Langmuir 2007, 23, 5477-5484
5477
Understanding the Dissolution of Zeolites Ryan L. Hartman and H. Scott Fogler* Department of Chemical Engineering, UniVersity of Michigan, 2300 Hayward Street, Ann Arbor, Michigan 48109-2136 ReceiVed December 21, 2006. In Final Form: March 2, 2007 Scientific knowledge of how zeolites, a unique classification of microporous aluminosilicates, undergo dissolution in aqueous hydrochloric acid solutions is limited. Understanding the dissolution of zeolites is fundamental to a number of processes occurring in nature and throughout industry. To better understand the dissolution process, experiments were carried out establishing that the Si-to-Al ratio controls zeolite framework dissolution, by which the selective removal of aluminum constrains the removal of silicon. Stoichiometric dissolution is observed for Type 4A zeolite in HCl where the Si-to-Al ratio is equal to 1.0. Framework silicon dissolves completely during Type 4A dissolution and is followed by silicate precipitation. However, for the zeolite analcime which has a Si-to-Al ratio of 2.0 dissolves non-stoichiometrically as the selective removal of aluminum results in partially dissolved silicate particles followed by silicate precipitation. In Type Y zeolite, exhibiting a Si-to-Al ratio of 3.0, there is insufficient aluminum to weaken the structure and cause silicon to dissolve in HCl. Thus, little or no precipitation is observed, and amorphous undissolvable silicate particles remain intact. The initial dissolution rates of Type Y and 4A zeolites demonstrate that dissolution is constrained by the number of available reaction sites, and a selective removal rate parameter is applied to delineate the mechanism of particle dissolution by demonstrating the kinetic influence of the Si-to-Al ratio. Zeolite framework models are constructed and used to undergird the basic dissolution mechanism. The framework models, scanning electron micrographs of partially dissolved crystals, and experimentally measured dissolution rates all demonstrate that a zeolite’s Si-to-Al framework ratio plays a universal role in the dissolution mechanism, independent of framework type. Consequently, the unique mechanism of zeolite dissolution has general implications on how petroleum reservoir stimulation treatments should be designed.
Introduction Understanding the science of material degradation is as important as knowledge of the basic interactions that bring about crystal growth. Dissolution reactions are a class of solid-liquid interactions central to a variety of processes occurring in nature1 and in the petrochemical, pharmaceutical,2 and electronics industries. The elucidation of reaction kinetics and the transport phenomena that govern dissolution processes facilitates a general understanding of how molecular building blocks are disassembled. A unique classification of solids, microporous materials are the focus of research work in biological systems,3,4 inorganic synthesis,5-10 and heterogeneous catalysis and adsorption processes.11 Zeolites are a class of microporous aluminosilicates defined by structure and composition.12 While a number of investigations having been undertaken on the synthesis of microporous materials,3-10,13 few have studied the dissolution * To whom correspondence should be addressed. E-mail: sfogler@ umich.edu. Tel: 1-734-763-1361. Fax: 1-734-763-0459. (1) Lasaga, A. C.; Luttge, A. Science 2001, 291, 2400. (2) Pekarek, K. J.; Jacob, J. S.; Mathiowitz, E. Nature 1994, 367, 258. (3) Cha, J. N.; Stucky, G. D.; Morse, D. E.; Deming, T. J. Nature 2000, 403, 289. (4) Cha, J. N.; Shimizu, K.; Zhou, Y.; Christiansen, S. C.; Chmelka, B. F.; Stucky, G. D.; Morse, D. E. P. Natl. Acad. Sci. U.S.A. 1999, 96, 361. (5) Bu, X. H.; Feng, P. Y.; Stucky, G. D. Science 1997, 278, 2080. (6) Bu, X. H.; Feng, P. Y.; Gier, T. E.; Zhao, D. Y.; Stucky, G. D. J. Am. Chem. Soc. 1998, 120, 13389. (7) Gier, T. E.; Bu, X. H.; Feng, P. Y.; Stucky, G. D. Nature 1998, 395, 154. (8) Davis, M. E.; Lobo, R. F. Chem. Mater. 1992, 4, 756-768. (9) Zhu, G.; Qiu, S.; Yu, J.; Sakamoto, Y.; Xiao, F.; Xu, R.; Terasaki, O. Chem. Mater. 1998, 10, 1483-1486. (10) Barton, T. J.; Bull, L. M.; Klemperer, W. G.; Loy, D. A.; McEnaney, B.; Misono, M.; Monson, P. A.; Pez, G.; Scherer, G. W.; Vartuli, J. C.; Yaghi, O. M. Chem. Mater. 1999, 11, 2633-2656. (11) Masel, R. I. Principles of Adsorption and Reaction on Solid Surfaces; John Wiley and Sons, Inc.: New York, 1996. (12) Breck, D. W. Zeolite Molecular SieVes: Structure, Chemistry, and Use; John Wiley and Sons, Inc.: New York, 1984.
of zeolitessa scientifically challenging problem that is of fundamental importance to a number of environmental processes including wastewater stream treatment,14-16 radioactive waste immobilization,17 and the acid stimulation of petroleum reservoirs.18,19 Acid stimulation is a technique commonly used to enhance reservoir productivity by injecting acids into oil-bearing formations to dissolve minerals comprising the porous medium. Naturally occurring zeolites generally exist within mineral formations indigenous to volcanic activity throughout the world.20 These zeolites, such as analcime, chabazite, heulandite, natrolite, and stilbite crystallize in the pore space and typically comprise less than 15 wt % of a sandstone reservoir formation. Zeolite minerals are particularly troublesome during stimulation of the well reservoirs, especially in the Gulf of Mexico, because of the precipitation of silicate and zeolite dissolution products. The precipitation and migration of dissolution products adds resistance to flow, and thereby reduces the production of oil or gas.19,21,22 Early experimental observations by Murata23 established a general rule for aluminosilicate dissolution and precipitation, including several zeolites, in hydrochloric acid. Those aluminosilicates exhibiting a Si-to-Al framework ratio less than 1.5 dissolve subsequently resulting in the formation of a gel, and (13) Kerr, G. T. J. Phys. Chem. 1968, 72, 2594. (14) Filippidis, A.; Godelitsas, A.; Charistos, D.; Misaelides, P.; KassoliFournaraki, A. Appl. Clay Sci. 1996, 11, 199. (15) Carland, R. M.; Aplan, F. F. ACS Symp. Ser. 1988, 368, 292. (16) Barrer, R. M.; Makki, M. B. Can. J. Chem. 1964, 42, 1481. (17) Ragnarsdottir, K. V. Geochim. Cosmochim. Acta 1993, 57, 2439. (18) Hartman, R. L.; Fogler, H. S. Ind. Eng. Chem. Res. 2005, 44, 7738. (19) Underdown, D. R.; Hickey, J. J.; Kalra, S. K. Proceedings of 65th Annual SPE Technical Conference and Exhibition: New Orleans, LA, 1990. (20) Dyer, A. An Introduction to Zeolite Molecular SieVes; John Wiley and Sons, Inc.: New York, 1988. (21) Rege, S. D.; Fogler, H. S. AIChE J. 1989, 35, 1177-1185. (22) Vaidya, R. N.; Fogler, H. S. Colloids Surf. 1990, 50, 215-229. (23) Murata, K. J. Am. Mineral. 1943, 28, 545.
10.1021/la063699g CCC: $37.00 © 2007 American Chemical Society Published on Web 04/13/2007
5478 Langmuir, Vol. 23, No. 10, 2007
Hartman and Fogler
Figure 1. Aqueous Si and Al concentrations of dissolving (a) Type 4A and (b) Type Y zeolites in 8.0 M hydrochloric acid. (c) Reduced time scale and Si concentration for Type Y dissolution.
those greater than 1.5 form a non-gel precipitate. Here, we can define the Si-to-Al framework ratio within an undissolved zeolite as
ν)
mole Si mole Al
(1)
The nature of dissolution, stoichiometric or non-stoichiometric, was not evident as visual observations without any measurement of aqueous silicon and aluminum concentrations were used as the basis of the general rule. More recently, others14,17,24 have reported that a zeolite’s Si-to-Al framework ratio influences dissolution in mildly acidic to alkaline solutions. Flippidis et al.14 observed that zeolites with low silica content, natrolite and thomsonite, were more reactive in distilled water than high-silica zeolites heulandite and stilbite. Ragnarsdottir17 demonstrated empirically that the reaction kinetics of various aluminosilicates, as well as the zeolite heulandite, are correlated to the Si-to-Al ratio for pH ranging from 2 to 12. Further, Cizmek et al.24 reported that the specific reaction rates of synthetic zeolites silicalite-1 and ZSM-5 decrease with decreasing Si-to-Al ratio in sodium hydroxide solutions. (24) Cizmek, A.; Subotic, B.; Aiello, R.; Crea, F.; Nastro, A.; Tuoto, C. Microporous Mater. 1995, 4, 159.
These observations show that a zeolite’s Si-to-Al framework ratio plays a universal role in the dissolution mechanism, independent of framework type. However, what role the Si-to-Al ratio takes on during zeolite dissolution is not fully understood, especially in acids with pH values used in reservoir stimulation (e.g., pH less than 1.0). Hartman and Fogler18 demonstrated that the mechanism of analcime, Type A, and Type Y dissolution by excess hydrogen ions is limited by the surface reaction under initial rate conditions. At high H+ concentrations, the initial dissolution rate, -ri*, is constrained by the number of available adsorption sites. The initial rate law derived assuming an adsorption followed by surface reaction follows a LangmuirHinschelwood type, analogous to the Michaelis-Menten equation,18 and is given as
-ri* )
Vmax[H + ] Km + [H +]
(2)
Where Vmax is the maximum dissolution rate, Km is the MichaelisMenten constant, and [H+] is the hydrogen ion concentration.18 Fundamental knowledge of the kinetic implications of the Sito-Al ratio on analcime dissolution was ascertained by recasting the dissolution rate in terms of a selective removal rate parameter, γ.18 When there is no precipitation, the ratio of the measured Si
Understanding the Dissolution of Zeolites
Langmuir, Vol. 23, No. 10, 2007 5479
Figure 2. Scanning electron micrographs of (a) initial unreacted Type 4A, (b) partially dissolved Type 4A particles recovered after 25 s of dissolution, (c) partially dissolved analcime recovered after 10 min, and (d) undissolvable analcime particles recovered after 100 min (modified from Hartman and Fogler25). Scanning electron micrographs of Type Y particles (e) before and (f) after treatment for 300 min. Transmission electron micrographs of Type Y particles (g) before and (h) after treatment for 300 min. Courtesy of Professor Abhaya Datye and Mangesh Bore at the University of New Mexico.
dissolution rate to the stoichiometric Si dissolution rate can be expressed by18
γ)
rSi measured Si dissolution rate ) ν0rAl stoichiometric Si dissolution rate
(
)
(3)
Fogler25
The parameter γ was later applied by Hartman and to quantify the influence of the Si-to-Al ratio throughout complete dissolution of the zeolite analcime. The dissolution of silicate species from the analcime framework was shown to be controlled by the selective removal of aluminum.25 Consequently, the dissolution of analcime by H+ leads to partially disintegrated silicate framework and a distribution of dissolved silicate species.25 The research work presented in the current investigation builds on previous zeolite dissolution science by developing a general understanding of how the dissolution kinetics are controlled by the Si-to-Al ratio and its implications on the nature of zeolite dissolution products. Here, we examine three fundamentally different and well-documented zeolites: Type 4A (ν0 (25) Hartman, R. L.; Fogler, H. S. Langmuir 2006, 22, 11163-11170.
) 1.0) which dissolves completely, analcime (ν0 ) 2.0) which undergoes partial dissolution, and Type Y (ν0 ≈ 3.0) which undergoes little dissolution. Structural characterization of each of these zeolites has been published elsewhere.12 Materials and Methods Zeolite Source and Reagents. Ultrapure Type 4A and Na-Type Y zeolites were obtained from Zeolyst International. All zeolite crystals appeared to range from colorless to white. Structural composition of each zeolite was determined in a previous investigation.18 Reactant solutions were prepared with deionized water and trace-metal grade hydrochloric and hydrofluoric acids. Dissolution and Precipitation Experiments. Zeolite dissolution and precipitation experiments were carried out in hydrochloric acid concentrations ranging from 1.0 to 8.0 M using a batch reactor described previously.18 Each zeolite, containing approximately the same number of moles of Si, was dissolved in 300 mL of acid solution at a temperature of 5.4 °C and a stirring rate of ∼500 rpm. In a previous investigation,26 experiments carried out at (26) Hartman, R. L. Ph.D. Dissertation, University of Michigan, Ann Arbor, MI, 2006.
5480 Langmuir, Vol. 23, No. 10, 2007
Figure 3. X-ray diffraction patterns of Type Y (a) before and after treatment for 300 min in (b) 1.0 M HCl and (c) 4.0 M HCl.
Figure 4. Si-to-Al framework ratio of undissolved analcime, Type 4A, and Type Y particles. temperature conditions above 15.0 °C resulted in high dissolution and precipitation rates. Thus, experiments were carried out at 5.4 °C in order to (1) accurately measure the dissolution rates and precipitation phenomena for short reaction times and (2) ensure a reaction-rate-limiting dissolution process rather than a diffusionlimited process.26 Samples were obtained using micropipettes and drawn through membrane filters (dpore ) 0.2 µm) at short time intervals. Filtration of the suspended particles stopped the dissolution reaction. Hence, filtrate samples were representative of dissolved mineral concentrations as a function of reaction time. The samples were immediately analyzed for dissolved Al, Si, and Na using inductively coupled plasma (ICP) spectrophotometry. The recovered filter cake was analyzed using a scanning electron microscope (SEM) and subsequently redissolved in a hydrochloric-hydrofluoric acid mixture and analyzed. Structural composition was determined via analysis with an ICP.
Results and Discussion Zeolite Dissolution and Precipitation. In the first series of experiments, Type 4A and Y zeolites were dissolved under conditions equivalent to those reported by Hartman and Fogler25 for the dissolution of analcime in 8.0 M hydrochloric acid. Type 4A zeolite, exhibiting a Si-to-Al framework ratio, ν0, of 1.06,
Hartman and Fogler
dissolved virtually instantaneously and stoichiometrically upon treatment with hydrochloric acid, as can be seen in Figure 1a. As is evident in Figure 1a, after the Si concentration reached and stayed on a plateau of 240 mmol/dm3 for 50 min, it started to precipitate as a gel. This precipitation is analogous to those results reported for analcime dissolution25 and consistent with the general rule described by Murata.23 However, Type Y which exhibits a Si-to-Al ratio of 2.56, dissolved non-stoichiometrically as shown in Figure 1b. It is readily seen in this figure that Al was selectively removed while the Si concentration remained negligible, confirming the existence of undissolvable silicate particles, as shown in the SEM micrographs in Figure 2. One observes in Figure 1c that expanding the time scale and Si concentration reveals that the Si concentration increases, reaches a maximum, and subsequently decays, which follows the general trend reported for Type 4A and analcime dissolution.25 Moreover, the Si and Al concentrations approach a corresponding maximum and plateau simultaneously, elucidating that the removal of Si is facilitated by Al dissolution. The dissolving zeolite particles were filtered and collected before and after acid treatments and then analyzed using a SEM. The crystals were also dissolved in hydrochloric-hydrofluoric acid mixtures and analyzed using ICP spectrophotometry in order to measure values of the Si-to-Al ratio of the partially dissolved zeolite particles. The scanning electron micrographs of Type 4A and Y particles demonstrate considerably different dissolution phenomenon, as can be seen in Figure 2. Before treatment in acid, Type 4A and Type Y crystals appeared to be unreacted, as shown in Figure 2a and e, and the Si-to-Al ratios within the particles, ν, were approximately the initial stoichiometric values of 1.06 and 2.56, respectively. Novel investigations by Lasaga and Luttge,1 Rufe and Hochella,27 and Brown28 demonstrated that non-microporous mineral dissolution leads to etched pits, resulting from natural dislocations or nucleation sites on mineral surfaces. One observes in Figure 2b that Type 4A crystals filtered after 25 s of treatment in acid exhibit similar attack on particle surfaces and is consistent with SEM images of analcime previously reported by Hartman and Fogler25 and shown in Figure 2c. Nevertheless, the Type 4A zeolite dissolved completely and stoichiometrically succeeded by silicate precipitation as shown (Figure 1a), and analysis of dissolving analcime particles for extended periods in hydrochloric acid revealed a unique mechanism by which the particles appeared to break up and remain inert to subsequent acid treatments,25 as can be seen in Figure 2d. As is evident in Figure 2e and f, no apparent change in Type Y crystal structure is observed in the SEM micrographs, which undergirds the general observations of Figure 1b. However, analysis with a transmission electron microscope (TEM) to further investigate Type Y structural change revealed a very interesting result. Analysis of TEM micrographs of Type Y particles before and after acid treatment (Figure 2g and h) demonstrated that the selective removal of Al transforms the Type Y particles from crystalline to amorphous and is in good agreement with the work of others.13,29,30 These observations are supported by the X-ray diffraction patterns shown in Figure 3, which confirm that Type Y particles are amorphous after treatment in acid. More than one mechanism must be considered when analyzing the experimental results presented here. Precipitated silicate could, in theory, deposit on Type Y particles forming an amorphous (27) Rufe, E.; Hochella, M. F. Science 1999, 285, 874. (28) Brown, G. E. Science 2001, 294, 67. (29) Ohsuna, T.; Terasaki, O.; Watanabe, D.; Anderson, M. W.; Carr, S. W. Chem. Mater. 1994, 6, 2201. (30) Kawai, T.; Tsutsumi, K. Adsorption 1998, 4, 225.
Understanding the Dissolution of Zeolites
Figure 5. Initial dissolution rates of Type 4A and Y particles for different values of ν, the Si-to-Al ratio.
Langmuir, Vol. 23, No. 10, 2007 5481
Figure 7. Experimentally measured and predicted aluminum concentration for Type Y dissolving in 4.0 M HCl.
Figure 8. A cartoon of repeating two-dimensional (a) Type A and (b) Type Y units undergoing dissolution.
Figure 6. Selective removal rate parameter, γ, for different values of the Si-to-Al ratio.
surface layer, which is in agreement with the hypothesis by Ohsuna et al.29 and Ragnarsdottir et al.31 However, precipitation leading to deposition within Type Y particles is unlikely because the diffusion of large molecules, or silicate polymers, is restricted by the pore space. Moreover, the selective removal of Al can facilitate a change in overall crystalline structure.13,29 Given that transmission electron micrographs reflect three-dimensional quantitative results translated into two-dimensional information, Figure 2h demonstrates conclusively that the Type Y structure is amorphous throughout and is confirmed by an X-ray diffraction analysis.26 The results of this study substantiate that a change in Type Y crystal structure can be attributed to localized framework modification, indicating that the selective removal of Al in excess HCl facilitates the formation of pseudo-disordered,30 undissolvable silicate structures. Figure 4 shows the Si-to-Al ratio of the undissolved zeolites as a function of reaction time revealing the contrasting dissolution phenomena of Type 4A, analcime, and Type Y zeolites. As is (31) Ragnarsdottir, K. V.; Graham, C. M.; Allen, G. C. Chem. Geol. 1996, 131, 167-181.
the case for analcime, the Type Y Si-to-Al framework ratio increases before approaching a plateau. One also observes in Figure 4 that virtually all Al has been removed from the Type Y framework and is consistent with the work of others.13,29,30,32-34 In contrast, however, for Type 4A crystals the Si-to-Al ratio remains stoichiometric (i.e., ν ≈ ν0 ) 1.0) throughout complete dissolution. These observations elucidate that Al is removed from the zeolite lattice at a faster rate than Si throughout complete dissolution when ν0 values are greater than 1.0. Application of the Selective Removal Rate Parameter, γ. To better understand the dissolution process for each zeolite, undissolved particles collected at different reaction times were redissolved in fresh hydrochloric acid according to a method described previously.25 It is readily seen in Figure 5 that measured Si and Al initial dissolution rates, -ri,0′, of Type Y approach zero with increasing values of the Si-to-Al ratio in the dissolving particle. A change in slope is observed in Figure 5 for Al removal from Type Y because the dissolution rate is constrained by the number of available reaction sites. Thus, subsequent treatments with fresh acid have a diminishing affect on the undissolved Type Y. One observes in Figure 5, however, that initial dissolution rates of Type 4A remain unchanged because there is no apparent change in the Si-to-Al ratio during the dissolution. (32) Scherzer, J. J. Catal. 1978, 54, 285. (33) Scherzer, J.; Bass, J. L. J. Catal. 1977, 46, 100. (34) Lynch, J.; Raatz, F.; Dufresne, P. Zeolites 1987, 7, 333.
5482 Langmuir, Vol. 23, No. 10, 2007
Hartman and Fogler
[Al] ) [Al]100% -
mundS0 exp[-k3f([H +])t] 4NAvoV
(6)
Where [Al]100% is the aluminum concentration at 100% dissolved, mund is the mass of undissolved zeolite remaining at time t, V is the reactor volume, and
ST ) S0 exp[-k3f([H +])t]
(7)
The kinetic rate constants k3 and Km were reported in a previous investigation.26 Samples were collected, filtered, and analyzed using ICP in order to experimentally measure the number of sites as a function of reaction time for Type Y dissolving in 4.0M HCl. As can be seen in Figure 7, plotting eq 6 as a function of dissolution time yields an excellent fit of the experimentally observed aluminum concentration. Given that the site density remains constant during Type A dissolution, i.e., ST ) S0 and ν * f(t), as revealed in Figures 4-6, integrating eq 4 gives
[Al] ) -
Figure 9. Zeolite frameworks before and after dissolution: (a, b) Type A, (c, d) analcime (modified from Hartman and Fogler25), and (e, f) Type Y.
The initial dissolution rates of Figure 5 can be reformulated in terms of a selective removal rate parameter γ (i.e., ratio of the measured Si dissolution rate to the stoichiometric Si dissolution rate),25 which better shows some important general trends. One readily observes in Figure 6 that the values of γ for Type 4A remain equal to one with no change in the Si-to-Al ratio, and hence the zeolite dissolves stoichiometrically. However, values of γ for analcime25 and Type Y are less than unity, resulting from the selective removal of Al. It follows that values of γ decrease as the stoichiometric Si-to-Al ratio, i.e., ν0, increases because there is insufficient Al to facilitate the removal of Si. Aluminum-Controlled Dissolution Kinetics. The abovementioned trends are further substantiated by applying a previously reported kinetic model for zeolite dissolution.25 The rate of aluminum dissolution, rAl′ (per mass), can be expressed in terms of the site density, ST (i.e., # per mass), for any zeolite by
rAl′ )
k3ST d[Al] 1 dST f([H +]) ) )dt 4NAvo dt 4NAvo
(4)
k3ST 4NAvo
(5)
Where
Vmax )
And f([H+]) is the rate law dependence of the H+ concentration given in eq 2, NAvo is Avogadro’s number, k3 is the rate constant for the surface reaction, and ST is the number of oxygen sites adjacent to each tetrahedrally coordinated aluminum. For Type Y zeolite, the number of reaction sites is time dependent, i.e., ST ) f(t), and an expression for the aluminum concentration in mmol/dm3 is readily derived by integrating eq 4 between the initial number of sites before dissolution, S0, and the number remaining at any given reaction time, ST, as
k3S0mund f([H+])t 4NAvoV
(8)
As described by eq 8, Type A zeolite dissolves at a constant rate until all reaction sites are depleted because all sites are accessible and the Si-to-Al ratio is 1.0. The order of magnitude of this slope is dependent on the maximum dissolution rate, Vmax, and was reported in a previous study to be on the order of 104 mmol/dm3‚min.18 It follows that the Type A dissolves virtually instantaneously, as shown in Figure 1a. These kinetic observations can be better understood by examining zeolite framework models. Framework Dissolution. The complex nature of zeolite crystal structures can sometimes make them difficult to visualize. Here, we simplify the frameworks in order to ascertain a general understanding of the contrasting dissolution observed in Figure 1. Figure 8 illustrates repeating units of Type A and Y framework flattened into two-dimensional space. Type A zeolite exhibits a stoichiometric Si-to-Al ratio of 1, and hence the framework consists of alternating Si and Al tetrahedra as illustrated in Figure 8a. It is well known that Al-O-Al fragments within a zeolite framework are excluded35 and Al atoms tend to be isolated from one another.36 Thus, we can construct simplified, two-dimensional repeating units of Type Y framework undergoing dissolution by H+ attack, as shown in Figure 8b. The cartoon of Figure 8a depicts the selective removal of Al atoms from Type A which in turn results in the release of Si and complete crystal disintegration. In contrast, however, the selective removal of Al from Type Y, exhibiting a Si-to-Al ratio of 3, leads to the formation of undissolvable silicate framework, as shown in Figure 8b. The mechanistic depiction of Figure 8 is examined in three dimensions by constructing zeolite frameworks using molecular models. As is readily seen in Figure 9a and b, the random and selective removal of Al (i.e., red tetrahedra) from Type A leads to complete dissolution of Si (i.e., black tetrahedra). As can be seen in Figure 9c and d for analcime, the release of silicon takes place when neighboring aluminum atoms are selectively removed from the framework.25 This removal results in structural collapse, leading to undissolvable silicate particles and a distribution of dissolved silicate species. In contrast, the selective removal of Al from Type Y crystals as shown in Figure 9e and f leads to an amorphous, intact silicate framework. A (35) Loewenstein, W. Am. Mineral. 1954, 39, 92. (36) Dempsey, E. J. Catal. 1974, 33, 497.
Understanding the Dissolution of Zeolites
Langmuir, Vol. 23, No. 10, 2007 5483
Figure 10. Dissolution and precipitation phenomena compared with each zeolite.
reaction mechanism whereby the selective removal of Al facilitates framework dissolution25 is in excellent agreement with experimental dissolution data and scanning electron micrographs of zeolite crystals undergoing dissolution. Universal Influence of the Si-to-Al Ratio. A comparison of the results of this study to each zeolite structure reveals the universal influence of the Si-to-Al ratio, shown schematically in Figure 10. Type A zeolites, exhibiting four aluminum tetrahedra adjacent to and surrounding each silicon tetrahedra, i.e., ν0 ) 1.0, dissolve stoichiometrically succeeded by silicate precipitation. These mechanistic observations being consistent with the research work of Oelkers and Schott37 for the dissolution of anorthite, a nonmicroporous feldspar with ν0 ) 1.0. In contrast, the zeolite analcime, where aluminum tetrahedra are adjacent to but not surrounding silicon tetrahedra, the removal of Al leads to partial Si dissolution and subsequent precipitation, and the undissolvable silicate particles remain intact.25 Comparatively, there is not enough Al removal from Type Y to weaken the structure and cause Si to dissolve, resulting in little or no precipitation and undissolvable silicate particles. Quartz, which has no aluminum, does not dissolve in hydrochloric acid.38-41 The mechanism of zeolite dissolution has general implications as to how zeolites dissolve in porous media during the acid stimulation of petroleum reservoirs. Native crystals, nucleated from a substrate such as quartz, are expected to remain impregnated when the Si-to-Al ratio is equal to or greater than about 3, even in the presence of excess HCl. Nevertheless, the selective removal of Al results in amorphous, undissolvable silicate framework. When the Si-to-Al ratio is less than about 3, but greater than 1, the selective removal of Al from zeolite minerals creates the potential for undissolved silicate particles to dislodge from the parent formation and undergo migration in a petroleum reservoir that has been acidized. This migration can lead to reservoir formation damage and thus reduce the flow of gas and oil.21,22,42 Furthermore, dissolved Si can precipitate and also result in formation damage during the acid stimulation of petroleum reservoirs.21 These scientific observations have significance on how zeolites dissolve in the presence of hydrofluoric acid, which has been shown to dissolve silicate minerals stoichiometrically.38,43,44 If there is not enough HF to (37) Oelkers, E. H.; Schott, J. Geochim. Cosmochim. Acta 1995, 59, 50395053. (38) Kline, W. E.; Fogler, H. S. J. Colloid Interf. Sci. 1981, 82, 103-115. (39) Brady, P. V.; Walther, J. V. Chem. Geol. 1990, 82, 253-264. (40) Knauss, K. G.; Wolery, T. J. Geochim. Cosmochim. Acta 1988, 52, 4353. (41) Iler, R. K. The Chemistry of Silica: Solubility, Polymerization, Colloid and Surface Properties, and Biochemistry; John Wiley and Sons, Inc.: New York, 1979. (42) Vaidya, R. N.; Fogler, H. S. SPE Prod. Eng. 1992, 325-330.
effectively dissolve all zeolite in the pore space, then fines could be generated by cleavage of O-Si-O groups, which in turn lead to detachment of partially dissolved zeolite from the formation. These basic mechanisms provide unique insight on how reservoir stimulation treatments should be designed.
Conclusions Experiments were performed to ascertain a general understanding of zeolite dissolution, revealing unique dissolution and precipitation phenomena. The Type 4A zeolite dissolves rapidly and stoichiometrically in hydrochloric acid solution followed by silicate precipitation. Non-stoichiometric dissolution takes place during Type Y dissolution resulting from the selective removal of aluminum. Measurement of aqueous product concentrations revealed that virtually all framework aluminum was removed from Type Y crystals, while the silicate framework remained intact. Examination of partially dissolved zeolite particles with a SEM confirms that Type Y crystals do not appear to dissolve. Nevertheless, etched pits and channels were formed on particle surfaces during Type 4A dissolution and are analogous to a previous investigation of analcime dissolution. Silicon and aluminum dissolution rates of the undissolved Type Y decrease by several orders of magnitude as the Si-to-Al framework ratio increases because the number of aluminum reaction sites is depleted. In contrast, Type 4A dissolution rates remain constant and the Si-to-Al ratio stoichiometric. Evaluation of the selective removal rate parameter demonstrates conclusively that aluminum plays a governing role in the dissolution mechanism of these fundamentally different zeolites. A previously developed dissolution mechanism was applied to the experimental results by constructing two- and threedimensional framework models, demonstrating that both Type 4A and Y dissolution are controlled by the selective removal of aluminum, as is the case in analcime. Comparison of the mechanism of analcime dissolution with Type 4A and Y zeolites demonstrates that, in general, the release of framework silicon takes place when adjacent and surrounding aluminum atoms are selectively removed. Thus, zeolite dissolution can result in stoichiometric framework degradation, silicate precipitation, partially dissolved silicate framework, or intact silicate framework dependent upon the initial Si-to-Al ratio. Consequently, the unique mechanism of zeolite dissolution has basic implications on how petroleum reservoir stimulation treatments should be designed. Acknowledgment. We thank the following members of the Industrial Affiliates Program on Flow and Reaction in Porous (43) Kline, W. E.; Fogler, H. S. Chem. Eng. Sci. 1981, 36, 871-884. (44) Kline, W. E.; Fogler, H. S. Ind. Eng. Chem. Fundam. 1981, 20, 155-161.
5484 Langmuir, Vol. 23, No. 10, 2007
Media at the University of Michigan for support: ChevronTexaco Energy Technology Company, ConocoPhillips Company, Nalco, Schlumberger Oilfield Chemical Products, Shell International Exploration & Production, and Total. Further, we acknowledge
Hartman and Fogler
Professor Abhaya Datye and Mangesh Bore at the University of New Mexico for kindly providing TEM micrographs. LA063699G
