Tuesday, March 13, 2012

Masterminding Mixing Technology

Mixing isn't exactly the most titillating topic, but it is an essential operation in the chemical process industries. Not only that, but it also has a substantial impact on a manufacturer's bottom line. In 1993, a major U.S.-based chemical company estimated that the value of mixing to the firm was in excess of $25 million annually. A recently published handbook on industrial mixing ( see p. 47 for a review of the book) estimates the cost of poor mixing to be as high as 0 $0 100 million/yr.

Mixing equipment has matured over the years and is now in an evolutionary state. New developments in design and performance are being driven by the characteristics of the product being manufactured. Today, formulations are becoming increasingly complex and viscous, and where a product's fluid characteristics are essential to its functionality (e.g., sunscreen, liquid detergents and conditioners), the rheology can be highly intricate. "In these cases, it is often essential to apply much higher shear rates than those achievable in agitated vessels, which drives the industries towards rotor-stator and similar devices," says Michael Butcher, marketing director of BHR Group's Fluid Engineering Centre (Bedfordshire, U.K.; www.bhrgroup.com).

The distinguishing feature of a rotor-stator (R-S) mixer is a high-speed rotor in close proximity to a stator (Figure 1). Typical rotor tip speeds range from 10-50 m/s. They are also called high-shear devices because the local shear rate they can achieve in a vessel (20,000-100,000 1/s) is much greater than that which is possible by a mechanical agitator. Charles Ross & Son Co. (Hauppauge, NY; www.mixers.com) offers an ultra-high-shear inline R-S device called the MegaShear that can do everything from dispersion to disintegration of difficult solids, such as polymers and elastomers, converting them into submicron-sized particles in a single pass. As fluid enters the center of the stator, pumping vanes on the rotor, which spin at 55 m/s, accelerate the product through grooves in the respective parts, but in opposite directions, "the result being an opposed flow collision that imparts tremendous shear forces upon the product," says Doug Cohen, vice president of technical services at Ross.

A secondary trend is the goal of using a high-shear mixer to disperse a dry powder directly into the flow of a liquid, replacing older design concepts that relied on an additional pump to transport solids from the solids eductor to the mixer. "This is a critical issue because many powders are extremely hard to disperse efficiently, and conventional technology was prone to persistent clogging and slow induction rates, which always drove up cost while they drove down throughput," says Scott Anderson, IKAWorks' (Wilmington, NC; www.ikausa.com) technical services manager. The firm's multipurpose homogenizer and disperser, the MHD 2000 utilizes an auger and paddle in place of a venturi device to feed the solids into the mixing chamber at rates of up to 700 lb/min, while wetting out resins, polymers and other materials that have viscosities of up to several hundred thousand cP - a feature that is not feasible with vacuum-type systems. It also significantly reduces aeration of the product.

"Rotor-stator mixer technology has been refined as opposed to revolutionized over the last 10 years," says Arthur Etchells, III, a mixing consultant with DuPont Corp. (Wilmington, DE; www.dupont.com). "What has changed drastically is our understanding of its design and operation, mostly through a trial-and-error approach to process development and scaleup, because there is no fundamental basis for this mixer's performance," he continues. Bridging the gap are two consortia, BHR's Fluid Mixing Processes Research Consortium and the highshear mixing research program spearheaded by Richard Calabrese at the Univ. of Maryland (College Park, MD; www.umd.edu). Both have engaged in research and development to elucidate the complex hydrodynamic environment in R-S devices.

Gaining a deeper understanding

Computational fluid dynamics (CFD) has become an essential tool, along with conventional and new experimental techniques, in the conceptualization, scaleup and understanding of mixer performance. 'However, time-dependent mixing flows, coupled with complex geometry, bring uncertainty to the CFD predictions, especially for turbulent flows," says Calabrese. "Models that couple the local reaction and mixing processes allow the simulation of the spatial variations of concentrations due to mixing and diffusion and thus, the rates of chemical reaction."

CFD software companies have put more options into their packages to help users build a more realistic model of the flow field in the vessel. This includes the use of using velocity input from the outlet of the impeller obtained by techniques such as laser Doppler velocimetry (LDV). Particle image velocimetry (PIV) has evolved to be a powerful technique for 2D and 3D whole-field velocity measurements and is especially useful for examining instantaneous spatial shear rates, but it is not as accurate as LDV for time-averaged measurements.

CFD is touted as a helpful tool in vendors' efforts, as well. Once an agitator is designed at Chemineer, Inc. (Dayton, OH; www.chemineer.com), the file may be sent to a preprocessor that sets up all the required information for the CFD software, including the grid and boundary conditions. The output is standardized and sent to the agitator designer. "It has proven most beneficial when designing agitators for non-Newtonian fluids," says julian Fasano, director of engineering and development at Chemineer. David Dickey, senior consultant at MixTech, Inc. (Dayton, OH; www.mixtech. com), isn't as quick to praise the benefits of CFD. "With regard to high-shear applications, CFD has only been shown as a research tool for tracking flows in R-S mixers. Regardless of the sales pitches, it does not work well in multiphase, dispersed, or nonNewtonian fluids."

One challenge with CFD is often the lack of data to validate the results. "In such situations, analysts must count on engineering knowledge and ' experience," adds Victor AtiemoObeng, scientist, engineering and market development, Dow Chemical.

Designing around scaleup issues

Scaleup methods for batch and semibatch mixing systems have been modeled extensively. However characterization of the physical and chemical parameters of multiphase systems with complex reactants and interfacial phenomena is extremely difficult and may limit the usefulness of these correlations, especially for CFD simulation. "One of the most difficult aspects of scaleup in homogeneous and heterogeneous reactions is the prediction and control of byproduct distribution. "These byproducts may be negligible on the lab or pilot scale, but may increase upon scaleup to production. An increase of as little as 0.1-1% in the amount of a particular byproduct may not be acceptable when it cannot be adequately removed by downstream processing," says Etchells. These impurities may affect the physical form, particle size, downstream liquid-liquid separation or foaming tendency of the product.

"The problem is that when scaling up, people do not always recognize the critical factors involved in the basic process - e.g., proper heat and mass transfer, especially if fast reactions are occurring," Etchells continues.

In conventional emulsions processing, critical parameters, such as mixing energy, mixing time, and heating or cooling times, are not easily transferred from the laboratory to production. With Velocys, Inc.'s (Plain City, OH; www.velocys.com) microchannel emulsifier, which creates dispersions for the coatings, food, and cosmetics industries, these parameters remain constant for all scales of processing.

During operation, the discontinuous and continuous phases enter the device and flow in alternating microchannels measuring 250-5,000-�m in dia. (Figure 2). The discontinuous phase flows through a porous substrate aligned between the two flow channels and forms droplets less than 1,000 nm that are clipped by the continuous phase, creating the emulsion, which exits through one outlet port. The system operates at temperatures in the range of 80-200�C, and at low pressures. "An emulsion might be made at 50 psig, compared to 20 ,0 000 psig in a homogenizer," says Laura Suva, manager of business development at Velocys. Pressure drops are 10 -0 10 bar, with services up to 100 ,0 000 L/h.

A similar system was unveiled in 2003 by the Institute f�r Mikrotechnik Mainz GmbH (Germany; www.immmainz.de). Called the StarLaminator, it is a micromixer capable of handling flowrates of up to 300 L/h at a pressure loss of 12 bar. The complete unit, with dimensions of 45 � 25 � 30 mm^sup 3^, consists of a stack of 320-1,600 mixing foils measuring 50 �m thick. The foils have microchannels with three different patterns. With a careful choice of the sequence, both educts are fed alternately in thin layers to a central channel, where the streams merge to become thin fluid sheaths. Mixing occurs via diffusion through the sheaths, then by secondary turbulence, depending on process conditions.

Moving towards continuity

Where process conditions and scale of operation allow, there is a move towards inline continuous mixing and even chemical reactions, which would allow the ability to control the reacting environment much more closely, thus permitting higher mixing and heat removal rates, elevated operating pressures, and more successful scaleup. "It is a form of process intensification that often allows reaction routes not possible within the constraints of stirred tanks, and it also requires less space," says BHR's Butcher.

"In the fine chemical industry, most reaction chemistry is done in batch reactors, where scaleup usually requires multiple units and presents risks of batch inconsistency," says Irv Gruverman, CEO of Microfluidics, Inc. (Newton, MA; www.microfluidicscorp.com). The firm recently commercialized a multiple stream mixer/reactor (MMR), based on its microfluidizer platform, that continuously produces uniform nanoparticles using multiple reactant fluid streams in an ultraturbulent reaction zone. 'The MMR optimizes fast chemical reactions and can be scaled easily to commercial production without losing product quality, since the equipment geometry does not change," Gruverman continues.

Current applications include making nanosuspensions for intravenous delivery of pharmaceutical insoluble actives, high-purity metal oxides for the electronics industry and dings nanoencapsulated in polymers, according to Mimi Panagiotou, Microfluidics' director of R&D. "We built a prototype comprising two microfluidizers operating in parallel that delivers <1 L/min of product and expect to scale this system to 80 -0 10 L/min without difficulty," she says.

Multiple reactant streams are fed in stoichiometric ratios at up to 400 ,0 000 psi through a pencil-eraser-sized channel within the mixing chamber. Flow is directed into nanometer-length inlet channels (0.01-0.05 cm^sup 2^), where fluid velocities of 30 -0 30 m/sec are reached and maintained for tens of milliseconds, creating a small amount of microcrystalline product nuclei to seed the reaction. The streams enter a second set of narrower channels (0.0001-0.001 cm^sup 2^) at 80-300 m/sec and mix for a few milliseconds before being released to a relatively large exit channel.

Depending on optimum conditions for a reaction, the geometry features macro-, meso-, and micromixing regimes (Figure 3). In the case of a very fast reaction, the degree of micromixing is maximized, generating energy dissipation values on the order of 10^sup 8^-10^sup 10^ W/kg, "far greater than that achieved near the impeller in any stirred tank reactor, homogenizer or R-S mixers," says Gruverman. The products are nanoparticles in the 10-100 nm size range, and can usually be held within 0 �0 100 %0 of the target particle size.

Shifting expertise

Many vendors and users acknowledge a downward trend in new products and services, in part due to mergers and acquisitions, such as the recent unification of Bran+Luebbe, Lightnin and Waukesha Cherry-Burrell to create SPX Process Equipment, Inc. (Delavan, WI www.spxprocessequipment.com), and Sulzer Chemtech's (Winterthur, Switzerland; www.sulzerchemtech. com) purchase of the mixer business of Koch-Glitsch. Users have cut their own engineering staffs, and now rely more on vendors or consultants to handle processing issues that in the past would have been handled inhouse. "Ten years ago, we had ten inhouse mixing experts. Today we have two," says I-Hwa Midey ChangMateu, director of coatings process technology at Rohm & Haas Co. (Philadelphia, PA; www.rohmhaas. com). "Much of the capital budget has been allocated to addressing security concerns," she adds.

Furthermore, with the large decline in sales for mixers in the past few years, vendors have reduced, or even eliminated, research in mixer technology. Consequently, says Shaffiq Jaffer, senior engineer and mixing specialist for Procter and Gamble's (P&G;West Chester, OH; www.pg.com) corporate engineering technical laboratories, some users are becoming more of the expert on the mixing equipment than the vendors from whom it was purchased. "A large part of being successful in developing new processes for existing hardware is leveraging external capability - linking up with consortia and academia," he says. 'When it comes to allocating resources, mixing equipment R&D is not a priority. At P&G, it accounts for much less than 1% of the R&D budget (which is 3-5% of the outside sales)."

Against this backdrop, mixing equipment design has gone beyond mechanical and costing considerations, with the primary objective being how best to achieve the key mixing process objectives, says Edward Paul, a mixing expert who recently retired from Merck & Co. (Rahway, NJ; www.merck.com).

Growing markets

In the pharmaceutical industry, impinging jets mixer/crystallizer technology has come of age. Merck holds the original patent for crystallization of pharmaceuticals using an impingingjets mixer, where reactants in a solvent precipitate are mixed with an anti-solvent (or non-solvent) via the collision of opposed jet streams to achieve a particle size of ~5-20 �m. "Intense micromixing and high supersaturation is responsible for the rapid crystallization of monodisperse micron sized particles, which exhibit improved bioavailability and stability," says Merck's Brian Johnson, who is responsible for the scaleup of processes that make pharmaceutical ingredients. Merck, which also patented the production equipment, is currently using the technology for the production of a commercial drug product.

In 2003, Pfizer patented the use of impinging jets for a reactive crystallization where two streams react in a rapid mixer and then crystallize to produce the pharmaceutical. To avoid infringement on Merck's intellectual property in future inventions, Pfizer has applied for a patent on an impinging plate device.

Meanwhile, Bristol Myers Squibb Co. (Princeton, NJ; www.bms.com) recently patented a technology for creating submicron crystals of a drug that involves two impinging liquid jets are positioned within a flask that is sonicated near the jet-impingement point by a 20-kHz probe. The ultrasound is claimed to enhance the mixing process and promote the formation of crystals smaller than that which is possible with the impinging jets alone.

Over the last five years, Robert Prudhomme, a Univ. of Princeton researcher, and Johnson have developed a process they call Flash Nanoprecipitation, where rapid mixing of reactant streams containing colloidal stabilizers is performed in an analytical, confined impinging jets (CIJ) mixer to produce nanoparticles of pharmaceutical agents. The system, which has been qualified, requires a mixing time well below 100 ms for ideal nanoparticle formation. "We have quantitatively shown that the time for mixing should be less than the time for precipitation, for optimum performance - hence, the need for specialty designed high intensity mixers to produce uniform and small particles," says Johnson.

In the plastics industry, manufacturers have sought to improve the properties of commodity plastics by mixing them with another plastic or additive. "A shortcoming of existing mixing technology is that only a limited variety of blend morphologies are producible at low compositions of a property modifier," says Dave Zumbrunnen, professor of mechanical engineering at Clemson Univ. (SC; www.clemson.edu). To overcome this, Zumbrunnen and a team of researchers at Clemson developed a continuous blending device called the SmartB lender, which produces polymer blends with novel properties by folding the two melts together, rather than by distributing one evenly throughout the other.

The device uses a principle of fluid dynamics known as chaotic advection to repeatedly fold a masterbatch or other component into a matrix polymer, forming a variety of controlled and repeatable polymer morphologies, from layers, ribbons and platelets to spongy interpenetrating structures. In a typical run, material is fed by two 0.75-in.-dia. single-screw extruders into a crosshead die from opposite ends and into a distribution head. The matrix material passes through a single central port, while the secondary resin or masterbatch goes through nine small ports arranged in a circle around the central one.

From the distribution head, the material enters a cylindrical blending chamber containing two 22-mm-dia. stirring rods that are programmed for a sequence of speed and/or directional changes -i.e., they take turns spinning three times faster than the other for a specific number of turns. Changing the rod rotation protocol can produce blends of differing morphologies without any type of equipment modification.

Clemson researchers have used the technology "to produce sponge-like blends of immiscible polymers, such as polypropylene (PP) and low-density polyethylene (LDPE) over broader compositional ranges (e.g., 70%:30% by vol.) than are achievable with conventional compounding equipment, yielding morphologies that improve PP's normally poor cold impact strength," he says. In this case, chaotic advection stretches and folds thinner and thinner layers of LDPE in the PP matrix. After repeated layering, the LDPE layers become so thin, they eventually rupture, letting the PP flow through the holes in the LDPE. Holes also form in the PP layers, creating a fibrous spongy structure out of the suffer PP.

Zumbrunnen says that the team is now performing developmental work with plastics for specific clients, extruding these materials into films and sheets for companies to test in their own labs. At least one machine supplier has been selected to produce commercial-scale equipment, and more agreements are expected in the near future.

No comments:

Post a Comment