Carbonation of metal silicates for long-term CO2 sequestration

Abstract

In a preferred embodiment, the invention relates to a process of sequestering carbon dioxide. The process comprises the steps of: (a) reacting a metal silicate with a caustic alkali-metal hydroxide to produce a hydroxide of the metal formerly contained in the silicate; (b) reacting carbon dioxide with at least one of a caustic alkali-metal hydroxide and an alkali-metal silicate to produce at least one of an alkali-metal carbonate and an alkali-metal bicarbonate; and (c) reacting the metal hydroxide product of step (a) with at least one of the alkali-metal carbonate and the alkali-metal bicarbonate produced in step (b) to produce a carbonate of the metal formerly contained in the metal silicate of step (a).

Claims

1. A process of producing a metal carbonate comprising: providing each of the following components: a source of magnesium hydroxide, an alkali-metal hydroxide, and a gaseous waste stream containing carbon dioxide; reacting said components together to form an alkali-metal carbonate or an alkali-metal bicarbonate from the reaction of said carbon dioxide with said alkali-metal hydroxide, while in the presence of said source of magnesium hydroxide; and reacting magnesium hydroxide from said source of magnesium hydroxide with said alkali-metal carbonate or alkali-metal bicarbonate, as said alkali-metal carbonate or alkali-metal bicarbonate forms in situ, to produce a magnesium carbonate composition.

2. The process according to claim 1, wherein said alkali-metal carbonate is sodium carbonate, and said magnesium hydroxide is reacted with said sodium carbonate to produce eitelite of the composition Na.sub.2Mg(CO.sub.3).sub.2.

3. The process according to claim 1, wherein said magnesium hydroxide is reacted with said alkali-metal bicarbonate to produce magnesium carbonate along with a hydroxide of the alkali metal.

Description

DETAILED DESCRIPTION OF THE INVENTION

(1) Many different types of metal-silicate feedstocks are amenable to carbonation by the invented process, including naturally occurring silicates such as those present in rocks and clay-rich formations, as well as silicates present in industrial waste products such as fly ash and waste concrete. Typically, the metal-silicate feedstock is composed of one or more calcium silicates, magnesium silicates, iron-bearing silicates (such as basalt), or mixtures thereof, although other types of silicates can also be used. Some nonlimiting examples of these silicates are described below. (Silicate feedstocks are referred to collectively herein as “metal silicates” with the understanding that this designation includes any natural or man-made material, in the crystalline or amorphous state, that contains at least one metal along with silicon. By this definition, aluminosilicates are metal silicates because they contain a metal, aluminum, along with silicon.)

(2) Calcium silicates include wollastonite (CaSiO.sub.3), calcic plagioclase [e.g., anorthite [(Ca.sub.x,Na.sub.1−x)(Al.sub.1+xSi.sub.3−xO.sub.8, where x≧0.9, and labradorite [(Ca.sub.x,Na.sub.1−x)(Al.sub.1+xSi.sub.3−xO.sub.8, where 0.5≦x≦0.7], calcium-rich fly ash, basalt (a volcanic rock rich in Ca, Mg and Fe), calcium-rich montmorillonite [nominally (½Ca,Na).sub.0.7(Al,Mg,Fe).sub.4[(Si,Al).sub.8O.sub.20](OH).sub.4.nH.sub.2O] and waste concrete. Contact-metamorphosed limestones frequently contain wollastonite, large quantities of calcic plagioclase are present in many different types of common rocks, and calcium-rich montmorillonite is found in special types of clay deposits. Basalt is a common rock in many terrestrial locations, and on the floors of the world's major oceans.

(3) Magnesium-rich silicates include olivine (specifically forsterite, Mg.sub.2SiO.sub.4), serpentine [Mg.sub.3Si.sub.2O.sub.5(OH).sub.4], and basalt. Significant masses of olivine- and serpentine-bearing rocks exist around the world, particularly in ultramafic complexes, and in large serpentinite bodies.

(4) Iron-bearing silicates include fayalite, Fe.sub.2SiO.sub.4, and various natural glasses (e.g., basaltic glass).

(5) The metal silicates used as feedstocks for the process can have a wide range of initial particle sizes. Typically, it is desirable to reduce the particle size of the metal silicate(s) prior to chemical treatment. For example, the particle size of the metal silicate(s) may be reduced to an average diameter of less than about 100 microns. Any suitable equipment can be used to reduce particle size.

(6) The process can optionally be conducted without heat pretreatment of the metal silicate feedstock.

(7) In a preferred embodiment of the invention, one or more metal silicates are transformed to one or more solid hydroxides by reaction with a caustic alkali-metal hydroxide, such as caustic soda (NaOH), in aqueous solution. This is the first step of a preferred process; i.e., the metal silicate(s) react with a caustic alkali-metal hydroxide to produce a hydroxide of the metal formerly contained in the silicate. This initial reaction is usually followed by physical and chemical segregation of the produced solid(s) and “depleted” caustic liquid. In addition, it may be desirable to separate the solid metal hydroxide(s) from any residual solid silicate and/or oxide material that forms as a byproduct of caustic digestion.

(8) Any suitable concentration of the caustic alkali-metal hydroxide in aqueous solution can be used to decompose the metal-silicate feedstock, including highly concentrated and very dilute solutions. The caustic solution is typically fairly concentrated, comprising, by weight, from about 30% to about 80% NaOH and from about 20% to about 70% water.

(9) In the final step of the preferred process, the metal hydroxide formed in the first step is reacted with alkali-metal carbonate (e.g., sodium carbonate) and/or alkali-metal bicarbonate (e.g., sodium bicarbonate) to produce a carbonate of the metal formerly contained in the metal silicate. This reaction can be induced at any suitable set of temperature-pressure conditions.

(10) An intermediate step in the preferred process involves reacting carbon dioxide with caustic alkali-metal hydroxide (e.g., NaOH) and/or alkali metal silicate (e.g., Na.sub.2SiO.sub.3) to produce alkali-metal carbonate (e.g., Na.sub.2CO.sub.3) and/or alkali-metal bicarbonate (e.g., NaHCO.sub.3), ±water and/or silica in either gelatinous or solid form. This step may or may not be followed by precipitation of the Na.sub.2CO.sub.3 and/or NaHCO.sub.3, which could be achieved by shifting the pH of the aqueous solution, or by evaporating off some of the water present.

(11) When all of the foregoing steps are carried out using straight flue gas as a source of CO.sub.2, capture, separation and sequestration of that gas is achieved in a single, integrated operation.

(12) Advantageously, the caustic material produced in the intermediate step of the process can usually be recycled back into the first step of the process. Thus, more generally, the invention relates to a process of carbonating a metal silicate which comprises the steps of: (a) reacting the metal silicate with a caustic material to produce a hydroxide of the metal; (b) reacting the metal hydroxide with a source of carbon dioxide to produce a carbonate of the metal and to produce reconstituted caustic material; and (c) introducing the caustic material from step (b) into step (a). The caustic material can be a caustic alkali-metal hydroxide or any other suitable caustic material.

(13) It has been discovered by the inventors that the intermediate and final steps of the process can be conducted at a pressure not greater than about 50 bars above the vapor pressure of pure water for the temperature of these two steps, typically not greater than about 30 bars, and more typically not greater than about 20 bars, and often not greater than about 10 bars. The initial step can be conducted at a pressure slightly below the vapor pressure of pure water for the temperature of that step. Achieving rapid chemical reaction at low pressure is a key technological advantage because relatively thin-walled pressure chambers will suffice to safely contain the aqueous liquids (±gas) as reaction proceeds. This will reduce the costs of commercial reactors built to implement the process on an industrial scale. Moreover, when total pressure is equal to the vapor pressure of the liquid phase, no investments in expensive pressure-intensifying equipment are required. On the other hand, higher fluid (liquid and/or gas) pressures at each step, particularly the intermediate step, could lead to more rapid and efficient chemical reaction, in which case additional capital expenditures to make the carbonation reactor more structurally robust, and to procure suitable pumping equipment, might be cost effective.

(14) More generally, the invention relates to a process of carbonating a metal silicate. The process comprises reacting at least the metal silicate and a source of carbon dioxide to produce a carbonate of the metal, wherein the reaction is conducted at a pressure not greater than about 50 bars above the vapor pressure of pure water for the temperature of the reaction.

(15) However, it may be beneficial to pressurize the CO.sub.2-bearing gas to a level above the vapor pressure of pure water for the temperature of the step in which it is reacted, prior to, or during, production of metal carbonate(s) and/or metal bicarbonate(s) in order to accelerate rates of carbonation. If CO.sub.2 is captured, separated and liquified by a another process, then pressures up to ˜64 atm (the vapor pressure of pure liquid CO.sub.2 at 25° C.) could be achieved simply by throttling flow of CO.sub.2 into the pressure chamber used to achieve carbonation.

(16) In each step of the preferred process, the extent to which aqueous liquids are agitated or stirred, and control of the proportions of phases as reaction proceeds, can be varied. In general, chemical reaction is accelerated by vigorously agitating or rapidly stirring the reactants as processing proceeds, and by maintaining high fluid/solid ratios.

(17) More generally, the invention relates to a process of carbonating a metal silicate which comprises the steps of (a) reacting the metal silicate with a caustic alkali-metal hydroxide to produce a hydroxide of the metal formerly contained in the silicate; and (b) reacting the metal hydroxide with a source of carbon dioxide to produce a carbonate of the metal formerly contained in the metal silicate of step (a). The invention also relates to a process of producing a metal carbonate. This process comprises reacting an alkaline-earth metal hydroxide with at least one of an alkali-metal carbonate, an alkali-metal bicarbonate, and carbon dioxide, to produce a carbonate of the metal formerly contained in the metal hydroxide.

(18) Some examples of reaction pathways for carbonating particular kinds of metal silicates are described below. It should be recognized that the invention is not limited to these specific examples.

(19) The invented process is advantageously used to carbonate calcium silicates. For wollastonite (nominally CaSiO.sub.3), the processing reactions are:

(20) ##STR00003##
(aq=aqueous, ↓=precipitate, liq=liquid), respectively. For anorthite (CaAl.sub.2Si.sub.2O.sub.8), the processing reactions are:

(21) ##STR00004##
(Other Ca-rich aluminosilicates carbonate by reactions similar to those for anorthite, with concomitant formation of various kinds of residual solid materials.) The NaOH “regenerated” in the third step of the process (e.g., Reactions 7 and 10) can be recycled to decompose additional calcium silicate in step 1 (e.g., Reactions 5 and 8), or to form additional Na.sub.2CO.sub.3 in step 2 (e.g., Reactions 6 and 9). Table 1 presents the results of experiments, performed by the inventors, that confirm Reactions 5-10.

(22) In step 1 of the calcium-silicate carbonation processes (e.g., Reactions 5 and 8), a solid, calcium-rich silicate feedstock is decomposed in an aqueous solution of caustic soda to produce crystalline sodium-calcium hydroxysilicate, or crystalline portlandite and crystalline hydroxy-sodalite. This step is followed by physical and chemical segregation of the precipitated solid(s) and “depleted” caustic liquid. In step 2 (e.g., Reactions 6 and 9), carbon dioxide is injected into a NaOH-bearing aqueous liquid, creating aqueous Na.sub.2CO.sub.3 and water. In step 3 (e.g., Reactions 7 and 10), the Na.sub.2CO.sub.3 generated in step 2 is reacted with the crystalline hydroxide(s) formed in step 1 to produce aqueous NaOH; and either crystalline calcite+crystalline pectolite, or crystalline calcite+crystalline carbonate-cancrinite. These solid reaction products are thermodynamically stable, environmentally benign, and sparingly soluble in meteoric water.

(23) The invented process also converts magnesium-rich silicates to magnesium-rich carbonates. For example, the following process reactions produce magnesite (MgCO.sub.3) from the magnesium-rich minerals olivine (specifically forsterite, Mg.sub.2SiO.sub.4) and serpentine [Mg.sub.3Si.sub.2O.sub.5(OH).sub.4]:

(24) ##STR00005##
[Net reaction: Mg.sub.3Si.sub.2O.sub.5(OH).sub.4+3CO.sub.2.fwdarw.3MgCO.sub.3(↓)+2SiO.sub.2(↓)+2H.sub.2O(liq)]. Table 2 presents the results of experiments, performed by the writers, that confirm Reactions 11-17.

(25) A key observation concerning Reaction 14 is that no heat pretreatment of the serpentine is required to achieve rapid and efficient production of Mg(OH).sub.2. This contrasts sharply with the so-called “direct” method for carbonating serpentine (by the reaction Mg.sub.3Si.sub.2O.sub.5(OH).sub.4+3CO.sub.2.fwdarw.3MgCO.sub.3+2SiO.sub.2+2H.sub.2O), which requires heat pretreatment of the serpentine at ˜600° C. to drive off structurally bound water. This extra step is necessary in the direct method of carbonating serpentine because water-bearing (hydroxylated) serpentine reacts sluggishly with CO.sub.2, whereas dewatered (dehydroxylated) serpentine is highly reactive. Dehydroxylation of serpentine makes the direct carbonation method very energy intensive and costly. In this regard, it is also noteworthy that the invented process, as applied to either olivine or serpentine, completely conserves the “rock solvent” (e.g., NaOH), which lowers overall processing costs. By contrast, in the direct method for carbonating serpentine and olivine, the rock solvent is (effectively) compressed, supercritical CO.sub.2, which is expensive to create due to the high capital and operating costs of the mechanical pumping that is required to achieve pressures as high as 185 atm.

(26) Another important discovery made by the inventors is that Mg(OH).sub.2 can be reacted with Na.sub.2CO.sub.3 at 200° C. and elevated CO.sub.2 fugacities to form the double carbonate eitelite [Na.sub.2Mg(CO.sub.3).sub.2] (see Table 2), which contains twice as much CO.sub.2 as magnesite (MgCO.sub.3). Crystallization of eitelite would permit twice as much CO.sub.2 to be sequestered per ton of mined Mg-rich rock. In addition, like magnesite, eitelite is thermodynamically stable, environmentally neutral, and only weakly soluble in meteoric water. However, a disadvantage of eitelite crystallization is that caustic soda is not “regenerated” simultaneously, as it is when magnesite crystallizes (see Reactions 13 and 16). Therefore, when either olivine or serpentine is used to produce eitelite, the cost-savings achieved by reducing the tonnage of mined Mg-rich rock are offset to a significant extent by the continuous need to replenish the supply of NaOH that is used to decompose the metal-silicate feedstock.

(27) It should be clearly understood that the chemical formulae for the solutes (substances dissolved in aqueous solution) in Reactions 11-13 and 14-16 (specifically NaOH, Na.sub.2SiO.sub.3 and NaHCO.sub.3) represent stoichiometric components in aqueous solution, not “real” aqueous species. This convention was adopted (see also Reactions 5-7 and 8-10) for the sake of generality and simplicity. The particular species in aqueous solution created by our process (presently unknown) are of considerable scientific interest; however, they need not be represented explicitly in sets of process reactions such as those above, because the solids that form and disappear in each process reaction, as well as the net carbonation reaction for each metal silicate, do not depend on the chemical formulae that are used to represent the compositions of solutes. A simple example serves to illustrate this point. In Reactions 11-13, the stoichiometric components NaOH, Na.sub.2SiO.sub.3 and NaHCO.sub.3 can be replaced by the ionic species OH.sup.−, SiO(OH).sub.3.sup.−, and HCO.sub.3.sup.−, with sodium ion omitted because it is neither consumed nor produced in any reaction. This leads to the following alternative carbonation pathway for forsteritic olivine:
Mg.sub.2SiO.sub.4+OH.sup.−+3H.sub.2O.fwdarw.2Mg(OH).sub.2+SiO(OH).sub.3.sup.−  (18)
SiO(OH).sub.3.sup.−+CO.sub.2.fwdarw.HCO.sub.3.sup.−+SiO.sub.2(↓)+H.sub.2O  (19)
Mg(OH).sub.2+HCO.sub.3.sup.−.fwdarw.MgCO.sub.3(↓)+OH.sup.−+H.sub.2O  (20)
Mg(OH).sub.2+CO.sub.2.fwdarw.MgCO.sub.3(↓)+H.sub.2O  (21)
[Net reaction: Mg.sub.2SiO.sub.4+2CO.sub.2.fwdarw.2 MgCO.sub.3(↓)+SiO.sub.2(↓)]. Comparing Reactions 18-21 with Reactions 11-13, it is evident that the solids consumed and produced, and the net reaction, are identical. Therefore, it should be clearly understood that the scope of our process for carbonating metal silicates includes various self-consistent sets of reactions—i.e., sets of reactions involving the same solids, with metal silicate digestion by one or more caustic metal hydroxides (such as NaOH)—wherein solutes are represented by aqueous species of varying composition and charge, rather than by stoichiometric components.

(28) The invented process may also convert iron-bearing silicates to iron-bearing carbonates, following carbonation pathways similar to those described above for Ca- and Mg-rich silicates. This is illustrated below for the iron-bearing silicate fayalite, Fe.sub.2SiO.sub.4:

(29) ##STR00006##
[Net reaction: Fe.sub.2SiO.sub.4+2CO.sub.2.fwdarw.FeCO.sub.3(↓)+SiO.sub.2(↓)]. It is likely that similar reactions will convert Fe-bearing silicate glasses (e.g., basaltic glass) to one or more Fe-bearing carbonates.

(30) In another embodiment of the invention, Ca-rich silicates are carbonated in a single step. In this alternative carbonation pathway, the silicate feedstock, aqueous NaOH, and Na.sub.2CO.sub.3 are reacted to produce crystalline calcite±crystalline sodium-calcium hydroxysilicate±crystalline hydroxy-sodalite±crystalline carbonate-cancrinite±residual silicate material. For example, using a concentrated aqueous solution of caustic soda (such as 50 weight percent NaOH in H.sub.2O) as the silicate solvent, and by adding abundant sodium carbonate to that solution to serve as a source of CO.sub.2, the aluminum-bearing calcium silicates anorthite and labradorite react with NaOH and Na.sub.2CO.sub.3 to form crystalline calcite±crystalline hydroxy-sodalite±residual aluminosilicate material at 200° C. and a total (fluid) pressure <15 atm (see Table 1).

(31) The single-step process for carbonating Ca-rich silicates is simpler than the three-step process described earlier, because physical and chemical segregation of the solids and liquids is not required. However, the one-step process has the disadvantage that carbonate-cancrinite is typically not among the solids that are produced. (In the “one-step” experiments performed by the inventors using Ca-rich silicates as a source of calcium, carbonate-cancrinite was only observed in experiments performed with Ca-rich fly ash.) A plausible explanation for this is that the aqueous fluids produced in one-step experiments were typically too basic to allow carbonate-cancrinite to crystallize. Therefore, in an industrial-scale implementation of the one-step process, it may be desirable, at some point, to lower the pH of the aqueous fluid to increase the possibility that carbonate-cancrinite will be produced. The formation of carbonate-cancrinite significantly increases the total “CO.sub.2-loading” of the solids produced, and therefore leads to more efficient and cost-effective carbonation of the Ca-silicate feedstock.

(32) In still another embodiment, the invention provides a means for carbonating magnesium and iron silicates in two steps. In step 1, the metal silicate(s) is (are) converted to Mg(OH).sub.2 and/or iron hydroxides(s)+Na.sub.2SiO.sub.3±SiO.sub.2 by reaction with caustic soda in aqueous solution (e.g., Reactions 11, 14 and 22). When this conversion is essentially complete, carbonation of Mg(OH).sub.2 and/or iron hydroxide(s) is (are) achieved by injecting CO.sub.2 into the aqueous solution to form NaHCO.sub.3(±Na.sub.2CO.sub.3)+silica gel and/or solid silica (step 2). MgCO.sub.3 is formed when the Mg(OH).sub.2 produced in step 1 reacts with NaHCO.sub.3(±Na.sub.2CO.sub.3) and/or aqueous CO.sub.2 (e.g., Reactions 13, 16 and 24).

(33) Physical and/or chemical segregation of solids and liquids is not required in the two-step process for carbonating Mg- and Fe-rich silicates; therefore, it is intrinsically simplier than the three-step process for carbonating forsteritic olivine (Reactions 11-13), and the four-step process for carbonating serpentine (Reactions 14-17). On the other hand, the two-step process has the disadvantage that silica and magnesite are produced simultaneously. It should be recognized that Reactions 11-13 and 14-17 produce separate “streams” of gelatinous/solid silica and crystalline magnesite, which allows each substance to be used as a feedstock for various commercial applications. In contrast, the two-step process for carbonating magnesium and iron silicates generates a single stream of solid material, consisting of intimately mixed gelatinous/solid silica and crystalline magnesite. This creates two difficulties. First, gelatinous silica (if formed) is readily mobilized by meteoric water, and is therefore an inherently undesirable waste product for near-surface terrestrial disposal. Second, when silica (in any form) and magnesite are intimately mixed, the high cost of separation effectively precludes any value-added commercial applications of each substance. Therefore, compared to separate masses of its individual components, an intimate mixture of silica and magnesite is an economic and environmental liability.

(34) The invention also relates to a system for sequestering carbon dioxide from a gas stream. The system includes a gas stream containing carbon dioxide, for example, a flue gas containing carbon dioxide or a stream of pure carbon dioxide gas. The system also includes a first reaction chamber for reacting a metal silicate with a caustic material to produce a hydroxide of the metal. The caustic material can be a caustic alkali-metal hydroxide, as discussed above, or any other caustic material suitable for the reaction. The system also includes a second reaction chamber for contacting the metal hydroxide with the gas stream containing the carbon dioxide to produce a carbonate of the metal. The first and second reaction chambers can be located in the same reactor or in different reactors. More than two reaction chambers can optionally be used in the system.

(35) The invention also relates to a system for carbonating a metal silicate. The system includes a supply of the metal silicate entering the system, and a source of carbon dioxide entering the system. The source of carbon dioxide can be a gas stream containing carbon dioxide, or it can be a compound containing carbon dioxide, such as an alkali-metal carbonate or bicarbonate. The system also includes a reactor structured for converting the metal silicate to a metal carbonate and silica with the use of a caustic material, and with the use of the carbon dioxide. Any suitable caustic material can be used, such as a caustic alkali-metal hydroxide. The reaction can include other reactants in addition to the metal silicate, the caustic material and the carbon dioxide. The metal carbonate and the silica exit the system as separate products. In a preferred embodiment, the metal silicate is magnesium silicate and the metal carbonate is magnesite. Typically, the magnesite and the silica exiting the system have a purity of at least about 90%. This is economically advantageous, and it contrasts with previously known processes that produce mixtures of magnesite and silica.

(36) The invention also relates to a system for recovering a useful metal from rock. The system includes a supply of rock entering the system, the rock containing the useful metal and a metal silicate, and a source of carbon dioxide entering the system. The system also includes a reactor structured for converting the metal silicate to a metal carbonate, with the use of a caustic material, and with the use of the carbon dioxide. As discussed above, any suitable caustic material can be used, and the source of carbon dioxide can be a gas containing carbon dioxide or a compound containing carbon dioxide. The system also includes apparatus for removing the useful metal from the rock. The apparatus can be located at any suitable location in the system. For example, magnetic apparatus can be used for removing magnetite from serpentine prior to the reactor. Alternatively, the reactor can produce the metal carbonate and a remaining rock portion, and the system can include apparatus for removing the useful metal from the remaining rock portion at a location subsequent to the reactor. The system also includes a stream of the metal carbonate exiting the system, and a stream of the useful metal exiting the system. The streams can be in any form, such as truckloads, trainloads, or other means of conveying the metal carbonate and the useful metal from the system.

(37) The carbonation pathways for Ca-, Mg- and Fe-rich silicates described above are similar to, but distinctly different from, the commercial chemical processes that are used to extract alumina (Al.sub.2O.sub.3) from bauxite (aluminum ore), and to generate caustic soda from trona (a rock rich in sodium carbonate, Na.sub.2CO.sub.3). In the treatment of bauxite ore by the well-known Bayer Process, caustic soda is used to remove reactive silica and iron oxides, and to dissolve aluminum oxides (gibbsite, boehmite and diaspore). Dissolution of silica by the caustic solution produces sodium silicate (nominally Na.sub.2SiO.sub.3), which quickly reacts with sodium aluminate (NaAlO.sub.2) to form crystalline hydroxy-sodalite [Na.sub.8(AlSiO.sub.4).sub.6(OH).sub.2]. This desilication of the solution is detrimental to the overall process because it consumes caustic soda, and the total mass of dissolved sodium aluminate is lowered. (With less sodium aluminate in solution, less high-purity gibbsite is precipitated in a later stage of the process, and as a second consequence, less caustic soda is regenerated by gibbsite precipitation. Regenerated caustic soda is recycled in the Bayer Process to treat additional batches of bauxite ore.)

(38) While the Bayer Process has several characteristics in common with the process described in this document—in particular, the use of caustic soda to decompose the metal-bearing, solid feedstock, and regeneration of caustic soda at a subsequent stage of the process—two major differences are also evident. The first is that the goals of the two processes are totally different: the Bayer Process was developed to produce a solid concentrate rich in aluminum (precipitated gibbsite), whereas the principal intent of the invented metal-silicate carbonation process is to form stable metal-carbonate compounds for long-term CO.sub.2 sequestration. Due to this key difference, the “ore” used for the two processes is much different. In the Bayer Process, deeply weathered, unconsolidated rock material is reacted because it is rich in alumina and poor in silica. In the present invention, the silica content of the metal-silicate feedstock is not a significant factor, except that a greater silica content generally means a lower Ca, Mg and/or Fe content, which is undesirable. The second key difference is that, in the present invention, caustic soda is regenerated during crystallization of one or more solid metal carbonates, whereas in the Bayer Process, caustic soda is regenerated during the production of high-purity gibbsite.

(39) In the commercial process of reacting trona with lime (CaO) to produce caustic soda plus calcite, the intent is solely to produce caustic soda; the entire amount of calcite formed as a byproduct is subsequently calcined to regenerate lime, which is recycled for reaction with additional batches of trona ore to produce more caustic soda. Thus, there is no CO.sub.2 sequestration achieved in the commercial treatment of trona. Moreover, trona contains little or no silica and alumina; consequently, its treatment to generate caustic soda does not consume or produce any significant amount of silicate material.

(40) The invented metal-silicate carbonation process may have the following practical uses and benefits:

(41) CO.sub.2 sequestration. The process was designed mainly to produce crystalline carbonates that persist indefinitely in most continental settings. The results of autoclave experiments presented in Tables 1 and 2 demonstrate that, under suitable conditions: Ca-rich silicates are readily converted to calcite (CaCO.sub.3)±carbonate-cancrinite [Na.sub.8(AlSiO.sub.4).sub.6CO.sub.3.2H.sub.2O], and Mg-rich silicates are quickly transformed to magnesite (MgCO.sub.3) and/or eitelite [Na.sub.2Mg(CO.sub.3).sub.2]. These four crystalline carbonate compounds bind CO.sub.2 indefinitely in most non-acidic terrestrial environments, and are completely harmless to all flora and fauna.

(42) Neutralization of highly acidic soils. Locally, calcium and magnesium carbonates can have commercial value as soil amendments. While adding carbonates to highly acid soils to increase pH ultimately releases CO.sub.2 to the atmosphere, significant environmental and economic benefits would accrue if the treated land was made more biologically productive. The amended soils might be used to grow crops or trees, either of which could have a total carbon sequestration potential higher than that of the crystalline carbonate amendment. Using calcium and magnesium carbonates for this purpose would also lessen demand for lime produced by calcining limestone, and this would help lower CO.sub.2 emissions to the atmosphere.

(43) Recovery of useful metals. Many rock formations contain useful metals (e.g., iron, copper, nickel and platinum) at concentrations that are currently uneconomical to mine. If, however, mining and grinding were already being performed to create a metal-silicate feedstock for carbonation, one or more metals that are not carbonated could be extracted as a byproduct(s), thereby reducing the costs of metal-silicate carbonation.

(44) Elimination of hazardous mine tailings. Mine tailings, consisting of crushed rock material from which metals or other valuable materials have been extracted—along with the “overburden” (soil and regolith) that is removed to access buried ore horizons—are an important waste problem for many active and abandoned mines. Thus, it is significant that a nearby mineral carbonation reactor might be able to use them as a source of calcium and magnesium. This would enhance environmental restoration, and reduce the costs of mining metal silicate(s) for carbonation.

(45) Production of high-purity silica. When olivine and/or serpentine is (are) carbonated by the invented process, the solid effluent produced in step 2 (Reactions 12 and 15) is high-purity silica, which can be refined for use in manufacturing silica-based desiccants, silica brick, silicon carbide, and various types of glass. High-purity silica is also a potential source of elemental silicon—the foundation material for numerous semiconducting electronic devices. Finally, it may be possible to use amorphous silica to form melanophlogite, a silica-rich compound with a cage structure that can accommodate as many as six CO.sub.2“guest” molecules for every 46 molecules of SiO.sub.2. Creation of substantial amounts of melanophlogite would significantly increase the total CO.sub.2-loading of the solids generated by the invented process.

(46) Production of high-purity magnesite. When olivine and/or serpentine is (are) carbonated by the invented process, the solid effluent produced in step 3 for olivine carbonation (Reaction 13) and in steps 3 and 4 for serpentine carbonation (Reactions 16 and 17) is high-purity magnesite, which can be used to produce magnesite cement.

(47) Capture and separation of CO.sub.2 from flue gas. Another potential application of the invented process is especially important, as it would greatly reduce costs associated with capture and separation of CO.sub.2 at fossil fuel-fired power plants, cement factories, and steel mills. Specifically, step 2 in the process (e.g., Reactions 6, 9, 12 and 15) permits CO.sub.2 to be captured and separated from flue gas by bubbling the gas through a NaOH- and/or Na.sub.2SiO.sub.3-bearing aqueous liquid. The CO.sub.2 would be transformed to aqueous±crystalline Na.sub.2CO.sub.3 and/or NaHCO.sub.3, and the nitrogen-rich gas effluent could either be refined to produce high-purity nitrogen for commercial use, or simply released harmlessly to the atmosphere.

(48) TABLE-US-00001 TABLE 1 Results of Experiments with Solid Calcium Silicates Date/ Temperature/ XRD Solid Run Products, and Duration.sup.@ Starting Materials.sup.@,# Pressure Number Initial/Final Solution pH.sup.& Jul. 10, 2002/ anorthite + NaOH + H.sub.2O 200° C./ LMAS001-2 portlandite + hydroxy-sodalite + 72 hrs <15 atm LMAS003 hydrogrossularite(?) Aug. 1, 2002/ anorthite + NaOH + 200° C./ LMA059 calcite + hydroxy-sodalite 72 hrs Na.sub.2CO.sub.3 + H.sub.2O <15 atm Oct. 4, 2002/ prereacted anorthite + 200° C./ LMA070 calcite + carbonate-cancrinite, 72 hrs Na.sub.2CO.sub.3 + H.sub.2O <15 atm pH: 12.5/≧14 Aug. 20, 2002/ labradorite + NaOH + H.sub.2O 200° C./ LMA060 portlandite + hydroxy-sodalite + 72 hrs <15 atm cancrinite + hydrogossular(?) Aug. 28, 2002/ labradorite + NaOH + 200° C./ LMA061 calcite + hydroxy-sodalite + 72 hrs Na.sub.2CO.sub.3 + H.sub.2O <15 atm hydrogrossular + minor AUCP, pH: 12.5/≧14 Oct. 8, 2002/ prereacted labradorite + 200° C./ LMA071 calcite + carbonate-cancrinite, 72 hrs Na.sub.2CO.sub.3 + H.sub.2O <15 atm pH: 12.5/≧14 Nov. 8, 2002/ Ca(OH).sub.2 + Mg(OH).sub.2 + 200° C./ LMA077 calcite + brucite, pH: 12.5/≧14 72 hrs Na.sub.2CO.sub.3 + H.sub.2O <15 atm Nov. 26, 2002/ wollastonite + NaOH + H.sub.2O 200° C./ LMA081 sodium-calcium 72 hrs <15 atm hydroxysilicate Dec. 2, 2002/ wollastonite + NaOH + 200° C./ LMA083 calcite + sodium-calcium 72 hrs Na.sub.2CO.sub.3 + H.sub.2O <15 atm hydroxysilicate, pH: 12.5/≧14 Dec. 10, 2002/ Ca-rich fly ash + NaOH + 200° C./ LMA084 sodalite + AUCP 72 hrs H.sub.2O <15 atm Dec. 13, 2002/ Ca-rich fly ash + NaOH + 200° C./ LMA085 calcite + carbonate-cancrinite + 72 hrs Na.sub.2CO.sub.3 + H.sub.2O <15 atm sodalite + AUCP, pH: 12.5/≧14 Dec. 17, 2002/ Ca(OH).sub.2 + Na.sub.2CO.sub.3 + H.sub.2O 200° C./ LMA089 calcite, pH: 12.5/≧14 72 hrs <15 atm Jan. 3, 2003/ prereacted wollastonite + 200° C./ LMA094 calcite + pectolite, pH: 72 hrs NaOH + Na.sub.2CO.sub.3 + H.sub.2O <15 atm 12.5/≧14 Sep. 16, 2002/ basalt + NaOH + H.sub.2O 200° C./ LMA064 portlandite + sodalite + brucite 72 hrs <15 atm Sep. 30, 2002/ basalt + NaOH + Na.sub.2CO.sub.3 + 200° C./ LMA063 brucite + AUCP 72 hrs H.sub.2O <15 atm Oct. 14, 2002/ prereacted basalt + NaOH + 200° C./ LMA059 calcite + carbonate-cancrinite, 72 hrs Na.sub.2CO.sub.3 + H.sub.2O <15 atm pH: 12.5/≧14 .sup.@Sources of solid-silicate starting materials: wollastonite-Willsboro, NY; anorthite-Grass Valley, CA; labradorite-Nain, Labrador, Canada; basalt (USGS standard BCR-2)-Portland, OR; Ca-rich fly ash, Joseph City, AZ; Ca(OH).sub.2 (portlandite)-commercially manufactured, reagent-grade chemical compound; Mg(OH).sub.2 (brucite)-commercially manufactured, reagent-grade chemical compound. NaOH (caustic soda) was added to each unreacted starting sample as a 50 weight percent solution of NaOH in H.sub.2O. .sup.#For each solid silicate, “prereacted” means reaction with a 50 weight percent solution of NaOH in H.sub.2O at 200° C., P < 15 atm, before the experiment was performed. In all cases, this step had the effect of converting the silicates to crystalline hydroxide(s) prior to reaction with Na.sub.2CO.sub.3 to form one or more crystalline carbonate compounds. .sup.&Chemical compositions of crystalline reaction products: calcite—CaCO.sub.3; portlandite—Ca(OH.sub.2); hydroxy-sodalite—Na.sub.8(AlSiO.sub.4).sub.6(OH).sub.2; sodium-calcium hydroxysilicate—NaCaSiO.sub.3(OH); carbonate-cancrinite—Na.sub.8(AlSiO.sub.4).sub.6CO.sub.3•2H.sub.2O; brucite—Mg(OH).sub.2; pectolite—NaCa.sub.2Si.sub.3O.sub.8(OH); hydrogrossular—Ca.sub.3Al.sub.2(SiO.sub.4).sub.3•xH.sub.2O. AUCP = additional unidentified crystalline phase(s).

(49) TABLE-US-00002 TABLE 2 Results of Experiments with Crystalline Magnesium Silicates Date/ Temperature/ XRD Solid Run Products, and Duration.sup.@ Starting Materials.sup.@,# Pressure Number Initial/Final Solution pH.sup.& Jun. 10, 2002/ serpentine + NaOH + H.sub.2O 200° C./ LMAS001-1 brucite + relict serpentine 72 hrs <15 atm Jul. 1, 2002/ olivine + NaOH + H.sub.2O 200° C./ — brucite 72 hrs <15 atm Jul. 16, 2002/ prereacted olivine + 200° C./ LMAS006 brucite 72 hrs Na.sub.2CO.sub.3 + H.sub.2O <15 atm Sep. 20, 2002/ serpentine + NaOH + H.sub.2O 200° C./ LMA069 brucite + trace AUCP 24 hrs <15 atm Sep. 23, 2002/ serpentine + NaOH + H.sub.2O 200° C./ LMA067 brucite + trace AUCP 72 hrs <15 atm Sep. 25, 2002/ Mg(OH).sub.2 + Na.sub.2CO.sub.3 + H.sub.2O 200° C./ LMA072 brucite, pH: 12.5/12.5 72 hrs <15 atm Oct. 17, 2002/ serpentine + NaOH + H.sub.2O 22° C./ LMA073 serpentine + brucite 72 hrs 1 atm Oct. 18, 2002/ serpentine + NaOH + H.sub.2O 22° C./ LMA075 brucite + minor relict 72 hrs at 1 atm, serpentine 22° C., then 3 200° C./ hrs at 200° C. <15 atm Nov. 15, 2002/ Mg(OH).sub.2 + Na.sub.2CO.sub.3 + H.sub.2O 300° C./ LMA078 brucite 72 hrs <86 atm Nov. 19, 2002/ Mg(OH).sub.2 + Na.sub.2CO.sub.3 + H.sub.2O 375° C./ LMA079 brucite ~1 hr <220 atm Dec. 23, 2002/ Mg(OH).sub.2 + Na.sub.2CO.sub.3 + H.sub.2O 200° C./ LMA088 eitelite + minor relict brucite 72 hrs ~60 atm* Dec. 30, 2002/ Mg(OH).sub.2 + H.sub.2O 200° C./ LMA090 magnesite + minor relict 72 hrs ~60 atm* brucite, final pH: 5.5 Jan. 17, 2003/ Mg(OH).sub.2 + Na.sub.2CO.sub.3 + H.sub.2O 300° C./ LMAS014 brucite + eitelite, pH: 11.5- 72 hrs ~60 atm* 12/12 Mar. 17, 2003/ Mg(OH).sub.2+ Na.sub.2CO.sub.3 + H.sub.2O 100° C./ LMAS017 brucite + magnesium carbonate 72 hrs <1 atm hydrate Mar. 21, 2003/ Mg(OH).sub.2 + Na.sub.2CO.sub.3 + H.sub.2O 80° C./ LMAS022 brucite + magnesium carbonate 72 hrs <1 atm hydrate, pH: 12.5/12.5 Apr. 07, 2003/ Mg(OH).sub.2 + Na.sub.2CO.sub.3 + H.sub.2O 125° C./ LMAS030 brucite + magnesium carbonate 72 hrs <2 atm hydrate, final pH: 11.5 Apr. 16, 2003/ Mg(OH).sub.2 + NaHCO.sub.3 + H.sub.2O 200° C. — magnesite + minor eitelite + 72 hrs <15 atm minor relict brucite .sup.@Sources of crystalline silicate starting materials: olivine-Twin Sisters Peak, WA; serpentine (antigorite variety)-Cedar Hill Quarry, Lancaster County, PA; Mg(OH).sub.2 (brucite)-commercially manufactured, reagent-grade chemical compound. NaOH (caustic soda) was added to each unreacted starting sample as a 50 weight percent solution of NaOH in H.sub.2O. .sup.#For each solid silicate, “prereacted” means reaction with a 50 weight percent solution of NaOH in H.sub.2O at 200° C., P < 15 atm, before the experiment was performed. In all cases, this step had the effect of converting olivine or serpentine to brucite prior to reaction with Na.sub.2CO.sub.3 or NaHCO.sub.3 to form one or more crystalline carbonate compounds. *A total (fluid) pressure of approximately 60 atm was achieved by injecting CO.sub.2 into the headspace of the autoclave, and keeping that space connected to an external cylinder filled with liquid CO.sub.2 at room temperature. .sup.&Chemical compositions of crystalline reaction products: olivine—Mg.sub.2SiO.sub.4; serpentine—Mg.sub.3Si.sub.2O.sub.5(OH).sub.4; brucite—Mg(OH).sub.2; eitelite—Na.sub.2Mg(CO.sub.3).sub.2. magnesium carbonate hydrate—MgCO.sub.3•xH.sub.2O. AUCP = additional unidentified crystalline phase(s).

(50) In accordance with the provisions of the patent statutes, the principle and mode of operation of this invention have been explained and illustrated in its preferred embodiments. However, it must be understood that this invention may be practiced otherwise than as specifically explained without departing from its spirit or scope.