Nanoscale iron is a material that is generating interest as a groundwater remediation technology for the reductive dehalogenation of various solvents, treatment of oxyanions such as chromate, and other constituents that may respond to strong abiotic reductive processes such as perchlorate. During initial development, it was thought that the unique character of nanoscale particles was due solely to the small size. Current studies show that the structure and composition of the particle is also critical to both reactivity and longevity. These newly understood relationships could have measurable implications for advancing the maturity of the technology as a practical tool for in-situ remediation.

Introduction

The use of granular zero-valent iron in permeable reactive barriers is a maturing technology that saw initial application in the middle 1990s. Nanoscale iron offers distinct advantages over conventional iron for in-situ applications in groundwater because they can be delivered to the subsurface without the need for excavation. This removes limitations due to depth, site structures and facility operations. A nanoscale colloid is small enough to flow through the pore space and associated pore throats of many granular geologic matrixes. The issues governing potential transport are not so much direct filtration, but the effects of gravity and the attractive forces at work on the mineral and colloid surfaces; this also includes agglomeration of the colloids in the injection solution.

In the earlier stages of development of the nanoscale iron colloid, it was thought that the high reactive rates of the colloids were solely due to size, since as size decreases the surface area per unit of mass increases. Surface areas of 30 square meters per gram were obtained with highly reactive 20- to 60-nanometer colloids.

Practical controls on in-situ iron colloid delivery

Figure 1

As the technology developed, it became apparent that there were a number of forces at work controlling the in-situ deliverability of the colloid. These forces can best be summarized as interfacial attractive and gravity. Attractive forces can be broadly divided into two categories: electrostatic and Van der Walls attraction.

Electrostatic forces are simply controlled by pH. Solid surfaces in water have a point of zero charge; at a pH above that point, surfaces will have a negative charge, and a positive charge below that point. As shown in Figure 1, the point of zero charge for most silicate minerals is in the range of 2.5. Under normal pH conditions, the electrostatic charge of a mineral matrix is negative. Consideration should also be given to trace minerals such as iron oxides with a point of zero in the range of pH 7. The point of zero charge for metallic iron is approximately 6.5. The ideal condition for mobility is to establish a pH at which the iron colloid has the same charge as the mineral matrix and is therefore electrostatically repulsed from mineral surfaces.

Van der Walls forces originate from the atomic nucleus. Practical effects of these forces can be best estimated by the Brownian velocity of the colloid in suspension. For the Van der Walls forces to have effect, the repulsive electrostatic forces must be overcome. If the colloid can get close enough to a surface, it will stick. Higher Brownian velocities allow that to happen and, as shown in Figure 2, the smaller a colloid, the greater the Brownian velocity.

The same forces govern interactions between nanoscale iron colloids dispersed in an injection solution; the colloids tend to be attracted to each other and agglomerate, losing the ideal size configuration. Dispersants can be used to aid in the control of agglomeration. But more fundamentally, the number of colloidal particles per unit mass of iron is inversely proportional to the cube of the colloid diameter. A 500-nanometer colloid is 10 times larger than a 50-nanometer colloid. A gram of 50-nanometer colloid has 1000 times more colloid particles than a gram of 500-nanometer colloid. Since colloid aggregation rates are proportional to the square of the particle number, the 500-nanometer colloid is a million times less susceptible to agglomeration compared to the 50-nanometer colloid.

The third in-situ force at work is gravitational settling. The larger a particle, the faster it will settle due to the force of gravity. Taking into account Brownian velocity and the effects of gravitational settling, there is an ideal size range for the in-situ application of nanoscale colloids; colloids larger than 200 nanometers have a low Brownian velocity; colloids smaller than 600 nanometers have a slow settling velocity.

However, colloids in the size range of 200- to 600-nanometers have much less surface area than those in the 20- to 60-nanometer range. Thus, some reactivity is theoretically lost. This has created the need to enhance or control colloid reactivity by means other than just size. This, in turn, touches on the manufacturing process.

Manufacturing of a field-practical nanoscale iron colloids

Figure 2

Current production methods for nanoscale iron colloids may be divided into two primary approaches: bottom-up and top-down. Bottom-up colloids are made by assembling individual atoms. Within the bottom-up approach, there are a number of potentially applicable processes, including chemical precipitation of metal from soluble salts. An example is chemical reduction using sodium borohydride with ferrous chloride in water. The top-down approach uses the attrition of larger particles of metal to produce colloids of the appropriate size. This approach uses variations of mechanical comminuting, such as high energy ball milling. Figure 3 is a photomicrograph of an early production run of top-down iron colloids.

Controls on colloid reactivity

Figure 3

In use, the iron particles undergo anaerobic corrosion when they react directly with halogenated solvents, or when they react with water to produce hydrogen that is used in subsequent dehalogenation reactions. These are a complex series of reactions. Aside from the simple reactivity effect of small size and surface area, effects due to the availability of individual atoms (and their associated electrons, which is the prime driver of the reactive process) also become apparent. The number of atoms located at the surface or in the interfacial region near the surface increases as particle size decreases: 1 to 2 percent of the atoms in a 100 nanometer particle, 10 to 15 percent at 10 nanometers, and 20 to 30 percent at 5 nanometers.

A colloid produced by chemical precipitation or reduction methods will be nano-structured. This means that the colloid will have distinct nano-crystal domains with sharp boundaries between crystals. The grain boundaries are typically only one atom thick and there is low dislocation density in the crystal structures. The reactivity of the bottom-up colloid is controlled by the colloid size and resulting surface area, with some additional control offered by chemical amendments to bulk or surface of the colloid to be discussed.

Alternatively, a colloid produced by mechanical attrition will be nano-crystalline. In this case, the crystal domains in the colloid are small, relative to the overall colloid size. The individual crystal domains are separated by wide amorphous transition regions that exhibit a very high dislocation density. These amorphous transition regions have many available electrons and are highly reactive. The size and intensity of dislocation density of the amorphous boundary regions rather than the absolute size of the colloid dominate the reactivity of the top-down colloid.

Colloid longevity versus reactivity

Long-term reactivity of the colloid is important for in-situ remediation applications. Iron particles are corroded when they react directly with halogenated solvents or with water to produce hydrogen. The products of the corrosion process are iron oxides and ultimately a range of iron-containing minerals. The iron metal may also react with other inorganic constituents dissolved in groundwater such as bicarbonate, sulfate, nitrate and other species that are subject to reduction reactions. These iron oxides and other mineral byproducts coat and passivate the surface of the iron particle.

Hydrogen diffusion through dislocations and grain boundaries is much faster than in the ordered crystal lattice. In addition, such defects act as deep traps for hydrogen storage. These structural features also provide a means of overcoming passivation films that develop under field applications. The passivating oxide film can be broken down by anions present in groundwater, especially chloride (fortuitous for the treatment of chlorinated VOCs). This localized chloride dissolution of the passivating film takes place at weak points, including grain boundaries and dislocations, which lead to the exposure of the underlying metal allowing further reaction.

The effect of colloid composition

Nanoscale colloids have been produced with trace quantities of other metals that enhance the chlorinated VOC reaction kinetics. These bimetallic systems are largely (but not always) based on iron as the primary metal, with palladium, copper, nickel and tin possibly used as a secondary metal. The addition of these metals in trace amounts (as low as 0.03 percent by weight) enhances reaction kinetics in a catalytic fashion. Palladium has been the dominant catalyst metal used.

Additions to the bulk iron chemical composition are termed solutes. Solute chemicals cause significant changes in the physical and chemical properties of iron at relatively low (hundreds of parts per million) concentrations. As an example, the bottom-up colloid produced by aqueous precipitation with borohydride yields a product with 10- to 12-atom percent boron in the bulk of the colloid and 22-atom percent boron on the surface of the colloid.

Boron appears to be a solute in the iron system that may offer superior control over the rate at which iron dissociates water to form hydrogen compared to pure metallic iron. As discussed above, controlling the rate of disassociation is an important contributor to the longevity of nanoscale iron colloids in-situ. Chemical amendments offer further control on the reactivity of the bottom-up type of colloid. As an example of another solute that may affect performance, there is sulfur present in a variety of nanoscale iron manufactured by a third method, where nanoscale iron oxides are produced by attrition then partially reduced to iron metal with hydrogen. Sulfur compounds in these colloids appear to enhance portions of the chlorinated VOC dehalogenation pathway. In summary, top-down colloids offer reactivity control based on their structure while bottom-up colloids offer control based on their internal and surface composition.

Conclusion

Nanoscale iron colloids offer a unique in-situ treatment option for chlorinated VOCs. From a practical perspective, these materials must be affordable, available in ton quantities and deliverable. In 1998, the cost was $5,000 per kilogram, with production capacity in the kilogram range. Progress to date has been significant, with costs lowered more than an order of magnitude and readily available quantities in the 1,000-kilogram range. Given reasonable cost and availability, the physical ability to deliver the colloids is critical for in-situ applications.

The ideal size for delivery is in the range of 200 to 600 nanometers. Colloids of this size must be manufactured to produce enhanced reactivity that is not due to small size alone. In addition, corrosion processes are at work on the colloids in the in-situ groundwater environment, and the colloids must have the ability to provide continued reactivity through passivating coatings.

A colloid can be too reactive; the generation of hydrogen to the point of causing champagne-like fizzing is impressive, but not ideal for long-term in-situ applications. Nanoscale colloids can be used for in-situ treatment of dissolved plumes, treatment of source areas (where a highly reactive version of the colloid could be desirable), in conjunction with biological reductive treatment systems such as in-situ reactive zones, and with surface-based reactors. Current production methods offer nanoscale colloids with a means of reactivity and passivation control at costs and volumes that make a variety of field applications of the technology practical. PE