There are two general hypotheses to explain this result.One is that the amorphous state of P(3HB) in vivo can becaused by a simple physical-kinetic limitation [153,154]within granules based on their small size (stabilized by anamphiphilic shell). This assumes that the rate of P(3HB)crystallization in vivo is governed by the rate of sponta-neous nucleation events, which depends upon the granulevolume and therefore on the third power of the granuleradius. The upper limit for the rate constant of spontaneouscrystal nucleation in isolated P(3HB) at 30◦C (the tempera-ture of bacterial growth) is 2.5 events mm−3s−1[155,156].The corresponding rate of nucleation for a typical storagegranule of diameter 0.25 m is 2.0 × 10−11events s−1. Pro-vided the granules do not coalesce and are not exposed to foreign nucleating agents, native granules should havea crystallization half-life of at least 3.4 × 1010s, or >l000years [112]. Even at a maximum nucleation rate of less than100 events mm−3s−1, the average lifetime for a granulebefore crystallization should still exceed several months.The in vivo state is therefore proposed to be underkinetic, not thermodynamic control [112,154]. Horowitzand Sanders simulated the formation of granules usingsynthetic surfactant to coat the surface of submicron-sizeartificial granules, which were shown to be amorphous andstable in suspension, and which had the same size, den-sity, morphology and molecular mobility as native granules[112,157].