whole surface covered, further increases in density occur by growth above and
below the surface.
The biomass in SSF is distributed spatially in one of two forms. Unicellular
organisms such as bacteria and yeasts grow as a moist film on the surface of the
substrate particle. This film will have a constant biomass density, and the intercellular spaces will be occupied by moisture. Mycelial organisms such as fungi
and streptomycetes may extend above and below the substrate surface. Steric
limitations associated with branching frequency and branch angles prevent
high packing densities. Maximum packing densities can be 15–34% of available volume [91, 108]. In unmixed fungal SSF processes, within the substrate
particle and within thin moisture films at the substrate particle surface, the
spaces between hyphae will be occupied by an aqueous phase. However, the
hyphae extending above the surface will typically be in direct contact with air
[109]. In processes with mixing, the mixing action tends to squash the fungal
mycelium to make a moist film at the surface, which will behave similarly to a
biofilm of unicellular microorganisms [110].
Penetration into substrates is an important phenomenon which has been observed experimentally but has not yet received modeling attention. Ito et al.
[111] showed that the concentration of penetrative hyphae decreased exponentially with depth in rice koji. Depths of penetration by hyphae of Rhizopus
oligosporus into soybeans during the production of tempe vary from 0.4 mm to
2 mm, with the majority of penetrative hyphae extending to depths of
0.4–0.7 mm [112, 113]. The average frequency of penetration into the soybean
was one penetration per 785 mm
2 of surface area. The majority of penetrations
occurred in the intercellular spaces during growth on soybean cotyledons.
Penetration into substrates where the cellular structure had been disrupted was
easier. In soybean flour, R. oligosporus penetrated to depths of 0.5–1 mm, with
a maximum depth of penetration of 5–7 mm [113].
Penetrative hyphae can potentially play an important role in making the substrate accessible to enzymes [113]. On the basis of expected molecular diffusion
coefficients of proteins in solutions of 10 –11 m 2 s –1 , enzymes would be expected
to diffuse to depths of about 1 mm after 40 h, or even less if lower effective
diffusion coefficients of around 10 –12 m 2 s –1 occurred, such as might be expected within solid substrates [113]. Therefore enzyme diffusion occurs at a
rate similar to hyphal penetration rates. As a result, the contribution of hyphal
movement and release of enzyme from the hyphal tip would be significant in
allowing the enzyme to reach greater depths than if the enzyme was simply
released from the surface. Note that in modeling work to date enzyme release
has been assumed to occur only at the interface between the substrate and biomass phases [84, 85, 114].
Early modeling studies of growth ignored the microbial growth form, simply
expressing it as a concentration per area of substrate surface [83, 84]. Effectively
these models assume that glucose crossing the surface is converted instantaneously into biomass. There is no diffusion of glucose in the biomass
layer. Most of the more recent models have modeled growth in the form of
biofilms, assuming a phase of constant density which increases in height during
growth [85, 114, 115], and in which the value for oxygen diffusivity is that for
90
D.A. Mitchell et al.
below the surface.
The biomass in SSF is distributed spatially in one of two forms. Unicellular
organisms such as bacteria and yeasts grow as a moist film on the surface of the
substrate particle. This film will have a constant biomass density, and the intercellular spaces will be occupied by moisture. Mycelial organisms such as fungi
and streptomycetes may extend above and below the substrate surface. Steric
limitations associated with branching frequency and branch angles prevent
high packing densities. Maximum packing densities can be 15–34% of available volume [91, 108]. In unmixed fungal SSF processes, within the substrate
particle and within thin moisture films at the substrate particle surface, the
spaces between hyphae will be occupied by an aqueous phase. However, the
hyphae extending above the surface will typically be in direct contact with air
[109]. In processes with mixing, the mixing action tends to squash the fungal
mycelium to make a moist film at the surface, which will behave similarly to a
biofilm of unicellular microorganisms [110].
Penetration into substrates is an important phenomenon which has been observed experimentally but has not yet received modeling attention. Ito et al.
[111] showed that the concentration of penetrative hyphae decreased exponentially with depth in rice koji. Depths of penetration by hyphae of Rhizopus
oligosporus into soybeans during the production of tempe vary from 0.4 mm to
2 mm, with the majority of penetrative hyphae extending to depths of
0.4–0.7 mm [112, 113]. The average frequency of penetration into the soybean
was one penetration per 785 mm
2 of surface area. The majority of penetrations
occurred in the intercellular spaces during growth on soybean cotyledons.
Penetration into substrates where the cellular structure had been disrupted was
easier. In soybean flour, R. oligosporus penetrated to depths of 0.5–1 mm, with
a maximum depth of penetration of 5–7 mm [113].
Penetrative hyphae can potentially play an important role in making the substrate accessible to enzymes [113]. On the basis of expected molecular diffusion
coefficients of proteins in solutions of 10 –11 m 2 s –1 , enzymes would be expected
to diffuse to depths of about 1 mm after 40 h, or even less if lower effective
diffusion coefficients of around 10 –12 m 2 s –1 occurred, such as might be expected within solid substrates [113]. Therefore enzyme diffusion occurs at a
rate similar to hyphal penetration rates. As a result, the contribution of hyphal
movement and release of enzyme from the hyphal tip would be significant in
allowing the enzyme to reach greater depths than if the enzyme was simply
released from the surface. Note that in modeling work to date enzyme release
has been assumed to occur only at the interface between the substrate and biomass phases [84, 85, 114].
Early modeling studies of growth ignored the microbial growth form, simply
expressing it as a concentration per area of substrate surface [83, 84]. Effectively
these models assume that glucose crossing the surface is converted instantaneously into biomass. There is no diffusion of glucose in the biomass
layer. Most of the more recent models have modeled growth in the form of
biofilms, assuming a phase of constant density which increases in height during
growth [85, 114, 115], and in which the value for oxygen diffusivity is that for
90
D.A. Mitchell et al.