Neutron scattering as a tool for industrial research

Neutron scattering as a tool for industrial research

PHYSICA[ ELSEVIER Physica B 213&214 (19951 1 5 N e u t r o n scattering as a tool for industrial research S.K. Sinha*, D.J. Lohse, M.Y. Lin Corporat...

438KB Sizes 0 Downloads 66 Views

PHYSICA[ ELSEVIER

Physica B 213&214 (19951 1 5

N e u t r o n scattering as a tool for industrial research S.K. Sinha*, D.J. Lohse, M.Y. Lin Corporate Research, Exxon Research and Engineering Company, Annandale, NJ 08801, USA

Abstract The unique properties of the neutron make it a potentially valuable tool for industrial research, but for historical reasons, its application in this area has been rather limited. The last few years have seen an increase in the number of industrial laboratories which have availed themselves of the enormous potential of the neutron to characterize novel materials and processes which impact technology. We give a brief overview of some of these applications, with particular reference to the petroleum and chemical industries.

1. Introduction It is no secret that what one might call "basic" or "fundamental" research is under considerably increased pressure and scrutiny by corporations, funding agencies, politicians, and a public less and less willing to write "blank checks" to support such research. One might legitimately argue that such an attitude is short-sighted, but nevertheless, it is a fact which scientists must deal with over the next several years. Neutron-based research has had (and continues to have) an enormous impact on fundamental studies in condensed matter and polymer physics, chemistry, materials science, and biology and has helped provide a basic understanding of the structure and excitation spectra of a variety of forms of condensed matter, ranging from the familiar to the exotic. The majority of papers presented at this conference attest to the important role of neutron scattering in the above fields. However, the potential of neutrons as a unique probe for the advanced characterization of materials and processes of importance to industry is only slowly being recognized, and even today relatively few industrial researchers are familiar with this potential. Thus, intelli*Corresponding author. Present address: Advanced Photon Source, Argonne, IL 60439, USA.

gent exploitation of neutron-beam research in these areas has important implications for the future health and stability of this field of science. While industrial applications of neutron scattering may be found in all branches of industry, we shall concentrate on examples of applications in the petroleum, chemical and engineering industries, which abound with many materials problems for which neutrons are uniquely suited. By its nature, much of this research involves the study of rather disordered and complex materials which are far removed from the highly ordered crystalline materials on which neutron scatterers have traditionally concentrated their efforts.

2. Thermodynamics of polyolefln blends The ability of deuterium labeling to provide contrast between polymers without substantially affecting other physical properties such as miscibility or crystallinity has made neutron scattering quite useful in a number of cases where the chains overlap with each other to a great extent. The first, and perhaps most important, example of this is the determination of the dimensions of polymer molecules in the bulk and in concentrated solutions, but SANS has also provided valuable insight into the nature of blends, block copolymers, and networks.

0921-4526/95/$09.50 ~ 1995 Elsevier Science B.V. All rights reserved SSDI 0 9 2 1 - 4 5 2 6 ( 9 5 1 0 0 0 4 9 - 6

2

S.K. Sinha et al. /Physica B 213&214 (1995) 1-5

Mixtures of several polymers are of increasing scientific and commercial importance. The main reason for this is that in this way a wide range of properties can be obtained to produce a number of useful materials, generally at low cost. Most of these blends are immiscible and form multiphase materials, due to the small entropy of mixing for long-chain molecules. A large number of experimental methods have been developed and applied to blends, mainly to determine the morphology of such multiphase mixtures. Direct data on the miscibility of blends has been difficult to obtain, and most efforts in this area have looked for the boundary between singleand multi-phase behavior. Recently, SANS has been shown to be able to determine the interactions directly, and so has been useful to establish quantitatively the basis for miscibility in many cases. Blends of polyolefins (e.g., polyethylene, polypropylene) are important items of commerce. Around 60 billion pounds of polyolefins are produced each year worldwide corresponding to a global business of ~ $25 billion per year, and a large fraction of these are mixtures of different types of polyolefins. These are widely used in a number of applications, including the automotive, construction, packaging, and electricals markets. Despite this commercial importance, very little work has been done to determine directly the thermodynamics of mixing in such blends. This lack of basic data on the thermodynamics of polyolefin mixtures proceeds from two basic causes: the complications due to crystallization for blends in the solid state, and the similarities in physical properties (density, refractive index) between these materials in the melt. Both factors make it difficult to determine miscibility and interactions at any temperature. This is an unfortunate state of affairs, as one might expect from first principles that many polyolefin blends of commercial interest are near phase transitions as they are commonly used. Developing such an understanding of the miscibility of the polyolefins should have a number of technological implications in terms of how these mixtures are processed to form new materials, and what properties they have. From a scientific point of view, polyolefin blends are also of interest as models for polymer blends in general Mixtures of structurally different species of saturated hydrocarbon polymers, such as polyolefins, represent one of the simplest sorts of blends for study due to their chemical simplicity. Because the polymers are aliphatic hydrocarbons (empirical formula: CH2, in all cases) the change in intermolecular energy with mixing should be relatively small and purely dispersive in origin. Permanent dipole effects and specific interactions are absent. So a more complete understanding of the origins of miscibility in polyolefins should help the understanding of polymer mixtures in general.

The focus of research on the miscibility of polymers is usually the determination of the Flory interaction parameter, ~. The miscibility of the two polymers is largely determined by the value of X, and so it becomes an important object of the research on any blend. This can be determined by SANS within the general framework of the random phase approximation (RPA) [1], which provides a relationship for the structure factor of a two-component, single-phase mixture obeying F l o r y Huggins Staverman theory [ 2 4 ] . Assuming incompressibility, one gets the following expression for the scattering function S(q) of the blend of components 1 and 2: 1

1

1

-+ vlNadplPl(q) vzN2d~zPz(q)

S(q)

2 -~, v

(1)

where Pl(q) and PE(q) are the normalized (Pi(0)= 1) form factors (Debye functions) of the component polymers. If the pure-component single-chain scattering functions P~(q) are known, one can directly obtain values of Z for any single-phase blend at a variety of temperatures and compositions. The scattering functions for a blend of two polyolefins, one labeled with deuterium, at several temperatures are shown in Fig. 1. The increase in scattering intensity at low wave vectors as temperature decreases is due to the increase in Z. Over the last several years in Exxon's Corporate Research Laboratories, we have been engaged in a study of the basic thermodynamics of many such mixtures of polyolefins by SANS, and have obtained such data on over 50 blends involving some 30 different polyolefins [-5-7]. From such a wealth of data we have begun to extract some general conclusions about the thermodynamics of TEMPERATURE

oF le0

DEPENDENCE

X(q) .vs.

H-97-A/D-88

FO~

~50~)

*

100

q

o00°0 o.°°° ..... .....

o

6~e Ikl'c lg~'c 170"C

Oo 60 u o

"~

"o 80.

*

40-

. . .o

°..~,lp ".:

2

I

0.00

0.02

0.04

q(A-')

0.06

0.06

Fig. I. S(q) for a blend of two polyolefins, one labelled with deuterium, at several temperatures (from Ref. [6]).

S.K. Sinha et al. / Phvsica B 213&214 (1995) 1 5

mixing that can apply not only to the polyolefins but to other polymers as well. A major portion of this work has involved random copolymers of ethylene and butene, covering the whole range from polyethylene to polybutene [7]. The interactions between a pair of copolymers, as measured by the Flory interaction parameter, )~, depend not only on the relative difference in butene content between the copolymers but also on the absolute level of branching; the Z values are larger the greater the butene content. This means that the commonly used mean-field theory for copolymer mixing which relies on a single parameter to describe the interaction between ethylene and butene groups cannot explain these data. Rather, we have found that the simplest explanation is to assign a value to each copolymer (at each temperature) which can be used as a solubility parameter in the Hildebrand regular solution scheme to describe the Z parameters seen.

3. Oil recovery

It is estimated that for existing discovered reservoirs of oil, only 30-50% of the oil in the reservoir is recovered by conventional techniques, the remainder staying trapped in the reservoir. It is estimated that ~ 1 trillion barrels of oil remain to be recovered from existing reservoirs, provided it can be done at an economic cost which is favorable relative to the (presently low) global market price of the oil. In Canada, it is estimated that reserves amounting to 2 trillion barrels of oil (greater than the reserves in the Middle East) exist in the tar-sands of Alberta, although it is relatively low-quality and difficult to recover and refine economically at present. Thus, the economic incentives for improved and economic methods of enhanced oil recovery are great. Following the initial methods of oil recovery, relying on natural gas pressure and water or steam flooding, enhanced oil recovery processes generally rely on fluids injected into the reservoir to mobilize the remaining oil [8]. The barriers encountered in such processes are related to the so-called "'sweep efficiency" of the process (which may be drastically decreased by phenomena such as "viscous fingering", due to viscosity differences between the injected fluid and the reservoir fluids) and also to the mobility of the trapped oil, which is decreased by the interfacial tension between the oil and the surrounding fluids (brine or the injected fluid). These problems often necessitate the use of additives such as polymers and surfactants as viscosity modifiers and interfacial tension-reducing agents. Thus, it is important to know the behavior of polymers and surfactants in solution, particularly in the supercritical fluids (such as CO2) often used in oil recovery, and to understand how they affect the rheological properties of the

3

fluid under shear. It is also important to know the adsorption properties of these molecules on the surfaces of the rock (which may be water-wet, oil-wet, or of mixed character) and their structure at the oil/brine interface and how they modify the interracial tension. Finally, it is also important to know the phase behavior of the oil components which are miscible with various injected solvents, and in the case of the "heavy oil" in the tarsands to understand its internal structure and what gives rise to its high viscosity. In each of these problems, neutrons can be an important probe on length-scales ranging from the molecular to the macroscopic. Neutron radiography and neutron microtomography can be used to study the flow and entrapment of various solvents through core samples of reservoir rock. Sufficient contrast may be obtained simply by deuterating one component (thus, changing the scattering cross-section) or by introducing neutron-absorbing nuclei (such as B 1°, Li 6, Cd, Gd, etc.) in solution into one of the components. At the Corporate Research Laboratories of Exxon, in conjunction with NIST, a neutron tomographic capability with ,~ 30 p.m spatial resolution has been developed (the goal being ultimately ~ 1 ~tm spatial resolution) and has been used to study the draining of fluid from a porous rock. Neutron radiography and SANS may also be used to study phase separation in oil/solvent mixtures as a function of concentration, temperature, and pressure, again by the trick of deuterating one component. This may be very helpful in mapping out the phase diagrams for such systems (which are often opaque to light and difficult to study by other techniques). The phase behavior of fluid mixtures inside porous media is obviously of interest for the reasons mentioned above, and again, neutron scattering has been an important tool in studying phenomena such as the phase separation of binary fluids in porous media [9-11], in particular, mixtures which are nominally critical. This problem, namely the effect of confinement or randomness on critical behavior is, of course, an area of scientific interest in its own right. But it may actually be relevant for oil recovery because, as the supercritical solvent flows through the rock, it will pick up in solution a mixture of light and heavy hydrocarbons which will separate into a mixture of two phases. The phase containing the light hydrocarbons will have greater mobility, leaving the heavier phase behind, and the process will repeat itself. Eventually, under certain circumstances, the equilibrium phase moving through the rock will be a mixture of solvent and hydrocarbons at the critical point. Fig. 2 illustrates the SANS scattering data for a near-critical and nominally contrast-matched water/lutidine fluid mixture inside a model porous medium, namely Vycor glass, where the pore diameter is typically ~ 8 n m [10]. The data shows a lutidine-rich adsorbed layer on the internal surface of the pores

4

S.K. Sinha et al. /Physica B 213&214 (1995) 1 5 100

[ t T=85"C

.

.

.

.

.

.

CONVENTIONAL ANII~B~II I ¢ ¢

'

,'no,-m,nt.,

0+01

mi~lJo

.en~o~

0.1

q (A "I)

k,

nemxlJc N~ !o~hmle

'~Bt~f,IW la~,'mnf

Fig. 3. Some examples of micellar surfactant phases in solutions.

Plug

Cqa=uJe

g Tube

a/r o

Fig. 2. (a) SANS spectra taken at IPNS of nominally contrastmatched water/lutidine in porous Vycor glass, taken at several temperatures. Data taken at IPNS (from Ref. [10]). (b) Schematic of interpretation of SANS data in Fig. 2(a) in terms of phase diagram proposed by Liu et al. [from Ref. [11]).

(deduced from the peak in the scattering at ~0.035,~ 1) and "nano-capsules'" of water-rich phase forming inside the pores in the two-phase region (deduced from a Lorentzian squared term in S(q)), in agreement with a theoretical phase diagram for phase separation inside finite cylindrical pores proposed by Liu et al. [12]. Wetting layers, microdomain formation and long-time relaxation and hysteresis effects in the two-phase region appear to be the hallmarks of binary-fluid phase separation inside porous media, as deduced from experiments on other systems as well [13]. Although the pores in reservoir rocks are much larger (typically > 1 ~tm in size), one may expect similar behavior there, an understanding of which is necessary to model and optimize oil-recovery processes. As noted above, microscopic wetting layers (of oil or water) may be deduced from the presence of structure or peaks in the nominally "Porod" region of the SANS patterns from the pore surfaces when the rest of the fluid in the pores is arranged to be contrast-matched with the solid (typically SiO2).

Surfactants and short block copolymers of hydrophilic and hydrophobic chains (which act like surfactants) form a variety of interesting structures in oil/water microemulsions by forming micelles which may be spherical, cylindrical, or sheet-like (Fig. 3). These can then order in phases which have long-range order such as lamellar, hexagonal, or cubic phases or can form disordered bicontinuous phases. Many of these have been elucidated with neutron-scattering experiments [14]. The structure of surfactants or block copolymers at oil/water interfaces is of obvious interest to the field of chemically enhanced oil recovery. Neutron reflectometry has played a unique role in characterizing the density profiles of such systems (surfactants, polymers, and even polymers associated with surfactants) at air/water interfaces, but very little has been done at actual oil/water interfaces, because of the problem of getting the neutron beam in and out through a fluid medium. The in-plane ordering of the molecules at the interfaces is a difficult problem for neutrons because of intensity limitations, and has been mainly studied by synchrotron X-ray grazing incidence diffraction from liquid surfaces [ 15]. The systematics of the conformation and association of polymers and surfactants at such interfaces and how they modify the interface tension are very active current areas of research.

4. Other applications Due to lack of space, in this overview, we shall simply touch very briefly on other important applications of the use of neutrons in research of actual or potential importance to industrial technology. One of the applications of neutron scattering that is rapidly increasing in popularity with industrial organizations is the use of high-resolution

S.K. Sinha et al. / Physica B 213&214 (1995) 1-5

neutron diffraction to measure residual stress in structural components [16]. This technique is by now wellknown and is the subject of several talks at this Conference so we shall not dwell on the technical details of the method here. Neutron reflectometry has been used to study corrosion associated with the formation of oxide layers on metallic films in electrochemical cells [17]. Increased neutron fluxes would make it possible to study off-specular scattering from surfaces and thin films, as is the case with X-rays, and one could study the in-plane morphology of the pits forming in the corrosion process 1-18]. Neutron reflectometry has also been used to study the magnetization density profile at the surfaces of magnetic materials used for recording devices, and the magnetic field penetration into superconductors [19]. SANS studies of polymer melts and solutions under shear are being carried out, which will ultimately impact our understanding of the rheology of non-Newtonian fluids used as viscosity modifiers in many industrial processes. With significantly more powerful neutron sources, it should be possible to study the dynamics of chain conformation changes in real time under actual processing conditions. Zeolites are some of the most important catalytic materials used in the chemical industry, and neutron diffraction has been a very important tool in characterizing the crystal structures of these materials [20]. SANS studies have also recently been very useful in characterizing the nanopore morphology of some of the recently developed novel zeolite materials [21]. SANS has been used to study the morphology and internal surface structure of new nanophase materials used as precursors for high-strength ceramics [22, 23], and even the details of the sintering process as such materials are consolidated. Molecular agents which inhibit the precipitation of wax crystals in diesel fuels, for instance, or the growth of gas hydrate crystals which block gas pipelines, are of considerable industrial importance. In order to better understand how they work, one may turn to SANS and neutron diffraction, which will undoubtedly play important roles in analyzing the nucleation and growth of crystals from solution or melted mixtures of hydrocarbons. We have tried to present a very brief overview of some of the fascinating research areas where neutrons are making or will make important contributions. We must recognize, however, that real industrial problems are often messy and multidimensional, with no well-defined "crank" to turn, as there often is in problems involving well-defined model systems in physics. Thus, often no clean, elegant and well-defined solutions may exist. Neutron scattering may provide only one clue to the puzzle of why a material or an industrial process behaves as it does. However, neutron scattering by its very nature as a basic microscopic probe is always close to the heart of how material systems, even complex ones, behave at the micro-

5

scopic level, and it is to be hoped that out of this generic understanding really innovative ideas for new materials and processes might arise which will be used by industry.

References [1] P.G. deGennes, Scaling Concepts in Polymer Physics (Cornell University Press, Ithaca, 1979). [2] P.J. Flory, J. Chem. Phys. 9 (1941) 660. [3] M.L Huggins, J. Chem. Phys. 9 (1941) 440. [4] A.J. Staverman and J.H. Van Santen, Recl. Tray. Chim. 60 (1941) 76. [5] D.J. Lohse, N.P. Balsara, L.J. Fetters, D.N. Schulz, J.A. Sissano, W.W. Graessley and R. Krishnamoorti, in: New Advances in Polyolefins, ed. T.C. Chung (Plenum, New York, 1993) p. 175. [6] R. Krishnamoorti, W.W. Graessley, N.P. Balsara and D.J. Lohse, J. Chem. Phys. 100 (1994) 3894, 3905. [7] W.W. Graessley, R. Krishnamoorti, N.P. Balsara, L.J. Fetters, D.J. Lohse, D.N. Schulz and J.A. Sissano, Macromolecules 27 (1994) 2574, 3073, 3896. [8] M. Baviere (ed.), Basic Concepts in Enhanced Oil Recovery Processes (Elsevier Applied Science, London, 1991). [9] S.B. Dierker and P. Wiltzius, Phys. Rev. Lett. 58 (1987) 1865. [10] M.Y. Lin, S.K. Sinha, J.M. Drake, X.L. Wu, P. Thiyagarajan and H.B. Stanley, Phys. Rev. Lett. 72 (1994) 2207. [11] J.S. Huang, S.K. Sinha, W.1. Goldburg, J.V. Maher and S.K. Satija, Physica B 136 (1986) 291. [12] A.J. Liu, D.J. Durian, E. Herbolzheimer and S.A. Safran, Phys. Rev. Lett. 65 {1990) 1897; L. Monette, A.J. Liu and G. Grest, Phys. Rev. A 46 (1992) 7664. [13] B.J. Frisken, D.S. Cannell, M.Y. Lin and S.K. Sinha, Phys. Rev. E, to be published (1995). [14] R. Zana (ed.), Surfactant Solutions: New Methods of Investigation, Surfactant Science Series, Vol. 22 (Marcel Dekker, New York, 1987). [15] F. Leveiller, D. Jacquemain, M. Lahav, L. Leiserowitz, M. Deutsch, K. Kjer and J. Als-Nielsen, Science 252 (1991) 1532. [16] J.H. Root, T.M. Holden, J. Schrrder, C.R. Hubbard, S. Spooner, T.A. Dodson and S.A. David, Mat. Sci. Technol. 9 {1993) 754. [17] D.G. Wiesler and C.F. Majkrzak, Physica B 198 (1994) 181. [18] C.A. Melendres, Y.P. Feng, D. Lee and S.K. Sinha, J. Electrochem. Soc. 142 (1995) LI9. [19] C.F. Majkrzak, Physica B 173 (1991) 75; G.P. Felcher et al., Rev. Sci. Instrum. 58 (1987) 609. [20] J.M. Newsam, Materials Sci. Forum 27/28 (1987) 385. [21] C.J. Glinka, J.M. Nicol, G.D. Stucky, E. Ramli, D. Margolese, Q. Huo, J.B. Higgins and M.E. Leonowicz, MRS Proceedings (1994), to be published. [22] A.J. Allen, G.G. Long, H.M. Kerch, S. Krueger, G. Skandan, H. Hahn and J.C. Parker, in: Proc. 8th CIMTEC World Ceramics Congress and Forum on New Materials, Florence, Italy (1994). [23] T. Freltoft, J.K. Kjems and S.K. Sinha, Phys. Rev. B 33 (1986) 269.