Metals and materials play a pivotal role in society as their properties impart unique functionality to engineered structures and consumer products. Metals are theoretically infinitely recyclable; however, the functionality and design of consumer product complicate recycling due to their ever more complex structures producing unliberated low grade and complex recyclates. Metallurgical smelting ingenuity, good technology and intelligent use of thermodynamics and transfer processes gets metallurgists and recyclers a far way down the path of creating high recycling rates from a large range of primary concentrates and recyclates. However, the 2nd Law of Thermodynamics teaches us the practical limits of recycling in terms of entropy creation, which is determined by the complexity of the recyclates and hence to the economics of processing/ technology and metal/energy recovery. The usual simple accounting type tools do not rise to the challenge. Therefore, a key issue for the creation of ''sustainable systems'' and hence the minimization of waste (or in other words achieve high recycling rates) is the creation of optimal industrial ecological systems with optimally linked Best Available Techniques (BAT). This must maximize the recovery of materials from ores and recyclates within the boundaries of consumer behaviour, product design/functionality, thermodynamics, legislation, technology and economics. Examples will show how recyclate quality/grade predicted by recycling models affects entropy creation, while also reviewing various published methodologies. This paper shows that simulation models are a prerequisite to designing ''sustainable'' systems as these can predict recyclate grade/ quality/losses/toxicity of streams, the link to entropy and economics and the realization of company ideals and mission statements in this regard. In other words, to dematerialize society requires detail input by engineers, their predictive tools and economic based design approaches to engineer a sustainable future.