Formulation is the process used to formulate a new product or material. The result is a recipe specifying the additives along with special mixing directions. There are two extremes of formulation. The least complex formulation usually involves modification of an existing recipe to meet changed customer requirements. This is sometimes accomplished by changing one ingredient (increasing, decreasing, or replacing). The most complex formulation problems often arise when no existing material meets all the customer needs. Cost is always a factor often followed by the necessity to function in an extreme environment and maintain a good appearance or something similar. Coleman, Ernie (2/10/2010) More Starting Making the Recipe Testing Plackett-Berman Design of Experiments Recommendations Background Literature Review Research Strategy Materials Properties Testing Processing  

More Sometimes the approach is to consider processing and design factors as well as just mixing additives. Examples of these include crosslinking, multi phase or multi layer designs, foams, grafting, and other combinations of these techniques to some of the materials or additives used in the final formulation. Many times grafted additives (like some impact modifiers) are far more effective than simple additives.

Starting Here are some ideas when you start from a wide open situation. Think about one polymer because it is inexpensive, a second because it has a key property, and a third to add more properties or to compatibilize the first two polymers. Just be sure at this stage to use materials at a level of about 15% or more based on final composition. The reason for the 15% suggestion is that testing errors of 10% are expected at this stage and 10% material would not make any difference if its contributions were linear. Even a 10% cost reduction would be relatively small, so use 15% at least of all major materials.

Making the Recipe After these key decisions have been made, go through the list of additives that are necessary to fabricate and additives to enhance properties. Select what is needed and pick an amount that is logical.

Testing Select three to five different formulations and make them. Test two key properties on each and select the best two and poorest. Add the ingredients in the best two and subtract the ingredients of the poorest to get the next formulation to try. This may not be exact because all samples may not contain the same ingredients, but it will suggest one (or more) samples to make up for the next trial. At this stage, you can even use the properties and ingredients of a preexisting material and force it into the system. This technique of rewarding the best two and punishing the poorest will speed you on your way towards formulating a new material that is a winner. Of course there are many other ways to do formulation. If you have other design of experiment ideas, use them.

Plackett-Design of Experiments Another technique is to use the Plackett-Berman design of experiments. Experiment -------A--B--C--D--E--F--G No. 1--x--o--x--x--o--x--o No. 2--o--x--x--o--x--o--x No. 3--x--x--o--x--o--x--o No. 4--x--o--x--o--x--o--x No. 5--o--x--o--x--o--x--x No. 6--x--o--x--o--x--x--o No. 7--o--x--o--x--x--o--x No. 8--o--o--o--o--o--o--o This is a typical P-B design for 8 experiments to test the effects of 7 variables at two different levels. In this design, x is high and o is low. Experiment 8 has the minimum amounts of all 7 variables. The value of this experimental design is that seven variables (A-G) can be tested in a single experiment. It is important to run every formulation in the design in the exact way specified in the design and to run the individual experiments in a random fashion. There are other experimental designs that can be used to accelerate new product development as well as upgrading of old formulations. The advantage of these tools is efficiency of research. They have the common theme of make many changes in a systematic fashion and then being able to factor out exactly which changes made the greatest contributions to the required improvements or upgrades. The P-B design allows processing variables like high and low temperatures, two different screw designs, etc. to be incorporated along with no and high loadings of impact modifier or to explore two different types of flame retardant packages (where one is ‘low’ and the other is ‘high’). This may be stretching this type of design to the limit, but it can be useful in finding new solutions to old problems and better solutions to new commercial needs.

Recommendations Background In order to avoid problems later, it is essential to know what to invent and the processes that will be used to make the product. At the same time it is helpful to know how others have solved the same or similar problems. The only way to get this information is to do a literature search along with a current market awareness study.

Literature Review A good literature search will include: · patents · technical journals · trade journals · chemical abstracts · engineering abstracts, etc. · salesmen There are several excellent computerized information retrieval services available today. However, many of these services do not go back in time before 1967 or 1950 at best. It must be remembered that basic science and technology in plastics was developed, described, and patented from the 1920’s to the 1960’s. These data must be retrieved by the old fashioned hand search methods in some cases. Even in the area of recycling and compounding with recycled feed stock, the old literature should be consulted. Recycling is not new It has been practiced since the first plastics were made. Modern recycling is based upon past work, so this information should be consulted. The same is true of nano technology, even though it was not calld by that name. You must be sure that what you do is patentable or new; otherwise expensive laboratory time is wasted. What do you do if you end up with a process that infringes an existing patent? An information scientist can be of great help in speeding the process of searching and making sure that key information is found so that your work is of high value and not a repeat of earlier findings. Every useful reference that is found is worth approximately two years of lab work and at least one year of lapsed time. A missed reference by similar reasoning might set back your research by a similar amount of time and effort. Literature searches are retrospective. Customers live in the present and future. The only way to relate to a current or future need is to ask the customer and/or end user, and to keep asking them questions throughout the program. This can be done by your own sales force or by outside market consultants. In some cases both are used. In some cases, you must see the customer yourself and establish a working relationship which takes time. Properly done, it will be of immeasurable value.

Research Strategy After the literature search is well underway and before lab work begins, the research strategy must be decided. If the technology is unique, all types of patent coverage (product, process, and application)can be used to protect the invention. The key strategy will be keeping the technology confidential until the patent is filed. Along with this is the need to cover as many products, processes and applications as possible consistent with secrecy and early filing date. Seldom will this be the case. Usually you will be working in areas that are not new but many important inventions are still possible. There is no need for super secrecy but time is a key factor in order to bring the product to the market while the specifications are still open and early processing trials are still underway. In these cases, the R&D is more open and customer interaction should be encouraged. Customers will tell what is needed. When the needs have been met, patent action can be considered. The key properties in all cases are cost effectiveness, price and value-in-use. In other words; cost, cost, and cost. All R&D work must be based on meeting cost goals. This is a necessary concession to the commercial world, and it is the only way to be sure that technically successful R&D work will reach the marketplace. All R&D workers must know the commercial value of their new products and constantly make simple cost estimates about their work to be sure that there is room for profit. This is probably the most often overlooked point in product development work. To sum up to this point: 1-know the cost 2-know the customers 3-know the competition 4-know the technology 5-know the market

Materials Properties In order to meet demanding property and processing requirements at a competitive cost, begin with the material properties. Of course if there is an overriding reason to consider processing first, do it, but do not get locked into only one way to process. This is unavoidable with small shops that have only one compounding line, but even small companies are free to rent time at outside facilities to evaluate alternative compounding machines. Begin your work by writing down all physical properties that must be met in the composite material and in the end product ( if known ). [Orientation and other process related factors must be kept in mind throughout the work.] Check to be sure that no existing material already meets or comes within 20% of all mechanical/physical properties or within the same range of thermal properties. If such a material exists, find out why it is not suitable for use. Document this point and make sure this information is widely distributed within your company or department. You do not want to reinvent something that has already been done. Since most mixes, blends, and composites of two polymer components (plus fillers or reinforcing fibers sometimes) are probably already offered commercially, the solution to your problem will almost always require three or four major ingredients. You will be fortunate if only three major (over 15% of the final product by weight )ingredients work. In recognition of these basic facts, start with blends of four or sometimes five major ingredients, and never use less than 15% of any single ingredient in the initial formulation. The reason for this is to insure that the contribution of each ingredient will be detectable even with relatively large testing errors. Use statistically designed programs to optimize formulas. Clear or transparent materials are a special case. There are only two ways to mix materials and keep them transparent: match refractive indexes, or keep particle size of ingredients much smaller than the wave length of the light (generally ¼ of the wavelength) whenever particles are added as a second phase. Testing Many end uses require specific minimum mechanical and thermal properties. Chemical and electrical properties may enter into some applications. Clearly, it is not possible to run more than two or three screening tests on each sample that you make and still keep time and costs within reasonable limits. Many assumptions are necessary to keep cost of development down, to get to the goal in reasonable time, and to allow other workers in the company some testing time for their projects. Prepare samples by a process that is as close to the intended end use as possible. Extrude sheet and stamp or cut out test specimens if the product is for an extrusion application. Better yet, test the specific extruded shape if this is at all possible. Mechanical tests on compression molded specimens do not predict performance for many extrusion or injection molded applications. This is not the place to cut corners or to try to save costs. Fabricate test specimens by the same method as your customer uses. Compression moldings frequently give different property values than injection molded or extruded samples. ISO, ASTM and other standard tests usually require enough specimens for each test in order to reduce test errors to a few per cent. By accepting large test errors in screening ( say 20-50% ), it is possible to test only one or two specimens. Here are the steps so far: 1. Formulate with a three to five major ingredients plus necessary minor additives. 2. Fabricate test samples by the same general method as your customer uses. 3. Test only one or two specimens (but hold others in reserve) during the scouting stage. Set your new product goals high (20% high in mechanical or 10 degrees above the temperature specification, for example). This applies only to the tests that are run during the screening part of the project. If the goal is 10 Pa try for 12 and don’t settle for anything less (at this point). If impact specification is 1.0, you set 1.2 as the only goal you will accept. Think big and set high goals. You will succeed more often than you ever expected. Even your failures will be winners.

Processing Processing can be reactive or simple mixing. In reactive processing, the ingredients are reacted chemically with one or more reactive coupling agents or tie layers. This reactive processing is growing in importance but the majority of commercial compounding is still in the passive or non-reactive category. Care must be taken to avoid breaking fibers but at the same time all ingredients must be dispersed in order to be cost effective. The use of multiple feed hoppers or more than one mixing or compounding device overcomes these common problems of passive compounding. In order to be sure that the product(s) of extrusion are within the expected range of composition, some simple corroborative tests like density; test bar mass, or per cent ash should be run on all lots. A file of this type of information will be useful in analyzing competitive products in the future.If your laboratory has access to all types of process equipment, you will be overcome with options. There are several ways to narrow down the choices. 1-use the least expensive equipment 2-use the most intense mixing equipment 3-use process selection as one of the variables in the experimental design to select materials and Do not worry about getting all the answers on the first round. 4-ask someone you trust and use that process. 5-pick two or three quite different processes and use them at random with occasional cross checks. Above all, do not try to optimize the process or prematurely rule out a processes until the first rounds of material testing are completed. When making process decisions, make large changes until the product is within 10% or a few degrees of the goal properties. 6-After the first round After the data from the first round of experiments are available, it is time to see what is known about cost and performance. You are now far more knowledgeable. Reread references, especially those that seemed unusual, difficult to believe, or unlikely to succeed. You may see them in a different light at this point. Recheck some basics. Did the melt temperature exceed the melting point or minimum processing temperature of the major material with the highest processing temperature or melting point? Did you match viscosities to make mixing of different materials more efficient? Should an ingredient be dropped because it is not cost effective? This is a difficult decision to make, but it is best to do it now before excessive effort is expended on a favorite material that is not doing anything beneficial. Now is the time to bring in minor ingredients for specific reasons. Antioxidants or processing aids are required to stabilize resins for short times at high temperatures during melt processing. Needs for this type of additive should be obvious at this point Coupling agents, dispersing agents, mold release, internal lubricants (viscosity reducing additives), etc. can be brought into the formulations at this point, if they have not been added sooner. Remember, minor ingredients that are known to be essential for function—stabilizers for PVC, flame retardants for FR materials, pigments for colored materials—must be included in the first round. What we are doing at this point is taking note of product deficiencies and testing specific trace or minor additives to provide corrections or improvements. 7-Other options With a good experimental design, you will be close to your product goals in just days from the start of a new program. This assumes full time work on the project and priority in testing. This also assumes reasonable availability of processing machinery and access to the patent department to keep them informed of the developing technology. Find a way to keep marketing and sales informed of progress and involved with the importance of your program. Remember, when talking to technical people to use technical language, and when talking to business people and management you must use the language of money and profit. Personal contact, telephone calls, and the written word must all be used in order for communications to be effective. What do you do now to move the program forward and keep the momentum? There are two options: continue in the laboratory phase as before; or push to get your best candidates into customer trials with you there to observe. Current business needs must be balanced against patent and long term protection of company technology. In well over 90% of all situations it is better to go for current business if the candidate material comes close to customer needs and processes easily. If there is a hint of processing problems at this point, do not go to the customer.Low properties can be tolerated better than a plastic that can not be fabricated, in my opinion. At this time it is necessary to reevaluate the end use requirements and change your goals to better reflect the true current end use needs of the customer. Remember that the plastic automobile spring was originally specified in carbon fibers, but it was a successful commercial application with the less expensive and lower property level of glass fibers. This was a mid-point change. Know your customers better and work closely with them at all stages of product development. This is especially important when major specification changes are being discussed, and you can only know about this if you are in close contact with your customer. The above represents some techniques that have been used in solving compounding problems at different companies and with several different materials. Ernie Coleman 8/24/2009

Recent US Patents

Recent Journal Articles

Review Articles

Editor's Notes

Comments