Faster formulations

The chemicals industry today is faced with increasing pressure over environmental regulations, health and safety standards, increased customer habit analysis, and the requirement to maintain competitiveness in a global market. All of these factors put pressure on the formulations department to perform to higher levels of efficiency and effectiveness.

In the age of the global chemical company, international communication and co-ordination are issues as never before. As the world has become more unified, the benefits of having separate national organisations have diminished. Chemicals companies have expanded into global markets and are leveraging their global presence to take advantage of economies of scale and different local strengths.

Formulations are also becoming more standardised around the world in order to achieve economies of scale. However, they can never be completely globalised for two main reasons.

Firstly, habits, and perceptions are very different around the world. For example, a sweeter, milkier formulation of chocolate is formulated for the British market than for the rest of Europe to satisfy the nation’s taste. Physical environments also vary. For example, Brazil’s water supply is one of the softest in the world, so there is no need to include water-softening ingredients in Brazilian washing powder. In other markets such as the UK, about 50 per cent of the chemicals in the washing powder are used to soften the water in order to obtain a better wash. As a result, Brazilian powder would not work in the UK, and the powder marketed in the UK uses ingredients superfluous in Brazil.

Secondly, regulatory conditions vary from country to country. For example, Thai law states that in order to call a product washing powder, it must contain a certain percentage of surfactants. Even if a better formulation was produced using fewer surfactants, the product could not be sold as a washing powder in Thailand. The regulatory restrictions are arguably most applicable to the pharmaceuticals industry, which must ensure that their ingredients meet the standards of individual countries.

These examples serve to illustrate the fact that different formulations still need to be produced for different markets. In order to work efficiently, formulators must be able to access data about the existing formulations, regulations, and market environments around the world. The goal of the formulator in a global chemicals business is to make the formulations as similar as possible. The production process and the formulations portfolio must be rationalised to achieve economies of scale, then adapted and fine-tuned for different markets.

There are very few formal courses available in formulation chemistry, and most new formulation chemists are trained in-house. A lot of the experimental work in developing new products through formulations involves a trial and error method. However, this is backed up with an in-depth understanding of how the chemical properties and the processing of the mixtures relate to the final characteristics of the new product, such as the shelf life, biodegradability, taste, or cleaning power.

Naturally, this new product development must take place in a context of consumer requirements, environmental regulations, and ingredient availability. ‘Best practice formulations’, the goal of every formulations department, means fulfilling the requirements for the new product through the most efficient use of available resources in the shortest possible design cycle, resulting in a product with maximum robustness.

Scientists whose task it is to produce new chemical formulations rely on a large wealth of experience in formulation design. Typically they know the recipes (formulae) and ingredients (suppliers’ chemicals) inside out, and have been working in the same department for years. Training a new formulations chemist requires a large investment, and the loss of an old hand through retirement or career progression is a severe blow to the department. With an ageing and increasingly mobile workforce, management watches skills and knowledge leave the company, with the replacement prospects at best being expensive; at worst non-existent.

Workforce attrition is only part of the problem. Even with the same expert workforce, many formulations departments are not working as efficiently as they might. With laboratory results stored in legacy systems, it takes expert company knowledge and experience to know that the results and methods are there, let alone to locate them quickly. This situation is compounded in global R&D set-ups where scientists in laboratories in different countries are working on similar problems. Time differences and language barriers mean that in many cases scientists find that it is easier to repeat an experiment than to find previous results. Mergers and acquisitions, common in the chemicals industry, further cloud the issue, adding different data storage systems to the equation.

Knowledge management is offering real solutions to these problems.

Storing information in a corporate formulations-focused data repository means that scientists can access results and methods from anywhere in the world. They can use their time and resources on new research, rather than repeating old work. Using computer technology to manage an organisation’s knowledge also helps to plan for and minimise the erosion of skills, ensuring that projects run smoothly and efficiently.

An example of a situation where a formulations knowledge management system could have had a huge impact is the recent currency crisis in SE Asia. Due to the devaluation of currency, those markets could not afford to import external ingredients for their washing powders. Detergents companies had to respond to this situation as quickly as possible by producing new products that would satisfy customers in the new environment.

Formulators working under these time pressures needed immediate access to information. They needed to know which ingredients could be easily substituted, and what the minimum quantities were for essential ingredients. The critical success factor was their ability to harness existing corporate knowledge in order to get a new product to market as quickly as possible.

Knowledge management and intelligent tools for the design of experiments help chemists to design formulation experiments taking full advantage of the existing corporate knowledge base. They also ensure that the corporation has a record of what happened when, and, most importantly, why, so that this knowledge is recorded and made available to other scientists.

Knowledge management offers a great improvement to a chemical specialism that relies so heavily on the knowledge of individuals. It also ensures that previous experimental knowledge can be reused the next time a new formulation is required. Scientists working on similar projects at a later date or in different locations will have access to this knowledge, which will save time and money in repeated effort. Storing the failed formulations as well as the successful ones ensures that only one person will go down each dead end.

Molecular Simulations Inc. (MSI) has created a new knowledge management and formulations design system that enables scientists to develop and manage the information and knowledge inherent in the formulation process. MSI’s Formulation-Assisting Software Toolkit (FAST) is an Oracle-based system for formulation design.

Figure 1 captures the overall view of the Software Toolkit, as it embraces the product development process. When a formulator is given a goal of creating an improved product formulation, the first step is to avoid ‘re-inventing the wheel’ or going down a known dead-end. Providing easy access to existing product data immediately saves time and lets the scientist take advantage of existing corporate knowledge.

The FAST system provides access to existing corporate databases, along with data mining tools to assist in the search. If, after consulting the database, it becomes apparent that the formulator will need to carry out new experiments in order to create the new formulation, FAST provides an Experimental Design Tool that helps the formulation chemist design the experiments in the most efficient manner.

The data from the experiments could be added as-is to the FAST database. However, a far more effective way of capturing and distributing this information is to ‘build’ a correlative model from it, and store the model in the database for subsequent re-use. A correlative model is a statistically derived equation which, given inputs describing formulation ingredients and usage conditions, predicts an experimentally-measured property.

Combining models with data management provides an ideal technology for exporting business knowledge throughout an organisation. FAST provides a general framework for letting formulators access models to make tentative electronic formulations (‘eFormulations’). Formulators can try ‘what-if’ virtual experiments at a fraction of the cost and time of laboratory ones, and so can rapidly narrow of the range of potential products.

Once models for all relevant performance properties (including cost) have been developed, then the process of practical formulation design can move on to the ‘real world’ of trading properties off against each other in the context of rapidly changing market conditions, in order to find the best overall product. The FAST optimiser tool allows for automatic multi-objective optimisation, where goals for each different property may be independently set (e.g. maximise, minimise, ignore). The optimiser then automatically varies the ingredient levels to simultaneously achieve as far as possible all specified goals. The user can choose between ‘weighted optimisation’, appropriate in a case where all goals can be simultaneously satisfied, and ‘pareto optimisation’, where a decrease in performance of one property will be accepted in exchange for an increase in another property.

Finally, the formulator visualises the results of the data mining or virtual experiment, in the multi-dimensional formulation+property space. This is done with the Spotfire tool, a third party product provided by Spotfire, Inc. In some cases, customers want to integrate the lists of possible ingredients within FAST with toxicological and production databases. One such link is to the toxicity information provided by Ariel Research Corporations tools. These provide information on a geographical basis (state, country) about the use of ingredients. This data is extremely important to the formulators as they work to develop global brands or master formulations.

FAST is currently in use in formulation laboratories around the world. It is available exclusively to members of the MSI’s Formulations Consortium.

Due to the high level of customisation required for any knowledge management system, MSI has chosen a consortium model of product development.

The consortium model has proven to be an effective mechanism for developing new computational technology and transferring it to industry. MSI has pioneered the consortium model, and it is also used by other companies in the same field, such as Oxford Materials.

Consortium development is a cost-effective approach to software development. Companies join a consortium with other companies in a similar business with similar problems and needs. Jointly, they agree on the best course of software development at a pre-competitive stage. The resultant software is tailored to the needs of the consortium members, but has been achieved at a fraction of the cost of contract or in-house development. Most of the key software developments remain exclusive to consortium members for an agreed period of time. The support and maintenance overhead is shared by the consortium members and is the responsibility of the software company that serves the consortium.

Consortium development of scientific software is the ideal for many pharmaceutical and chemical companies. The client is sure that they get the software that they want, individually tailored to their needs. The software company gains valuable insight into the workings of their clients’ industry, and after an agreed period of time they are able to sell similar solutions with the support of a sound relationship with a leading company. The lock-in mechanism of a consortium, whereby the resultant product is only available to the other members of the consortium ensures that the company retains a competitive advantage by using the software.

MSI consortia are groups of industrial researchers, MSI scientists, and consultants offering particular expertise in relevant technologies. These groups focus on developing, validating, and applying computation to a specific research area. Membership provides software, formal input to MSI product plans, regular meetings at which experiences and ideas are shared, and dedicated application support. It can also offer early access to technology.

Since its inception in MSI, over ten years ago, the consortium model has been highly successful - particularly in establishing simulation in new application areas. Consortia provide nuclei from which wider use grows. Members enjoy the competitive advantages of early adapters - a technology lead, a greater knowledge base, and a better long-term return on their investment. MSI currently runs seven consortia, each of which focuses on a different technology area. MSI’s most recent materials science consortium is the Formulations Consortium, which provides knowledge management and decision support software for the pharmaceutical and chemicals industries. Current members include Unilever Research PLC, Henkel KGaA, and Clorox Inc. On joining the Formulations Consortium, Dr. Wolfgang Gawrisch, Henkel’s Corporate Vice President, Research/Technology commented: "The Formulations Consortium will allow us to provide product developers in all business sectors of the Henkel Group with the latest available tools, significantly accelerating our development process."

Knowledge management for chemical formulation is coming of age, and manufacturers are realising that there is real power to be had through applying computational techniques to formulation design. Companies such as MSI are making huge headway in providing comprehensive solutions to the challenge of knowledge management for formulation design.

Katriona Knapman is scientific communications manager at Molecular Simulations Inc. She can be contacted at: kknapman@msi-eu.com