc;ft D/D0 vs. a— allows T to be evaluated from f5). Self-diffusion coefficientc measu-ong were caeried out as soon as powder sample infilteation by SiO2 sol was finished. In Fig.9) these dependencies aire peesented for different dwell times. Id tesult the dependencies prmvedto be asymptotic and they does not be affected to dwell time,
being also a property of powder itself. For PMMA powder sample the tortuosity T was found out to be 2.27.
Conclusions
In this work the NMR Imaging method was successfully applied to study PMMA colloidal crystal infiltration in process of inverse opals preparation. Some facilities of this technique were tested for noninvasive visualization of liquid phase interaction with porous media. Wetting front propagation dynamic, wetting front instability, structural properties of samples were studied. So, bulk sample permeability was measured to be 0.3-1.7 mD, for powder sample both the fractal dimension and tortuosity were measured to be 1.3 and 2.27 appropriately. Such information can assists to adjust some technological parameters for colloidal crystals infiltration. MRI as a tool proved to be helpful for monitoring of essential stage of inverse opals preparation. It is worth noticing that the NMR Imaging possibility in this field of research now is hardly exhausted and further activity is looking forward.
References
[1] Cao G., Brinker C. J. // World Scientific. 2006. Vol. 1. P. 628.
[2] Cai M, ZongR., Li B., Zhou J. // J. Mat. Sci. Lett. 2003. Vol. 22. P. 1295-1297.
[3] Waterhouse G. I. N., WaterlandM. R. // Polyhedron. 2007. Vol. 26. P. 356-368.
[4] John S. // Phys. Rev. Lett. 1987. Vol. 58. P. 2486-2489.
[5] Park S. H, Xia Y. // Langmuir. 1999. Vol. 15. P. 266-273.
[6] Schroden R. C, Al-Daous M, Blanford C. F, Stein A. // Chem. Mat. 2002. Vol. 14. P. 3305-3315.
[7] Коптюг И. В., Сагдеев Р. // Успехи химии. 2002. Вып. 71. С. 899-949.
[8] Stapf S., Han H. S. I. NMR Imaging in Chemical Engineering. John Wiley & Sons, 2006. P. 560.
[9] Коптюг И. В., Сагдеев Р. // Успехи химии. 2002. Вып. 71. С. 672-699.
[10] Коптюг И. В., Сагдеев Р. // Успехи химии. 2003. Вып. 72. С. 183-212.
[11] Коптюг И. В., Лизова Ф. Ф., Ковтунов К. В. и др. // Успехи химии. 2007. Вып. 76. С. 628-645.
[12] Amin M. H. G., Richards K. S., Chorley R. J. // Magn. Reson. Imaging. 1996. Vol. 14. P. 879-882.
[13] Amin M. H. G., Hall L. D., Chorley R. J. // Magn. Reson. Imaging. 1994. Vol. 12. P. 319-321.
[14] Callaghan P. Principles of nuclear magnetic resonance microscopic. Clarendon Press, Oxford,
1991.
[15] Bernstein M. A., King K. F., Zhou X. J. // Elsevier Academic Press, 2004. P. 1040.
[16] Schroden R. C., Balakrishan N. // University of Minnesota, 2001. P. 32.
[17] Cueto-Felgueroso L., JuanesR. // Water resources research. 2009. Vol. 45. P. 1-23.
[18] Kao C. S., Hunt J. R. // Journal of environmental engineering. 2001. Vol. 127. N. 2. P. 162-169.
[19] Posadas A., Quiros R., Tannus A., Crestana S., Vaz C. M. // Nonlinear processes in geophysics. 2009. Vol. 16. P. 159-168.
[20] Feder J. Fractals. Plenum press, New York. 1988.
Применение ЯМР-томографии для исследования пропитки коллоидных кристаллов полиметилметакрилата в процессе получения инверсных опалов
Е.В. Морозов"'5, О.В. Шабанова б, О.В. Фалалеева
а Институт физики им. Киренского СО РАН Россия 660036, Красноярск, Академгородок, 50 б Специальное конструкторско-технологическое бюро
«Наука» КНЦ СО РАН Россия 660049, Красноярск, Мира, 53
В работе представлены результаты ЯМР томографического исследования пропитки коллоидных кристаллов полиметилметакрилата золем SiO2 и жидкостями с различными значениями поверхностного натяжения и динамической вязкости. С помощью данного метода выявлен устойчивый характер фронта пропитывания в крупнокристаллических образцах, измерена его динамика и рассчитана из экспериментальных данных проницаемость пористой среды. Обнаружена неустойчивость фронта при пропитывании порошкообразных образцов, рассчитана фрактальная размерность и извилистость порового пространства. Продемонстрирован высокий потенциал метода для неразрушающего контроля процессов пропитки ПММА при получении инверсных опалов.
Ключевые слова: ЯМР томография, инверсные опалы, коллоидные кристаллы ПММА, пропитывание, неустойчивость фронта пропитки.
