RESUMO
Prevention of fungal spoilage is a key microbiological issue for the shelf life of fat spreads. Our aim was to assess and model the scope of (natural) antimicrobials for extending shelf life of spreads (water-in-oil emulsions). Production conditions were established to make 60% model fat spreads with reproducible droplet size distributions. The mould vulnerabilities ranged from 1 to 20 weeks. The system allowed feasibility testing of lytic enzymes (Novozym 234) and LMW compounds against Penicillium roqueforti, a key-spoilage mould. The action of Novozym 234, carvacrol, undecanol and dihydrocarveol was benchmarked against sorbate and preservative-free controls under ambient and chilled conditions. Novozym 234 was ineffective to prevent outgrowth of P. roqueforti. Carvacrol, undecanol and dihydrocarveol had limited effects on shelf-life extension compared to sorbate. Fungal growth boundaries of (un-)preserved spreads were modelled. The emulsion droplet size distribution (DSD) was first captured in a mechanistic parameter DSD-I (I = Influence). DSD-I was a move away from the mean droplet diameter D3,3 as sole quantitative droplet-size distribution parameter for mould susceptibility of emulsions. DSD-I is a combination of available water droplets and surface area to initiate and sustain fungal outgrowth. Followup experiments showed that modelling D3,3 and distribution width (e(sigma)) instead of DSD-I gave better results for emulsions with high e(sigma). Empirical predictive models were subsequently developed for the effects of D3,3, e(sigma) and undissociated sorbic acid (HSO) on the shelf life of emulsions.