Patent Grant
7943084
The present invention relates to the surface treatment of metal powders with a chlorosilane compound to improve the flow characteristics and packing density of the powders.
Flowability is an important characteristic for all types of powders in industrial processes, including, for example: the manufacture of pharmaceuticals; powder metallurgy; the powder injection molding (PIM) of metal, ceramic and/or composite powders; the formulation of energetic materials; and other processes and applications involving the flow, mixing, dispersion, or fluidization behavior of powders. Flowability is often characterized by the flow rate of a specific mass or volume of powder that passes through a funnel with specified cone angle and aperture [Wanibe, Y.; Itoh, T. New Quantitative Approach to Powder Technology; J. Wiley: New York, 1998, Ch. 10.]. A related method is to vary the size of a funnel aperture to determine the minimum diameter through which the powder will freely flow [a) Taylor, M. K.; Ginsburg, J.; Hickey, A. J.; Gheyas, F. “Composite method to quantify powder flow as a screening method in early tablet or capsule formulation development.” AAPS Pharm Sci Tech, 2000, 1(3) article 18. [http://www.aapspharmscitech.org/pt0103/pt010318/pt010318.pdf last accessed Feb. 2, 2004. b) Gioia, Alberto “Intrinsic flowability: a new technology for powder-flowability classification.” John Morris Scientific. http://203.147.186.54/html/Hanson/flodex_report.htm last accessed Feb. 9, 2004.]. Another dynamic approach is to put a powder sample in a drum that rotates at a constant rate and measure the frequency with which the powder avalanches down the side of the drum [“Aero-Flow Powder Flowability Analyzer Model 3250” TSI Incorporated http://www.tsi.com/powder/products/3250/3250.htm last accessed Feb. 2, 2004.].
A static parameter known to correlate with flowability is the angle of repose of a powder pile, which reflects whether the particles cling together to form a steep pile or slide off one another to form a low mound [Bose, A. Advances in Particulate Materials; Butterworth-Heinemann: Boston, 1995, 300-302.]. Typically, the angle of repose parameter is measured by allowing powder to flow freely through a funnel onto a flat surface to form a pile. The angle between the flat surface and slope of the pile is the angle of repose. Dynamic tests of powder flowability may use specialized equipment, such as a computer controlled Jenike shear cell [Puri, V. M.; Ladipo, D. D. U.S. Pat. No. 6,003,382, Dec. 21, 1999.], and may require samples on the order of 100's of grams. By contrast, the angle of repose measurement requires no specialized equipment and can be adapted to a scale utilizing only a few grams of powder. While both angle of repose and shear flow tests generally differentiate powders based on their flow characteristics, the exact correlation between the tests is not known.
The flowability, or fluidity, of a powder depends on many factors. For example, Table 1 (taken from Rimai, D. “Particle-Substrate Interactions: Microscopic Aspects of Adhesion” NexPress Solutions LLC. Rochester N.Y.) lists both sample dependent and condition dependent factors.
1“The Nature of Powders” http://www.freemantech.co.uk/pages/natofpow/nature-powders.html last accessed Feb. 2, 2004.
At a microscopic level, the theory of powder flow is based on theories about adhesion, with the dominant theory being that of Johnson, Kendall, and Roberts (JKR) [Johnson, K. L.; Kendall, K.; Roberts, A. D. Proc. R. Soc. London: Part A, 1971, 324, 301.]. Basic assumptions of the theory include: (1) elastic deformation of spherical particles that vary the contact region, (2) a small contact radius compared to the particle radius, (3) only local interactions in the region of contact are considered—no long range forces are considered to act on the particles [Rimai, D. “Particle-Substrate Interactions: Microscopic Aspects of Adhesion” NexPress Solutions LLC. Rochester N.Y., http://www.clarkson.edu/projects/fluidflow/courses/me537/Rimai—1.pdf last accessed Feb. 2, 2004.]. The JKR theory has been applied to the study of contact surfaces between alkylsilane-coated aluminum and polymers (motivated by the importance of adhesion in the performance of composite materials) [Emerson, J. A.; Giunta, R. K.; Miller, G. V.; Sorensen, C. R.; Pearson, R. A. “The effect of surface contamination on cohesive forces as measured by contact mechanics.” Mat. Res. Symp. 2000, 629, 871-876.].
The cohesive force acting between two polymer surfaces is described as dispersive (non-polar, hydrophobic interactions). In contrast, for metal powders such as aluminum the cohesive forces are predominantly polar. The polar forces arise from hydroxyl group surface termination and physisorbed water molecules on the metal oxide layer that covers most metal surfaces (similar in nature to the well known hydrogen bond formation in molecular chemistry). A realized object of the present invention is to reduce the polar cohesion between metal powder particles by replacing surface hydroxyl groups with hydrophobic alkylsilanes. The prospect of using a hydrophobic surface-modifying agent to improve flow has been broached in a number of patents that describe the use of substituted chlorosilanes and substituted alkoxysilanes [a) Yukiko Ozaki, Satoshi Uenosono, Kuniaki Ogura, Iron-based powder composition for powder metallurgy having higher flowability and higher compactibility and process for production thereof, U.S. Pat. No. 6,503,445, issued Jan. 7, 2003. b) Yukiko Ozaki, Satoshi Uenosono, Kuniaki Ogura, Iron-based powder composition for powder metallurgy excellent in flowability and compactibility and method, U.S. Pat. No. 5,989,304, issued Nov. 23, 1999. c) Takaaki Kohtaki, Masaaki Taya, Masami Fujimoto, Developer containing insulating magnetic toner flowability-improving agent and inorganic fine powder, U.S. Pat. No. 5,547,796, issued Aug. 20, 1996. d) Tetsuya Kobayashi, Haruo Fujii, Motoi Katoh, (Yokohama, JP), Tatsuya Kobayashi, Toshiaki Miyashiro, Naoki Enomoto, Akihiko Uchiyama, Yoshiro Saito, Developing apparatus using one component toner with improved flowability, U.S. Pat. No. 5,307,127, issued Apr. 26, 1994. e) Helmut Knorre, Joachim Fischer, Gerhard Pohl, Process for preventing caking and obtaining flowability of alkali chlorides and salt mixtures thereof U.S. Pat. No. 4,107,274, issued Aug. 15, 1978. f) Helmut Knorre, Joachim Fischer, Gerhard Pohl, Preventing caking and obtaining flowability of alkali chlorides and salt mixtures thereof, U.S. Pat. No. 4,051,228, issued Sep. 27, 1977.]
In the direct reaction of chlorosilanes with hydroxy terminated metal surfaces, the expected products are the tethered or surface bound silanes and HCl. In the literature on modifying aluminum surfaces with alkylsilanes, frequent use is made of the condensation reaction of a silanol with a surface hydroxyl group to eliminate water and form a Si—O—Al linkage. The silanols are usually generated by hydrolysis of either chlorosilanes or alkoxysilanes. Volatile dialkyldichlorosilanes have been applied to metal oxide powders using a gas/solid reaction [Siegmar Laufer, Waldemar Roy, Process for the hydrophobization of higher dispersed oxides, U.S. Pat. No. 3,873,337, issued Mar. 25, 1975.]. Much of the work in the literature is directed to improving the bonding between aluminum and epoxies or polymers [a) Zhang M. C.; Kang E. T.; Neoh K. G.; Tan K. L “Surface modification of aluminum foil and PTFE film by graft polymerization for adhesion enhancement” Colloids and Surfaces A: Physicochemical and Engineering Aspects, 2001, 176(2), 139-150. b) Abel, M-L.; Rattana, A.; Watts, J. F. “Interaction of Epoxy Analogue Molecules with Organosilane-Treated Aluminum: A Study by XPS and ToF-SIMS” Langmuir, 2000 16(16), 6510-6518. c) Lee, I.; Wool, R. P. Thin Solid Films 2000, 379(1-2), 94-100.] and [Underhill, P. R.; DuQuesnay, D. L. “The reaction of water with abraded aluminum surfaces as studied by Fourier transform infrared spectroscopy” App. Surf Sci. 1999, 141, 138-140.]. Some pretreatments are reported to increase the number of surface hydroxyl groups available to react with the surface modifying silane [a) Kabayashi, G. S.; Donnelly, D. J. Boeing Co Report No. DG-41517, 1974. b) Boeing Process Specification BAC 5810, February 1991.]. Other pretreatments create Lewis acid sites on the alumina-like surface of the aluminum metal to increase the binding strength of surface modifying reagent [Puri, V. M.; Ladipo, D. D. U.S. Pat. No. 6,003,382, Dec. 21, 1999.]. In addition to alkoxy- and chlorosilanes, other surface modifiers have been tried, such as acid anhydrides [Schultz, J.; Lavielle, L.; Cane, A.; Comien, P., J. Mater. Sci. 1989, 24, 4363.]. However, the prior art does not disclose or suggest the direct reaction of chlorosilanes with hydroxy terminated metal oxide surfaces to improve the flow characteristics of metal powders such as aluminum, magnesium, titanium, aluminum-magnesium powders, and the like.
The novel process of the present invention provides a significant improvement in the flow behavior of reactive metal powders obtained by surface modification with alkylsilane reagents. The powders treated by this process become hydrophobic and are not readily wetted by water. Dramatically improved flow characteristics of the treated powders are demonstrated by lower angle of repose, higher pour density, higher packing density, and shorter times required for sieving. For example, the improvement in powder flow behavior was demonstrated by angle of repose measurements, which showed up to a 30° reduction for the treated powders. In one embodiment, the chlorosilane compound is diluted with a hydrocarbon solvent to facilitate efficient mixing with the metal powder. After mixing, the solvent and unreacted chlorosilane compound are separated from the metal powder by distillation or filtration. After drying, the treated metal powder may be sieved to remove small amounts of clumped material. Chemical analysis of the treated powders shows that small amounts (<0.05%) of silicon and chlorine are incorporated. In an alternative embodiment, a vapor phase treatment using a chlorosilane compound without a solvent may be used.
The process disclosed herein was readily scaled to 105-pound batches with a moderately high throughput. Further scale-up requires only larger scale equipment. Silane treated aluminum powders with nominal sizes of about 5-10 microns exhibited improved flow behavior, with higher solids loading under pressurized delivery and higher packing density.
Among a wide variety of chlorosilane reagents examined, the most effective to improve the flow characteristics of metal powders were trichlorosilanes with short-chain alkyl groups (e.g., methyl, isobutyl, tertiary butyl, and the like). Hanes with long-chain alkyl groups or only one chloro substituent were generally much less effective in altering the flow characteristics of metal powders. Accordingly, the trichlorosilanes with short-chain alkyl groups are the preferred surface modifying reagents for practicing the present invention. Repeated treatment of metal powders with silane compounds results in only slight additional improvement in flow characteristics. Gravimetric and Fourier Transform Infrared (FTIR) analysis indicate that only a small amount, estimated to be 0.1 weight percent or less, of a silane compound is incorporated in a treated powder. Thus, the improvement in powder flow characteristics provided by the present invention most likely results from a surface modification of the particle metal oxide layer in which interparticle hydrogen bonding interactions have been reduced by reaction of the chlorosilanes with the surface oxide layer hydroxyl groups. As the trichlorosilanes are most effective surface modifying reagents, binding of the silane to the particle surface at multiple sites may be an important factor in the improved flow behavior observed. Furthermore, tests on the effect of storage under high humidity show that the silane treated metal powders are less susceptible to degradation, as evidenced by clumping, than are untreated powders.
The improvement in metal powder flowability provided by the present invention has the potential for direct application to many processes and applications involving the flow or mixing behavior of metal powders, such as the manufacture of pharmaceuticals and lubricants, powder metallurgy, the powder injection molding (PIM) of metal, ceramic and/or composite powders, powder coating, composite material formulation and formulation of energetic materials. [a) Kubota, N. Thermochemical Aspects of Combustion; Wiley-VCH: New York, 2002, p. 76. b) Johnson, C.; Parr, T.; Hanson-Parr, D.; Hollins, R.; Fallis, S.; Higa, K. “Combustion and Oxidation of Metal Nanoparticles and Composite Particles” JANNAF 37th Combustion Subcommittee Meeting, Vol. 1. CPIA, November 2000, pp. 539-551.] Due to enhanced aerosol formation, the metal powders provided by the present invention may also be useful in fluidization processes and certain combustion applications, including the combustion synthesis of metal oxide powders. The term “fluidizability” as used herein shall mean the ease with which a powder may be made more fluid. It is known to those of ordinary skill in the art of the subject invention that a powder may be fluidized, such as in a fluidized bed or aerosol, for example. The improved dispersion properties of micron-sized or nano-sized powders provided by the present invention may render them useful in inkjet printing applications.
Our work with hydrophobic coatings on glass suggested to us that alkyl and perfluoro-alkyl chlorosilanes would be good candidates for investigation as metal particle surface modifying agents [Johnson, C. E.; Harris, D. C.; Nelson, J. G.; Kline, C. F.; Corley, B. L. “Strengthening of Glass and Pyroceram with Hydrophobic Coatings” NAWCWD TP 8536, April 2003.]. Because the chlorosilanes cannot undergo auto-condensation reactions, they are more stable than silanols. Such stability is an important consideration with regard to shipping and storage. In dry solvents, the chlorosilanes react directly with hydroxyl terminated metal oxide surfaces to form the Si—O—Al linkage and HCl, or react with physisorbed water to form the silanol which can undergo either auto-condensation to form oligomers, or condensation with the hydroxyl terminated surface to form more linkages. In wet solvents, a greater amount of oligomerization is to be expected. The number of potential bonds or linkages to stabilize the alkylsilane on the metal oxide surface increases proceeding from mono- to di- to trichloro alkylsilanes. For monochloro alkylsilanes auto-condensation yields only unreactive dimers, but the di- and trichloro alkylsilanes can form more complex siloxane chains and three dimensional siloxane nets, which can have multiple bonds or linkages to the metal oxide surface. In wet solvents, the di- and trichloro alkylsilanes could possibly form particles of the bulk siloxane [Rochow, E. G.; Gilliam, W. F “Polymeric Methyl Silicon Oxides” J. Am. Chem. Soc. 1941, 63, 798-800.].
Advances made in understanding how changing the surface modifying agent, solvent and metal powder affect the flowability of metal powders are disclosed herein in connection with the process of the present invention. The metal powders examined in connection with the present invention include aluminum powder (ranging in size from a nominal 2 microns to a nominal 10 microns in diameter), magnesium powder, magnesium-aluminum alloy, and titanium powders. It will be understood by one of ordinary skill in the art of the present invention that powders of other sized particles and of other metals and materials having hydroxyl terminated surfaces may be successfully treated by the process of the present invention and are within the scope presently disclosed invention. Alkylchlorosilanes, such as methyltrichlorosilane, are used as preferred particle surface modifying reagents. It is another realized object of the present invention to provide a simple, reliable and inexpensive process that is amenable to a high throughput scale-up, to handle powder lots on a scale of tens to hundreds of pounds, utilizing readily available reagents and equipment to provide significantly improved metal powder flow characteristics.
As used herein, the terms “tethered”, “linkage”, “linked”, “bound”, “binding”, “bond” and “bonded” shall be understood to refer to an organic substituent, including those shown in
Diagram 1 in the text is a diagram of the chemical structures of 12 representative surface modifying compounds.
Aluminum powders were produced by Valimet, Inc. The nominal diameters of the aluminum particles are: 10 micron (H-10 Al); 5 micron (H-5 Al); 3 micron (H-3 Al); and, 2 micron (H-2 Al). Alloy powders were also purchased from Valimet: Al/Mg 50:50 (−270 mesh) and Al/Mg 70:30 (−325 mesh). Magnesium and titanium powders, both −325 mesh, were purchased from Alfa-Aesar. The sample of −200 mesh Mg powder was produced by Hart Metals, Inc.
The silane-based hydrophobic reagents n-octadecyldimethylchlorosilane, (tridecafluoro-1,1,2,2-tetrahydrooctyl)methyldichlorosilane, Aquaphobe CM® (oligomerized dialkyldichlorosilanes), and Aquaphobe CF® (a mixture of oligomerized dichloro di(perfluoroalkyl) silanes) were purchased from Gelest Inc. Silanes purchased from Sigma-Aldrich are octadecyltrichlorosilane, chlorotrimethylsilane, methyltrichlorosilane, tert-butyltrichlorosilane, tert-butyl(chloro)dimethylsilane, isobutyltrichlorosilane, propyltrichlorosilane, trichloro (3,3,3-trifluoropropyl)silane and octadecyltrichlorosilane. In addition, (3,3,3-trifluoropropyl) dichloromethylsilane was purchased from Oakwood and dichlorodimethylsilane from Supelco. Trifluoroacetic anhydride was purchased from Sigma-Aldrich. All reagents were used as received. Solvents (typically 99.9% anhydrous grade) were used as received from Sigma-Aldrich.
With reference to Diagram 1, the structural comparisons of selected surface-modifying reagents are shown.
Methods of Metal Powder Treatment
Method I. About 1.5 grams (g) of aluminum powder was placed in a 12×75 millimeter (mm) test tube that was filled with a solvent. A silane reagent was added to the test tube to give about 2-10 weight percent solution and the tube was capped with a septum. Vigorous agitation was applied several times over the course of approximately 30 minutes (min). The test tube was then centrifuged, the supernatant siphoned off, and the treated powder washed with solvent in the test tube. After centrifuging again and siphoning off the solvent rinse, the powder was dried at 110° C. for a few hours in an atmospheric oven, before collecting it for an angle of repose measurement.
Method II. About 3 g of metal powder was placed in a 30 milliliter (mL) Gooch crucible and about 0.5 Molar (M) solution of a silane reagent (or other surface modifying reagent) was added, with some stirring, to cover the powder. After about 15 min, the reagent solution was filtered off and the sample rinsed on a glass frit with solvent. After the wash was filtered off, various drying procedures were applied to the samples: (1) air dry at room temperature; (2) air dry at up to 110° C.; or, (3) dry in a vacuum oven at up to 150° C. Portions of the samples were dried for various times from 15 min to overnight. To ascertain the best conditions for an oven treatment, 45 g of H-10 Al (10 micron) (lot 01-2149) was treated with methyltrichlorosilane in heptane, and then divided into three portions. One portion was placed in a vacuum oven at about 145° C., a second portion was placed in a conventional oven at 110° C. under air, and a third portion of the sample was allowed to dry at room temperature, over a period of hours, on the fritted funnel. A few grams were removed from each sample at regular intervals for angle of repose measurement.
Method III. Two 2.2-kg batches of H-10 Al (lots 00-1143 and 02-3019) were treated with methyltrichlorosilane in either pentane or hexane, following a modification of Method II. The H-10 Al powder was divided into 1.1 kg portions and each placed in a 2.8 liter (L) fitted funnel (medium or fine frit). All treatments were carried out by slurrying the powder on the frit and then vacuum filtering to remove the liquid. The H-10 Al powder was pretreated with water, washed three times with acetone and once with pentane (or hexane), then treated with about 0.5 L of a 0.2 M solution of methyltrichlorosilane in pentane (or hexane) and washed once with the alkane solvent. The treated H-10 Al was broken up thoroughly with a spatula and dried for several hours under vacuum at 130-150° C. The dried aluminum powder was sieved to −200 mesh.
Method IV. Method III was modified somewhat for subsequent treatments of Al powders in 2.3-kg batches. The Al powder was manually mixed in a 10-L beaker with a 0.2 M methyltrichlorosilane/heptane solution (prepared with 47 mL of the silane and 2 L of heptane). The mixture was transferred to a 4-L medium-frit filter funnel, vacuum filtered, and washed once with 2 L of heptane. During the washing step the solid (metal powder) was transferred back to the beaker for easier mixing. The washed solid (metal powder) was transferred to beakers and dried overnight under vacuum at 110° C. The dried powder was sieved to −200 mesh.
Method V. For treatment of up to 15 kg batches of H-5 Al (5 micron) powder, the mixing was conducted in an 18 L stainless steel slurry kettle, equipped with a pneumatically actuated anchor blade stirrer and a water-circulating jacket. In a typical experiment, 12 kg of aluminum powder was added from plastic bags through a narrow port in the kettle top. Eight liters of the 0.02 Molar methyltrichlorosilane (prepared by adding 4.7 mL portions of methyltrichlorosilane to 2 L bottles of anhydrous hexane) was poured into the kettle. The mixture was stirred at 90 revolutions per minute for 15 min. In order to distill out the hexane and unreacted methyltrichlorosilane, water at 90-95° C. was circulated through the kettle jacket. Within 10 min the temperature of a probe in the kettle reached 70° C., where it held while the solvent distilled out and was collected through a condenser and a collection bottle. When condensate stopped forming, the temperature increased to 76-95° C., where it was held for 30-90 min. The treated powder was poured or scooped from the kettle. Since this treatment procedure was conducted in a ventilated room instead of a fume hood, protective gear worn during handling the powder included a full-face respirator with filters appropriate for dust and organic chemicals, gloves, and coveralls. The treated powder was sieved through a stack of 100, 200, and 270 mesh sieves. The scale of the Method V procedure was increased to 48 kg batches using a 95 L capacity mixer.
Method VI (used with Mg). Another variation on method II was used to treat six 2.4-kg batches of magnesium powder. Different lots of Mg powder (−325 mesh, lot numbers F20N13, D23M05, C14N27) were mixed in the same proportion for each batch and treated with a 0.02 M solution of either methyltrichlorosilane or isobutyltrichlorosilane dissolved in anhydrous heptane. The powder was mixed with the silane solution in a 10-L beaker for 15 min, and poured into a 4-L medium-frit funnel for vacuum filtration. After filtering off the silane solution, the treated powder was transferred to bottles or beakers (without a solvent rinse) and dried for 6 hours (h) under vacuum at 120° C.-150° C. The dried magnesium powder was sieved to −325 mesh.
Method VII. A vapor phase treatment was conducted by passing hot methyltrichlorosilane vapor through a filter funnel containing aluminum powder placed on a flitted glass disk. The filter funnel was heated to 200° C. in an oil bath. A stream of nitrogen was first passed through a bubbler containing methyltrichlorosilane, and then through the filter funnel.
Moisture sensitivity test. A portion of the 45-g sample of silane-treated 1-1-10 Al (lot 01-2149), which had been heated in the atmospheric oven, was placed in a humidor over a saturated solution of cupric sulfate with an average relative humidity 88%, compared with the 25-30% relative humidity typical of ambient laboratory conditions. Portions of this sample were removed and analyzed for angle of repose.
Angle of Repose Measurement
An American Society for Testing and Materials (ASTM) standard procedure fixes the height of a cone of powder falling from a funnel and measures changes in the cone's diameter, calculating the angle of repose from a simple trigonometric relationship [C1444-00 Standard Test Method for Measuring the Angle of Repose of Free-Flowing Mold Powders: ASTM International, West Conshohocken, Pa., 2002.]. In this work, the method has been adapted to samples of a few grams, instead of the pound or more called for in the ASTM standard method. The diameter of the powder cone is fixed by catching the powder poured through a Pasteur pipet in a shallow cylinder, as illustrated in Diagram 2 [http://rxsecure.umaryland.edu/courses/PHAR535/lecture_handouts/Labs/Particle-lab.pdf last accessed Feb. 2, 2004]. The angle of repose is calculated from the measured height when the powder cone no longer increases in height as additional powder poured through the pasture pipet slides off the edge of the dish. The dish sits on weighing paper so any spilled powder can be collected and, along with the powder in the dish, reused to make a subsequent measurement. Prior to angle of repose analysis, a vial containing a few grams of powder was vigorously shaken for about 30 seconds to break up clumps. Each powder was measured seven times (three times for some high angle of repose powders), the results averaged, and the 95% confidence limit calculated. In general, an angle of repose for a powder does not have a value that is established with respect to a reference standard, and the values are reported in terms of 95% confidence intervals [C1444-00 Standard Test Method for Measuring the Angle of Repose of Free-Flowing Mold Powders: ASTM International, West Conshohocken, Pa., 2002.]. The term “angle of repose” shall be understood to refer to measurements made by the above-described method. Because the amount of powder used to make the measurement affects the precision of the measurement, the confidence intervals are two to three times greater than the ASTM standard.
The setup in
The quantitative results are also somewhat dependent on the detailed procedure for the more free-flowing powders (with an angle of repose that is about 30 degrees or less), because a rapid flow of powder can flatten the pile upon impact. For example, for a powder that exhibited a 23° angle of repose by the usual procedure, an angle of repose of 31° was obtained by slowly adding small amounts of powder to the pipet.
The treated powders with low angle of repose tend to become remarkably fluid-like when aerated by brief shaking, yet are rather resistant to flow when at rest and not aerated. The treated powders also have a greater tendency to form suspended dust. Flow rate measurements of treated H-5 aluminum powders by the method of ASTM B213-03 (Standard Test Method for Flow Rate of Metal Powder: ASTM International, West Conshohocken, Pa., 2003) were not successful due to inadequate flowability of the powder when at rest. There is a general tendency of these powders to continue in motion once set in motion. When larger amounts of powder flow out of a funnel the resultant pile has a wide base and slope that is relatively steep near the middle of the pile, but shallow along the vast outer parts of the pile. Due to this behavior, angle of repose measurement using the procedure of ASTM C 1444-00 will give very low values compared to the maximum slope that the powder will maintain. For example, a treated H-5 aluminum powder with a 39° angle of repose measured by the pipet method gave a value of 22° when a bulk sample flowed out of a funnel.
Gravimetric Analysis
To check for significant mass change upon silane treatment, careful measurements of the aluminum powder mass were conducted. Four samples of H-10 Al lot 01-2149 were treated with chlorotrimethylsilane, methyltrichlorosilane, trifluoroacetic anhydride and tert-butyl(chloro)dimethylsilane, respectively. At each step in the process, the samples were weighed five times and average masses and 95% confidence intervals determined.
Thermogravimetric analysis was conducted on a TA Instruments Hi-Res TGA2950 Thermogravimetric Analyzer. Analysis was carried out on H-10 Al (lot 01-2149) treated with tert-butyl(chloro)dimethylsilane in heptane, an untreated sample of H-10 Al (lot 01-2149) and an empty reference pan. The heating profile was: hold 30 min at room temperature (except for no room temperature hold on the treated sample), ramp at 20° C./min to 400° C., hold for 960 min at 400° C.
Scanning Electron Microscopy
Scanning electron microscope (SEM) images were collected with an Electroscan Environmental Model E-3 microscope. The samples were prepared by taking a few milligrams of powder and placing it on the middle of a sample holder (a small aluminum disk with a piece of carbon double stick tape on it). The powder was smeared out, using light pressure, with a spatula to ensure the particles adhered to the tape without large clumps (the imaging was of particles at the edge of the smear). Then, the sample was sputter coated with a layer of gold, about 15 nanometers. The digital images were adjusted for brightness and contrast.
Diffuse Reflectance Fourier Transform Infrared Spectroscopy
The diffuse reflectance, Fourier transform infrared (FTIR) spectra were collected with a Foundation Series diffuse reflectance accessory on a Thermo Nicolet 870 spectrometer, utilizing a mid-band, mercury-cadmium-telluride (MCT) detector. Each spectrum consisted of the Fourier transform of 256 co-averaged interferograms with nominal resolution of 8 cm−1 and multiplied by a Happ-Genzel apodization function. The single beam spectra of sample aluminum powders were ratioed to the single beam spectrum of an untreated aluminum powder reference to give the reflectance of the sample. The reflectance was scaled to unity in the featureless region of 2000-2500 cm−1. The scaled reflectance was converted to Kubelka-Munk units of remission with the standard formula: Reflectance, R=Rsample/Rreference; Kubelka-Munk Function, F(R)=(1−R)2/(2*R). Remission bears a similar relationship to reflectance as absorption does to transmission and can be viewed as directly comparable to an absorption spectrum. The wavenumber axes of the experimental data were scaled to those of literature infrared absorption spectra for convenient comparison.
Ion Chromatography for Chlorine Analysis
Samples for chlorine analysis were prepared as follows. One gram of aluminum powder was added to 20 mL of 50:50 water:ethanol solution, the mixture was sonicated for 1 h, then filtered through a 0.45 micron filter, then diluted with water to 50 mL. The concentration of chlorine was quantified by ion chromatography analysis.
Inductively Coupled Plasma-Atomic Emission Spectroscopy (ICP-AES) for Silicon and Trace Metal Analysis
For Si and trace metal analysis, 0.5 g of aluminum powder was first “wet” with a small amount of 10% ethanol/water solution, then digested in nitric acid using microwave digestion, then the samples were diluted with water. The concentrations of Si, Fe, Ni, and Ti were quantified by ICP-AES analysis.
Tap Density
Tap density of aluminum powders was determined by manually tapping a 50 mL glass graduated cylinder against a hard bench top, approximating the method described in ASTM Standard B 527-93 (Standard Test Method for Determination of Tap Density of Metallic Powders and Compounds, Reapproved 2000: ASTM, West Conshohocken, Pa., 2000), where the drop height is specified as 3 mm and the tapping frequency is between 100 and 300 taps/minute. Prior to weighing a sample for analysis, the powder was vigorously shaken in its container for about 30 seconds to break up clumps. The sample weight was typically 43 g for aluminum powders.
Discussion of Examples and Equivalents
An object of the present invention is to improve the flow characteristics of reactive metal powders. The H-10 aluminum powders from lots 00-1179 and 00-1134 were observed to have relatively good flowability, however, H-10 Al from lots 01-2149 and 02-3019 were observed to have relatively poor flowability. The results in Table 2 show that the angle of repose measurement differentiates between the powders that showed relatively good flowability (having approximately 40° angle of repose) and those which did not (having an angle of repose in the range 50°-55°. This is consistent with the correlation that a powder with greater flowability will have a smaller angle of repose than a powder that is less fluid. Powders with good flow behavior tend to have a low angle of repose, a low surface area, and a high flow index from shear flow tests. However, none of these characteristics alone is a reliable indicator of good flow performance. The H-10 Al powders were compared with untreated powders of smaller size and it is shown in Table 2 that the smaller diameter particle powders generally have higher angles of repose, with values similar to the highest values among the H-10 lots.
One known strategy to improve flowability of a powder is to add a “glidant” such as finned silica [Mei, R.; Klausner, J. F.; Shang, H.; Kallman, E. “On the improved flowability of cohesive powders by coating with fine particles” in Processing and Handling of Powders and Dusts; Ed.s Battle, T. P.; Henein, H.; TMS: Warrendale Pa., 1997, 225-236.]. Fumed silica was employed for this purpose in a 1% mixture with H-10 Al, lot 01-2149. The mixture appeared more fluid than the plain powder when poured and had a lower angle of repose, 45.6°±2.1°, compared to 55.2°±2.9° for the untreated powder, as is shown in Table 3.
A second strategy to improve powder flowability is to reduce the cohesive forces between the particles by treating the surface chemically. Chemical treatments are routinely done to reduce the adhesive forces between solid surfaces and liquids, such as water, by treating the surface with hydrophobic coatings [Johnson, C. E.; Harris, D. C.; Nelson, J. G.; Kline, C. F.; Corley, B. L. “Strengthening of Glass and Pyroceram with Hydrophobic Coatings” NAWCWD TP 8536, April 2003.]. In initial experiments aluminum powders were treated with silane reagents and then washed with solvent using a centrifuge for separation according to Method I. In some later experiments the powders were washed using filtration according to Method II, since this approach is more amenable to scale up.
Tables 3, 4, and 5 report the results for treated aluminum powders, showing that the angle of repose was decreased by as much as 30 degrees as a result of silane treatment. Table 3 lists angle of repose measurements (an average value is reported) for H-10 Al (lot 01-2149) treated with various reagents. Based on this initial screening of alkylchlorosilanes, methyltrichlorosilane was selected for further investigation because of its high efficacy in lowering the angle of repose, its relatively low cost and convenient properties as a volatile liquid. Table 4 reports further details for experiments using methyltrichlorosilane in which such parameters as the scale, solvent or adding a pretreatment, and the like are varied.
Surprisingly little, if any, lowering of the angle of repose was measured for samples of H-10 Al treated with commercial hydrophobic reagents (as shown in the entries for Aquaphobe CM® and Aquaphobe CF® in Table 3) or treated with alkylchlorosilanes with long alkyl chains (as shown in the entries for n-octadecyltrichlorosilane, n-octadecyldimethylchlorosilane and (tridecafluoro-1,1,2,2,tetrahydro-octyl)methyldichlorosilane in Table 3).
Table 5 shows the results of alkylsilane treatments of the smaller particle size aluminum powders.
aUntreated powder was too clumpy to flow through the angle of repose apparatus.
Diffuse reflectance FTIR spectroscopy was used to characterize the silane-treated aluminum powders. The diffuse reflectance spectrum for the n-octadecyldimethylchlorosilane treated powder is presented in
The high angles of repose for the n-octadecyldimethylchlorosilane, n-octadecyltrichlorosilane and (tridecafluoro-1,1,2,2,tetrahydrooctyl)methyldichlorosilane treated powders indicates that the treated powder's cohesive force is comparable to that of the untreated powder. This suggests that long alkyl chains may create a cohesive force that is dispersive in nature (i.e., non-polar interactions), and similar in magnitude to the predominantly polar cohesive forces acting between untreated aluminum particles. The performance of the commercial waterproofing silanes, however, require a different explanation. Aquaphobe CM® has only methyl groups attached to its siloxane backbone, and it failed to improve the angle of repose. A possible explanation for this observation is a reduced coverage of the particle surface by the silane.
In contrast to the long-chain alkyl chlorosilanes, trichlorosilanes with short chain lengths were very effective in lowering the angle of repose of treated metal powders. Methyltrichlorosilane treatment in particular yielded powders that appear very fluid when poured and that have an average angle of repose of about 31°, with a minimum value of 26°. The FTIR spectrum of methyltrichlorosilane treated H-10 Al powder, shown in
The trend of decreasing angle of repose with increasing number of chlorides attached to the silane is also consistent with the results for (3,3,3-trifluoropropyl)dichloromethylsilane (angle of repose 48.7°±2.5°) and trichloro(3,3,3-trifluoropropyl)silane (angle of repose 34.6°±3.9°). However, this trend is not observed in the case of tert-butyl(chloro)dimethylsilane (angle of repose 29.3°±2.3°) and tert-butyltrichlorosilane (angle of repose 29.5°±2.2°), whose angles of repose are indistinguishable. The FTIR spectrum of the H-10 Al powder treated with tert-butyl(chloro)dimethylsilane, shown in
Another dimension of the silane connectivity is the ability of the di- and trichlorosilanes to form complex siloxane oligomers or polymeric material. As noted above, the FTIR results do not rule out the presence of siloxane oligomers. To assess any significant change in mass following silane treatment, which would be indicative of the formation of a bulk second phase polymeric siloxane, four approximately 3 g samples of H-10 Al (lot 01-2149), were treated with different surface modifying reagents. The results are as follows for the change in mass of sample upon treatment (±the 95% confidence level): chlorotrimethylsilane (−1.8 mg±3.0 mg), methyltrichlorosilane (−3.8 mg±5.1 mg), trifluoroacetic anhydride (−2.0 mg±2.1 mg) and tert-butyl(chloro)dimethylsilane (−2.4 mg±2.3 mg), respectively. Since the confidence limits are close to or exceed the mass loss, the mass change upon treatment may be considered negligible. Similar results were obtained in TGA experiments that showed no significant difference among weight changes for TGA runs with an empty pan and a pan loaded with untreated or tert-butyl(chloro)dimethylsilane treated H-10 Al. Based on an estimation of approximately 0.1% by weight surface bound silane, the expected mass changes would be near the sensitivity limit of the TGA instrument for 10 mg samples. The TGA results, taken together with the gravimetric results demonstrate that there is no significant retention of a bulk second phase of siloxanes in the treated H-10 Al samples. Further, the SEM images shown in
Simple carboxylates, such as those derived from acetic acid [de Cheveigne, S.; Gauthier, S.; Klein, J.; Leger, A. “Adsorption of marked acetic acid on alumina studied by LETS” in Vibrations at Surfaces: Ed.s Caudano, R.; Gilles, J.-M., Lucas, A. A: Plenum, New York, 1982, 465-470.], and acid anhydrides [Schultz, J.; Lavielle, L.; Carre, A.; Comien, P., J. Mater. Sci. 1989, 24, 4363] are known to adsorb on alumina surfaces. The 1′-10 Al (lot 01-2149) exposed to trifluoroacetic anhydride in heptane for 15 min showed no significant change in its angle of repose (55.1°±2.8°). The scaled FTIR spectrum (not shown) did not exhibit any significant differences from the scaled blank FTIR, thus it may be concluded that the carboxylate groups did not attach to the surface of the particles sufficiently to alter inter-particle cohesive forces.
The number of surface binding sites is a factor that affects the binding of surface modifying agents. In the case of silane, it may be possible to increase surface coverage by increasing the number of hydroxyl groups on the particle surface by pretreatment with an acidic or alkaline wash. This approach has been reported for the preparation of silane coated aluminum surfaces, utilizing aqueous solutions at pH 5 or pH 9 [Underhill, P. R.; DuQuesnay, D. L. “The reaction of water with abraded aluminum surfaces as studied by Fourier transform infrared spectroscopy” App. Surf. Sci. 1999, 141, 138-140.]. Methyltrichlorosilane treated H-10 Al (lot 01-2149) powder, pretreated with pH 9 ammonium hydroxide solution (angle of repose 31.8°±2.9°), or with pH 7 water (angle of repose 32.4°±2.7°), does not have a different angle of repose compared with matched samples treated only with methyltrichlorosilane in heptane.
The high angle of repose for the (tridecafluoro-1,1,2,2,tetrahydrooctyl)methyldichlorosilane treated powder (angle of repose 56.1°±2.6°) suggests that the powder's cohesive forces are not mitigated by fluorination of the alkylsilane. Likewise a comparison of n-propyltrichlorosilane (angle of repose 34.5°±6.5°) treated powder with trichloro(3,3,3-trifluoropropyl)silane (angle of repose 34.6°±3.9°) treated powder shows no change in the angle of repose. The Aquaphobe CFS (angle of repose 47.3°±1.5°) treatment does show improvement compared to Aquaphobe CM® (angle of repose 56.0°±1.8°), however, uncertainty as to the relative degree of oligomerization of the Aquaphobe products as well as uncertainty regarding the differing lengths of the carbon side chains means these results are not strictly comparable. The results for H-5 aluminum (lot 87-8204) coated with the fluorinated polymer KEL-F (angle of repose 48.7°±2.6°) also showed a modest reduction of the angle of repose, but lesser in degree than that of H-5 Al treated with methyltrichlorosilane. These data suggest the conclusion that perfluoroalkyl groups offer no advantage over alkyl groups in reducing the angle of repose of aluminum powders of the sizes examined.
Five-pound batches of H-10 Al were successfully treated with methyltrichlorosilane by Method III to yield powders that had reduced angles of repose, (31.6°±2.4° versus 50.5°±3.2° for untreated lot 00-1143, and 29.0°±2.4° versus 53.8°±1.9° for untreated lot 02-3019). Testing of these treated samples showed improved flow characteristics. In these treatment experiments, the limiting process steps for increasing the throughput for the five-pound or larger scale treatments were the aqueous pretreatment and the use of a small vacuum oven to dry the treated powders. Since, as discussed herein, an aqueous pretreatment does not significantly effect the change in the angle of repose, this step was eliminated in subsequent treatments along with elimination of the washings with acetone and hydrocarbon, as set forth in Method IV.
With further reference to Table 5, treatment of smaller sized aluminum particles with the aforementioned surface modifying agents was usually very effective in reducing the angle of repose measured for H-5 Al, sometimes effective in reducing the angle of repose measured for H-3 Al, and not effective in reducing the angle of repose measured for the one lot of H-2 Al examined. A second treatment of H-3 (lot 01-2142) did show a further modest reduction of the angle of repose. However, a second treatment of the H-2 Al with methyltrichlorosilane did not further change the angle of repose.
The silane treatment process may be useful in improving the mixing and dispersion behavior of nano-sized powders, which tend to incompletely disperse and form high viscosity mixes due to their high surface area. Three nano-sized aluminum powders (110, 138, and 165 nm average diameter) were treated with a 0.2M iso-butyltrichlorosilane solution in heptane. The 138-nm powder was first treated according to Method II on a 2.3 g scale. The treatment was then conducted according to Method III on a 50 g scale, where there appeared to be some gas evolution 5 min after the reagents were mixed. The 110 nm and 165 nm powders were then treated according to Method IV on a 50 g scale, with the mixing conducted at 0° C., where no gas evolution was observed.
In most experiments the concentration of the silane reagent in the treatment solution had little effect on the angle of repose of treated H-3 and H-5 samples, with 0.02 M silane solutions about as effective as more concentrated silane solutions in reducing the angle of repose. However, in some large scale treatments of H-5 aluminum a 0.08M concentration of methyltrichlorosilane was more effective than the usual 0.02M concentration. The results for H-3 Al are shown in Table 5. It will be appreciated that the flow characteristics of methyltrichlorosilane treated H-3 Al powder are similar to those of methyltrichlorosilane treated H-10 Al powder, as high flowability aluminum powders were not previously available in sizes below 10 micrometers.
An improved procedure, Method V, was developed to treat aluminum powders on a 15 kg scale using a slurry kettle. In the procedure of Method V the solvent and any excess methyltrichlorosilane are distilled off rather than being separated from the powder by filtration. To facilitate solvent removal and product drying, hexane (boiling point 69° C.) is used in Method V rather than heptane (boiling point 98° C.), which is used in Method IV. Any residual methyltrichlorosilane (boiling point 65° C.-66° C.) distills off with the hexane. Method V yields a superior treated metal powder having angles of repose as low as 21°, compared to 42°±1.8° for the same lot of aluminum powder treated according to Method IV, as is also shown in Table 5. Additional drying in an oven occasionally resulted in a reduced angle of repose of treated metal powders prepared by Method V. In one case heating a treated H-5 aluminum powder overnight at 123° C. resulted in the angle of repose decreasing from 45° to 37.5°. The slurry kettle procedure is successful in reducing the angle of repose and in reducing labor and material costs. The additional reduction in angle of repose of treated metal powders prepared by Method V may result from the powder being stirred during the drying process, and/or from the surface modifying agent being heated to 69° C. while in contact with the aluminum particles.
Silane treated H-5 aluminum powders prepared by Method V were analyzed to determine the extent of incorporation of silicon and chlorine. The chlorine content (determined by a solvent extraction process) ranged from 0.012-0.015% for three powders treated with 0.02M methyltrichlorosilane solution. Chlorine was not detected in the untreated powder (level below 0.0024%). The total chlorine available from the silane reagent was 0.11%, indicating that approximately 10% of the available chlorine was retained in the treated powder. When the methyltrichlorosilane concentration was increased to 0.08M in the treatment process, the incorporated chlorine content in the treated powder increased to 0.021-0.023%.
Silicon content was determined by completely digesting aluminum powder samples in nitric acid. Untreated H-5 aluminum powders contained about 0.035% Si, while 0.02M methyltrichlorosilane treatment led to approximately 0.043% Si, and 0.08M methyltrichlorosilane treatment led to approximately 0.046% Si. The amount of Si attributed to the silane treatment is approximately 0.008-0.011%, which compares to 0.031% Si available from a 0.02M methyltrichlorosilane treatment. The combined amount of Si and Cl added by the treatment process is approximately 0.02-0.04%, indicating that the amount of the added silane material is less than 0.1% of the powder mass.
The flow characteristics of magnesium-aluminum alloy and magnesium powders were also examined. Table 6 lists the angles of repose for untreated samples of magnesium powder and two different magnesium-aluminum alloy powders. Table 6 shows that there is lot to lot variance between nominally equivalent powders from the same distributor. The −200 mesh sized Mg and Al-30% Mg powders had the lowest angles of repose.
The typically high angles of repose for magnesium powder and the magnesium-aluminum alloy powders are dramatically reduced by treatment with the surface modifying agents, as is shown in Tables 7, 8 and 9.
With reference to Table 9 the effect of other short-chain alkyl substituents on flowability is shown.
Unlike the treated aluminum powders that are relatively insensitive to drying conditions, the angle of repose of treated magnesium powder and treated magnesium-aluminum alloy powders is sensitive to the drying process, as can be seen by comparing the different heating conditions in Table 7 for the 50% magnesium-50% aluminum alloy powder and in Table 10 for the magnesium powder. In other experiments, the silane treated metal powders were heated under vacuum for various times in the temperature range of 110° C.-150° C. Trichlorosilanes with short attached alkyl groups reduced the angle of repose (with exceptions attributed to insufficient heating). However, the angle of repose was not reduced by treatment of Al-50% Mg powder with the monochlorosilane chlorotrimethylsilane (angle of repose 50.4°±1.6°). Neither was the angle of repose reduced by treatment of Mg powder with the long-chain alkylsilane octadecyltrichlorosilane (angle of repose 51.5°±2.4°).
With reference to Table 11 the effect of reducing the concentration of silane in solution on flowability is shown.
For scale-up to 2.4 kg batches of magnesium powder, both methyltrichlorosilane and iso-butylchlorosilane were used at a concentration of 0.02 M in anhydrous heptane, with drying for six hours under vacuum in a temperature range of 110° C.-150° C., and sieving the final product to −325 mesh according to Method VI. The silane treated −325 mesh Mg powder exhibited much better flow characteristics than did the untreated −325 mesh Mg powder. While the reduction in the angle of repose achieved is similar between the methyltrichlorosilane and iso-butylchlorosilane reagents, the iso-butyltrichlorosilane solution formed a more fluid suspension that was easier to transfer from the 10-L beaker to the 4-L medium-frit filter, and so offers a small advantage for further scale-up.
A variety of organic solvents were used in the silane treatment experiments, including pentane, hexane, hexanes, heptane, toluene, dichloromethane, and acetone. The acetone solutions of alkyltrichlorosilanes appeared chemically unstable, but were still effective at reducing the angle of repose of Mg and Al—Mg powders. Other solvents that do not react excessively with the silane reagent may be used.
In addition to the data tabulated in Tables 1-11 and shown in
The possible deterioration of powder flow behavior under high humidity was also examined. Clumping was so severe in the case of untreated H-10 Al (lot 01-2149) exposed to 88% relative humidity for one month, that an angle of repose measurement could not be made. In contrast, a sample of H-10 Al (lot 01-2149), treated with methyltrichlorosilane and stored under 88% humidity for one month showed no clumping or loss of flowability (angle of repose 31.4°±3.7°). A preliminary experiment comparing the angle of repose of tert-butyltrichlorosilane treated Mg powder before (average angle of repose 33.6°±2.3°) and after (average angle of repose 35.1°±2.0°) overnight exposure to 88% relative humidity shows that the flow characteristics are stable to brief exposures to high humidity. Moisture in the solvent or physisorbed water on the metal powders was not specifically controlled in these experiments, although it is known to affect silane coverage on glass [McGovern, M. E.; Kallury, K. M. R.; Thompson, M. “Role of Solvent on the Silanization of Glass with Octadecyltrichlorosilane” Langmuir 1994, 10, 3607-3614.]. Consistent with the enhanced stability to moisture, the treated powders were hydrophobic and could only be suspended in water with the aid of a surfactant.
An important aspect of the enhanced flowability of the treated metal powders is a substantial increase in both pour density and packing density.
It will be apparent to one of ordinary skill in the art of the present invention that many similar surface modifying reagents in addition to those identified in
While the present invention has been described in connection with what are currently considered to be the most practical and preferred embodiments, it is to be understood that the invention is not to be limited to the disclosed embodiments, but to the contrary, is intended to cover various modifications, embodiments, and equivalent processes included within the spirit of the invention as may be suggested by the teachings herein, which are set forth in the appended claims, and which scope is to be accorded the broadest interpretation so as to encompass all such modifications, embodiments, and equivalent processes.
The invention described herein may be manufactured and used by or for the government of the United States of America for governmental purposes without the payment of any royalties thereon or therefor.
| Number | Name | Date | Kind |
|---|---|---|---|
| 2917406 | McBride | Dec 1959 | A |
| 3873337 | Laufer et al. | Mar 1975 | A |
| 4051228 | Knorre et al. | Sep 1977 | A |
| 4107274 | Knorre et al. | Aug 1978 | A |
| 5307127 | Kobayashi et al. | Apr 1994 | A |
| 5445886 | Ogawa | Aug 1995 | A |
| 5547796 | Kohtaki et al. | Aug 1996 | A |
| 5989304 | Ozaki et al. | Nov 1999 | A |
| 6207101 | Beall et al. | Mar 2001 | B1 |
| 6224980 | Komai et al. | May 2001 | B1 |
| 6503445 | Ozaki et al. | Jan 2003 | B2 |
| 6558488 | Wanninger et al. | May 2003 | B2 |
| 7419527 | Gwozdz et al. | Sep 2008 | B2 |
| 20020128336 | Kolb et al. | Sep 2002 | A1 |
| 20030077448 | Ueta et al. | Apr 2003 | A1 |
| 20040009340 | Zhu et al. | Jan 2004 | A1 |
| 20040221682 | Gwozdz et al. | Nov 2004 | A1 |
| 20060239851 | Nelles et al. | Oct 2006 | A1 |
| 20080152909 | Kyotani et al. | Jun 2008 | A1 |
