Keywords:mixed cultures;modeling and control;operating parameters;polyhydroxyalkanoates (PHA);polymer characterization
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Abstract
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Summary: Numerous bacteria have been found to exhibit the capacity for intracellular polyhydroxyalkanoates (PHA) accumulation. Current methods for PHA production at the industrial scale are based on their synthesis from microbial isolates in either their wild form or by recombinant strains. High production costs are associated with these methods; thus, attempts have been made to develop more cost-effective processes. Reducing the cost of the carbon substrates (e.g., through feeding renewable wastes) and increasing the efficiency of production technologies (including both fermentation and downstream extraction and recovery) are two such examples of these attempts. PHA production processes based on mixed microbial cultures are being investigated as a possible technology to decrease production costs, since no sterilization is required and bacteria can adapt quite well to the complex substrates that may be present in waste material. PHA accumulation by mixed cultures has been found under various operational conditions and configurations at both bench-scale and full-scale production. The process known as “feast and famine” or as “aerobic dynamic feeding” seems to have a high potential for PHA production by mixed cultures. Enriched cultures submitted to a transient carbon supply can synthesize PHA at levels comparable to those of pure cultures. Indeed, the intracellular PHA content can reach around 70% of the cell dry weight, suggesting that this process could be competitive with pure culture PHA production when fully developed. Basic and applied research of the PHA production process by mixed cultures has been carried out in the past decade, focusing on areas such as microbial characterization, process configuration, reactor operational strategies, process modeling and control, and polymer characterization. This paper presents a review of the PHA production process with mixed cultures, encompassing the findings reported in the literature as well as our own experimental results in relation to each of these areas.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
1. Introduction
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Plastics occupy a high volume fraction in municipal landfills because of their relatively low density (0.9 g · cm−3). Substitution of synthetic plastics by biodegradable plastics can reduce almost 20% of the total waste by volume and 10% by weight.1
Polyhydroxyalkanoates (PHA) have been recognized as good candidates for biodegradable plastics because of their similar properties to conventional plastics and their complete biodegradability. Furthermore, PHA can be produced from renewable carbon sources, allowing for a sustainable and closed-cycle process for the production and use of such polymers.2 Although the most well-studied PHA is poly(3-hydroxybutyrate) (PHB), over 150 different hydroxyalkanoic acids are known as constituents of these biopolymers at present.3 From these, only the homopolymer of PHB and copolymers of 3-hydroxybutyrate (HB) and 3-hydroxyvalerate (HV), and HB and 3-hydroxyhexanoate (HHx) have been, so far, industrially produced.4
The cost of PHA production is still too high for PHA to become a competitive commodity plastic material. In order to obtain a sustainable industrial process based on PHA, its high manufacturing cost needs to be reduced. The most significant factor in the production costs of PHA is the price of the substrate and the corresponding fermentation strategies.5 The use of renewable carbon sources based on agricultural or industrial wastes, and the development of processes requiring lower investment can contribute to reducing the production costs.
PHA are polyesters that are synthesized and stored within the cell by various microorganisms. More than 300 different microorganisms that synthesize PHA have been isolated. Industrial production processes are based on the use of pure cultures of microorganisms in their wild form, such as Ralstonia eutropha, Alcaligenes latus, and Burkholderia sacchari.4 More recently, recombinant strains for cost-effective PHA production (properties include: rapid growth, high cell density, ability to use several inexpensive substrates, and simple polymer purification) have been developed by cloning the PHA synthase genes from many microorganisms, including Ralstonia eutropha (recently renamed as Cupriavidus necator).6 The production of PHA by recombinant Escherichia coli harboring R. eutropha can reach 80–90% of the cell dry weight. However, the high oxygen demand during fermentation, due to the high cellular concentration obtained, as well as the need for feed and equipment sterilization, may have a negative impact on the process economics when compared to the production with open mixed cultures.5
The interest in the use of mixed cultures for the production of polyhydroxyalkanoates has increased in recent years. In general, mixed cultures are microbial populations of unknown composition, which are able to perform specific intracellular and extracellular reactions, and are selected by the operational conditions imposed on the biological system. Mixed cultures selected for PHA production can have a high intracellular storage capacity due to operational conditions that limit their primary metabolism.
The storage of intracellular polymers by mixed cultures is observed in wastewater treatment systems for phosphorus removal (enhanced biological phosphorus removal; EBPR), which use alternating anaerobic-aerobic conditions. Polyphosphate-accumulating organisms (PAO) synthesize PHA under anaerobic conditions from external carbon sources and internal glycogen, and consume the PHA in the presence of oxygen or nitrate for cell maintenance, growth, and glycogen replenishment.7 Phosphorus is released in anerobiosis and taken up in aerobiosis/anoxia. Glycogen-accumulating organisms (GAO) are also present in EBPR systems and compete for carbon substrates with PAO. They also cycle PHA and glycogen in a fashion similar to PAO, but GAO do not cycle polyphosphate. Both groups of organisms reach a maximum PHA content of about 20% (grams PHA per gram cell dry weight).8
A PHA content between 30 and 57% was obtained under anaerobic-aerobic conditions in a process named “polyhydroxyalkanotates accumulating bacteria enhanced reactor” (PABER).9 However, the PHA content in this process was not stable. PHA-accumulating bacteria can be enriched in wastewater treatment processes with selectors for bulking control, or in biological nitrogen removal processes (with transient anoxic conditions). However, their capacity for PHA accumulation is quite low (20–30%).9 The polymer yield per substrate consumed is reported to be lower under anoxic conditions than under aerobic conditions.10
Satoh et al.11 and Takabatake et al.9 have proposed a microaerophilic-aerobic process. The oxygen limitation in the first step (microaerophilic) prevents the growth of microorganisms, while carbon is used for PHA production. In the fully aerobic phase, PHA are used as carbon and energy sources for growth and maintenance. A maximum PHA content of 62% was obtained, but, again, the PHA production was not stable.12
PHA storage by activated sludge under fully aerobic conditions can be a particularly important process, if the sludge is submitted to consecutive periods of external substrate accessibility (feast) and unavailability (famine).13 This process is currently known as “aerobic dynamic feeding” (ADF) or “feast and famine”. These conditions allow for the selection of an enriched culture with a high and stable capacity of PHA production (Serafim et al.14 showed that the intracellular PHA content can reach 65% using a pulse substrate feeding strategy). This process can be economically competitive with PHA production from pure cultures and it has the advantages of being simpler and requiring less investment and operating costs.
This paper presents a review on the production of PHA by mixed microbial cultures that are enriched using the aerobic dynamic substrate feeding strategy. Results from our own work and the work of others are presented in order to give a clear status of the process development. Future prospects for the process of PHA production by mixed cultures as a competitive alternative to pure culture production are also discussed.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
2. Mechanisms of Aerobic Dynamic Feeding
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
2.1. Fundamentals
In the presence of an external substrate, microorganisms have the option to use the substrate for growth or for accumulation of intracellular reserves. Conceptual models on PHA synthesis assume that accumulation occurs when growth is limited by external factors such as a lack of nutrients (for example, phosphorus or nitrogen) or internal factors such as an insufficient amount of RNA or enzymes required for growth.15–17 The latter mechanism is the most generally accepted explanation for the storage phenomena in the feast and famine process, where substrate availability alternates periodically. Starvation by substrate for a certain period of time can cause a decrease in the amount of intracellular components (RNA and enzymes) needed for growth. Following this period of substrate limitation, cells faced with an excess of substrate can take it up rapidly. However, the growth rate does not increase at a rate that corresponds with the substrate uptake rate. This difference is related to the fraction of substrate stored or used for cell maintenance. After starvation, storage occurs preferentially instead of cell growth because the amount of enzymes required for storage are lower than the RNA and enzymes needed for growth at maximum rate.15 During the famine period, the stored substrate is used for cell growth and cell maintenance. Operation of a reactor under feast and famine conditions can selectively enrich a mixed population with a high storage capacity.
The high PHA storage capacity of bacteria selected under feast and famine conditions was unequivocally proved in a large number of studies published in the past 10 years. However, some studies have demonstrated that storage of PHA by mixed cultures occurred without any external or internal growth limitation.18 Excess nutrients and sufficient experiment time for physiological cell adaptation were used by the authors to support this fact. Indeed, once the bacteria with PHA accumulation capacity have been selected, they can simultaneously grow and store PHA. Coincident storage of PHA and growth without an external limitation is commonly found in pure cultures such as Alcaligenes latus19 and Azotobacter vinelandii UWD.20
2.2. Metabolism
Currently, there is no experimental evidence regarding the metabolism performed by PHA-producing bacteria in mixed cultures. It is anticipated, however, that the metabolic pathways for carbon consumption and polymer storage are the same as was reported for pure cultures.
In activated sludge submitted to aerobic dynamic feeding, considering acetate as the carbon substrate, the acetyl-CoA (two carbons) produced is partially channelled to the tricarboxylic acid cycle (TCA) for growth and NAD(P)H production, and partially used for PHA production (Figure 1). For polyhydroxybutyrate synthesis, the main polymer formed from acetate, two units of acetyl-CoA condense to produce acetoacetyl-CoA, which is reduced to hydroxybutyryl-CoA at the expense of NAD(P)H, and finally gives the hydroxybutyrate monomer (HB, four carbons). PHB is metabolized when no external substrate is available.
Figure 1. Schematic representation of PHA production from different fatty acids. TCA, tricarboxylic acid cycle; HB, hydroxybutyrate; HMB, hydroxymethylbutyrate; HV, hydroxyvalerate; HMV, hydroxymethylvalerate. Dashed arrows indicate that more than one metabolic step can be involved.
Download figure to PowerPoint
Other short-chain fatty acids such as propionate, butyrate, and valerate can also be used by mixed cultures for the synthesis of different short-chain length PHA (scl-PHA) (Figure 1). Propionate is converted to propionyl-CoA, and when condensed with acetyl-CoA generates a hydroxyvalerate monomer (HV, five carbons). The acetyl-CoA needed for PHA production comes from propionyl-CoA through a decarboxylation step, while some acetyl-CoA is also used in the TCA for growth. Two other monomers, to a lesser extent, can also be produced from propionate: the condensation of two propionyl-CoA moieties gives rise to hydroxymethylvalerate (HMV, six carbons) and the condensation of one acetyl-CoA and one propionyl-CoA can also produce hydroxymethylbutyrate (HMB, five carbons). Butyrate and valerate can be used directly for the production of the corresponding hydroxyacyl-CoA (hydroxybutyryl-CoA and hydroxyvaleryl-CoA, respectively). The latter can be used for PHA synthesis, or partially decomposed for the production of acetyl-CoA and propionyl-CoA for energy, growth, and reducing equivalents.
Besides the aforementioned PHA, other short-chain length PHA can be produced from mixed cultures if longer-chain fatty acids are consumed. In this situation, fatty acids are metabolized by β-oxidation reactions, giving rise to a two-carbon chain precursor, acetyl-CoA, for an even number of carbons, or to acetyl-CoA and propionyl-CoA, for an odd number of carbons. These precursors follow the general metabolism described above, while precursors with a higher number of carbons can yield the corresponding hydroxyacyl-CoA. Valerate can also follow this metabolic pathway.21
In general, the global yield (which accounts for the substrate used for cell growth, polymer synthesis, and maintenance) for acetate, butyrate, and valerate is close to quantitative, indicating that almost all the carbon was used for these three processes. For propionate, the global yield is usually less than 1.0 g substrate/g substrate, due to the production of CO2 from decarboxylation that is not accounted for in respiration measurements. For the production of HV (2C + 3C; the main monomer produced), HB (2C + 2C), or HMB (2C + 3C), there is always the need to convert at least one propionyl-CoA (3C) into acetyl-CoA (2C).
In some studies from the literature, the sum of PHA storage, respiration, and cell growth yields does not match the amount of substrate consumed. It was hypothesized that part of the substrate was accumulated inside the cell in the form of low molecular weight intermediates22 or transformed into exopolymeric substances (EPS).23 It would be beneficial to identify conditions for which the substrate uptake is deviated to form EPS, since this corresponds to a loss of substrate that could have been used for PHA synthesis.
2.3. Microbiology
There are currently more than 300 different microbial species known to synthesize PHA.24 Little information is known about the microorganisms responsible for PHA accumulation under ADF. The first mention of a floc-forming biomass and filamentous micoorganisms in a sequencing batch reactor (SBR) with a high storage rate was made by Dionisi et al.22 Only three papers have reported the microbial characterization in these systems with more detail.25–27
Dionisi et al.25 reported the use of denaturing-gradient gel electrophoresis (DGGE) to characterize the evolution of the microbial community relative to the time of reactor operation. The most representative bands of the gel corresponding to day 80 were excised and sequenced. Uncultured Methylobacteriaceae bacterium clone M10Ba54 (98% homology), Flavobacterium sp. F3 (97% homology), both already described in activated sludge, Candidatus Meganema perideroedes strain Gr28 (99% homology), a filamentous organism able to store PHA, and Thauera chlorobenzoica (95% homology) were identified in this work. While Dionisi and co-workers stated that the dominant genera belong to Thauera, this organism was not described as being able to accumulate PHA and its capacity for PHA storage has not yet been unequivocally demonstrated.
In a recent study,26 after confirming the speciation of the population for PHA accumulation by DGGE, a clone library was constructed from the total DNA extracted from the SBR sludge. The screening of the clones was performed using amplified ribosomal DNA restriction analysis (ARDRA). Each of the 14 different operational taxonomic units (OTU) was sequenced. The most abundant taxonomic group obtained was the Betaproteobacteria, in which at least two different species of Thauera, two species of Alcaligenes, Comamonas sp., Achromobacter sp., and Pseudomonas sp. were present. Thauera sp. was the most abundant organism in the sludge, but neither Thauera sp. nor Achromobacter sp. have been shown to produce PHA. Bacteria from the Gammaproteobacteria group were detected, namely the genera Kluyvera, Pseudomonas, and Acinetobacter, although Kluyvera has also not yet been shown to produce PHA. Finally, Xanthobacter sp., belonging to the Alphaproteobacteria, and Curtobacterium sp. were also identified, where the latter has not yet been associated with PHA accumulation. Further investigation regarding the aforementioned microbial groups is necessary to certify their status as PHA producers.
A different approach was employed by Serafim et al.27 Using fluorescence in situ hybridization (FISH), the dominant organism present in a SBR operated under feast and famine conditions gave a positive signal to the Azoarcus sp. probe. Using Nile blue staining, it was possible to correlate the presence of PHA inclusions with the Azoarcus sp., and thus confirm this organism as a PHA producer.
The feast and famine process was also applied to pure cultures like Amaricoccus kaplicensis28 and Paracoccus pantothrophus.29 Both cultures showed a similar behavior to that of the mixed cultures submitted to ADF: PHA storage and growth while the substrate was available, and PHA consumption during famine conditions. The abundance of these strains in a mixed-culture process has not yet been evaluated.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
3. Process Operation
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
3.1. Substrates
Substrate cost is the main cost in large-scale PHA production. It has been estimated to be about 40% of the total PHA production costs.30 Therefore, a more cost-effective process should include the selection of cheap substrates, which can be effectively used by the microorganisms to synthesize PHA at high productivities, and that the resulting PHA polymer properties are suitable for a wide range of industrial applications. In the past decade, a wide variety of low-cost carbon substrates (e.g., renewable carbon sources), such as agricultural and industrial residues or by-products (e.g., starch, tapioca hydrolysate, whey, xylose, molasses, malt, and soy wastes), have been tested for PHA production by pure cultures.31,32 The PHA yields varied from 0.11 to 0.33 g PHA·g−1 substrate and volumetric productivities ranged from 0.05 to 0.90 g PHA · L−1 · h−1.31 To the best of our knowledge, only sugar cane molasses is being used at an industrial scale for PHB production (PHB Industrial S.A., personal communication).
Waste materials or industrial by-products can also be used in mixed-culture processes for PHA production. However, unlike pure cultures, carbohydrates are not directly stored as PHA by mixed cultures, they are preferentially accumulated as glycogen.33,34 PHA production from sugar-enriched raw materials requires a previous anaerobic fermentation step for their transformation into volatile fatty acids (VFA). This is the reason why the majority of the studies related to PHA production by mixed cultures are based on the use of organic acids.
Acetate is one of the most well-studied substrates for PHA production by mixed cultures submitted to feast and famine conditions. Table 1 shows examples of studies carried out with acetate and other substrates. Acetate is transformed into a homopolymer of PHB. The polymer production rate and yield depend on the reactor operating conditions used. The maximum values of these parameters were obtained by Serafim et al.14 where acetate was fed pulse-wise: polymer yield was 0.56 g PHB · g−1 substrate and the specific productivity was 0.77 g PHB · g−1 cell dry weight · h−1. Productivities in the range of 0.015–0.028 g PHB · g−1 cell dry weight · h−1 have been reported for pure cultures.35,36
Table 1. Biopolymer production by mixed cultures using different substrates and operational characteristics. Substrates HRT SRT Cycle length Substrate PHA storage yield PHA production rate PHA content PHA composition Ref.
h d h g · L−1 g PHA · g−1 substrate g PHA · g−1 cell dry weight · h−1 % mol:mol
a)Units of mg COD · mg−1 COD.
b)Units of g COD · g−1 COD · h−1.
c)Units of COD · COD−1.
d)Not reported.
e)Units of g COD · L−1.
f)Microaerophilic/aerobic.
g)Units of g PHA · g−1 COD.
Acetate/propionate/lactate(40%/20%/40%) 24 1 2 7.5 0.29 0.52 46.0 P(HB/HV)69:31 37
24 1 2 17.6 0.46a) 0.39b) 45.0c) NRd) 26
Acetate Batch – 2 ≈0.33e) 0.31 0.39 NR PHB 37
Propionate Batch – 2 ≈0.25e) 0.35 0.068 NR PHV 37
Lactate Batch – 2 ≈0.25e) 0.20 0.43 NR PHB 37
Acetate 8 1 4 0.38 0.22 0.26 11.8–20.4 PHB 51
Acetate 8 6.9–98 4 0.38 0.36–0.42 0.26–0.44 4.5–6.2 PHB 51
Acetate 24 1 2 1.0–8.0 0.32–0.38 NR NR PHB 22
Acetate NR 3.8 ≈4 ≈0.54 0.34–0.49 0.28 36.3 PHB 50
Acetate NR 3.8–19.8 4 0.18 0.29–0.43 0.22–0.26 NR PHB 49
Acetate NR 4 4 0.18 0.43–0.51 0.52–0.63 40.0 PHB 48
Acetate Batch – 6 0.08-0.20 0.38–0.44 0.43–0.52 47.0 PHB 39
Acetate 24 10 12 0.90 0.52 0.49 31.9 PHB 14
Acetate Batch – 12 3 × 1.8 0.56 0.77 65.0 PHB 14
Propionate Batch – 12 0.63 0.30 0.09 13.6 P(HB/HV/HMV)12: 61:27 4
Propionate + acetate Batch – 12 0.75 0.37 0.46 25.4 P(HB/HV/HMV)54:33:13 4
Acetate + glucose NR 6.1 4 0.36 + 0.36 0.43–0.65 0.16–0.27 NR PHB 42
Glutamic acid Batch – NR 1.2e) 0.058 0.0046 NR PHB 37
Butyrate Batch – 12 0.53 0.44 0.068 14.5 PHB 4
Valerate Batch – 12 0.47 0.37 0.072 14.3 P(HB/HV/HMV)32:52:16 4
Fermented food waste (MAA/AE)f) 12 2–12 6 0.68e) 0.086g) NR 37 PHB/PHA(0.74–0.77) 44
Fermented olive oil mill effluent Batch – 2 NR 1.05 0.57 54 P (HB/HV)96:4 43
Fermented molasses Batch – 12 4.0 0.46 0.10 31 P (HB/HV)47:53 This work
As shown in Table 1, acetate is preferentially transformed into PHB, which is a highly crystalline and stiff material, leading to brittleness. This lack of flexibility limits its range of industrial applications. More ductile, easier to mould, and tougher polymers can be obtained by incorporating HV units in the polymer chain. Copolymers of poly(3HB-co-3HV) can be synthesized by bacteria from higher-chain VFA, such as propionate, butyrate, and valerate (Table 1).
Propionate as the sole substrate was fed to a mixed-culture SBR operated under feast and famine conditions.4,37 A homopolymer of HV was produced in the first study while a copolymer of poly(3HB-co-3HV) or a terpolymer (HB/HV/HMV) was produced in the latter. Although HV is the main fraction produced from propionate metabolism, the formation of hydroxybutyryl-CoA, the precursor of the HB units, from the condensation of two acetyl-CoA units appears likely (see Figure 1 and Section 2.2). The difference in polymer composition observed in both studies is probably due to differences in the microbial population structure. The specific storage rate was higher in the latter study4 than in the first37 (0.09 g PHA · g−1 cell dry weight · h−1 and 0.068 g PHA · g−1 cell dry weight · h−1, respectively), but storage yields were similar (0.35 g PHA/g substrate37 and 0.30 g PHA/g substrate4). A maximum PHA content of 41% was obtained with propionate (unpublished results).
Propionate is often used in pure-culture processes as a precursor for the synthesis of HV units. For instance, R. eutropha produces poly(3HB-co-3HV) with approximately 45% 3HV content when propionic acid is used as the sole carbon source.38 This fraction is much lower than that of the copolymer produced by activated sludge, where 84% 9 and 83%4 of 3HV were produced from propionate as the sole carbon source.
Butyrate and valerate have also been used in the feast and famine process. Butyrate, was converted into a homopolymer P(3HB).4 The polymer yield (0.44 g PHA · g−1 substrate) was lower than the value obtained by the same culture with acetate (0.52 g PHA · g−1 substrate), but higher than feeding with propionate (0.30 g PHA · g−1 substrate). Valerate was stored by this culture as a terpolymer of P(HB/HV/HMV). The yield of polymer on substrate was 0.37 g PHA · g−1 substrate. The synthesis of PHA from propionate requires a decarboxylation step, which may also be the case for valerate. This likely explains the lower polymer yield from these two substrates.
In addition to VFA, other single substrates (found in some fermented streams) such as lactate, ethanol, and glutamate can be converted into PHB in the feast and famine process, but with a very low storage yield: 0.20 g PHA · g−1 substrate for lactate, 0.25 g PHA · g−1 substrate for ethanol, and 0.058 g PHA · g−1 substrate for glutamate.18,37
Mixtures of substrates are generally used to obtain copolymers with different monomer compositions, aiming at the tailored synthesis of PHA with given target mechanical properties. In a feast and famine process, a mixture of 50% of propionate and acetate was transformed in a copolymer poly(3HB-co-3HV) with a molar fraction of HV: HB of 51: 49.4 The quantity of HV per mole of carbon consumed was higher when acetate and propionate were supplied simultaneously than when only propionate was fed. Indeed, when both substrates are present, the acetyl-CoA units required for HV synthesis can be produced directly from acetate, leaving more propionyl-CoA available for hydroxyvaleryl-CoA synthesis (see Figure 1). The storage yield was 0.30 g PHA · g−1 substrate.
Chua and co-workers40 changed the relative concentration of valeric and butyric acids in a fed-batch reactor. When butyric acid was used as the sole carbon source, only the homopolymer P(3HB) was produced. The molar fraction of 3HV in the polymer tended to increase with the valeric acid concentration in the medium. Valeric acid alone resulted in a copolymer poly(3HB-co-HV) with a HV molar fraction of 54%. The incorporation of valeric acid in the substrate affected the polymer yield per substrate: it decreased from 0.69 to 0.32 g PHA · g−1 substrate as the valeric acid fraction increased from 0% to 80%, respectively.
A mixture of acetate, propionate, butyrate, and valerate was fed to a batch reactor and converted into a copolymer of poly(HB-co-HV) (with a storage yield of 0.39 g PHA · g−1 substrate).41 In this case, acetate and propionate were completely removed, while only a small fraction of butyrate and valerate was consumed. Mixtures of acetate and glucose were fed in a study carried out by Carta et al.42 Both glycogen and PHB were stored by the sludge (the storage yields were 0.51 g PHB · g−1 substrate and 0.74 g glycogen · g−1 substrate).
All of the aforementioned studies were carried out with simulated feeds supplemented with the organic substrates. The production of PHA by mixed cultures under feast and famine conditions was also studied with real wastewater. Studies involving carbohydrate-based wastes include a prefermentation step in order to produce VFA. This is the case of the study conducted by Dionisi et al.,43 in which olive oil mill effluents (OME) were converted in a continuous anaerobic reactor to a mixture of organic acids (acetic, butyric, propionic, isobutyric, and valeric), which were then fed to a batch reactor inoculated with an enrichment of PHA-accumulating bacteria. The storage yield [1 g PHA · g−1 VFA on a chemical oxygen demand (COD) basis] was much higher than the one obtained with synthetic substrate containing acetic, lactic, and propionic acids (0.39 g PHA · g−1 VFA on a COD basis).26 This was explained by the conversion of organic compounds other than VFA, which were also present in the fermented OME. The polymer content obtained was 54%.43 The nonfermented OME (containing only acetic and lactic acids as VFA) was also used, but the total amount of PHA produced (in mg COD · L−1) was three times lower than that obtained from the fermented OME.
The use of sugar cane molasses as a substrate for PHA production was investigated in our research group (unpublished results). A three-stage process was implemented consisting of acidogenic molasses fermentation [carried out in a continuos stirred tank reactor (CSTR)] followed by PHA production in a batch reactor inoculated with a PHA-enriched culture (in a SBR reactor under dynamic feeding conditions). Clarified fermented molasses samples that were produced under different CSTR steady state conditions (pH ranging from 5 to 7) were fed to the batch reactor for PHA accumulation. Polymer content varied from 27 to 31% PHA per cell dry weight and PHA yields on substrate ranged from 0.27 to 0.50 g PHA · g−1 VFA. The best result, with PHA yield per substrate of 0.49 g PHA · g−1 VFA and PHA cell content of 31%, was obtained with the effluent produced at pH 5, where a mixture of acetate, propionate, butyrate, and valerate (molar fraction of 0.52/0.20/0.26/0.01, respectively) led to the production of a copolymer of poly(HB-co-HV) with molar fractions of 47% HB and 53% HV. The global PHA yield was 0.13 g PHA · g−1 molasses.
Other real wastes that have been used for the production of PHA include fermented food waste44 and domestic wastewater.45 The storage yield was about 0.05 g PHA · g−1 COD (25 g PHA · kg−1 of dry food waste) in the first case and 0.44 g PHA · g−1 COD in the latter study.
3.2. Reactor Operational Strategies
The majority of the studies related to the production of PHA by the mixed cultures submitted to feast and famine conditions were carried out in SBR, operated with cycles of feeding, reaction, settling, and draw. The length of the total cycle varies from 2 to 12 h, depending on the study (Table 1), although the substrate feeding period was always very short. Figure 2 shows the carbon, oxygen, and ammonia transformations in a typical cycle of a SBR operated under dynamic substrate feeding. In this process, the sludge retention time (SRT) can be varied independently of the hydraulic retention time (HRT).
Figure 2. Evolution of acetate (•), PHB (□), ammonia (▵), oxygen (◊) and pH (▾) in a typical cycle of a sequencing batch reactor operated under feast and famine conditions.
Download figure to PowerPoint
In almost all of the published studies, a pulse-wise addition of substrate was carried out in batch reactors inoculated with biomass taken from SBRs enriched in PHA-accumulating bacteria. In some cases, more than one pulse of substrate was supplied in order to obtain higher PHA content and to avoid substrate inhibition.14,41,46 The addition of substrate by pulses has been controlled on-line by the rise in oxygen concentration in a SBR reactor operated for a period of almost 1 month.14 In this case, the length of the famine phase was fixed (10 h), while that of the feast phase varied freely according to the time required to consume all of the carbon supplied. Since the length of the famine phase was long and constant, the selective pressure that favors PHA-accumulating microorganisms was maintained during the reactor operation. As a consequence, a population with a high storage capacity was selected, with about 65% of PHA inside the cells.
The productivity of the mixed-culture process depends mostly on the selection of microbial cultures with a high storage capacity and on the cell concentration used for PHA accumulation. The “side stream” PHA production process, based on continuous culture selection followed by batch PHA accumulation, achieves both aims. This concept was used by Dionisi et al.26,37,43 for the production of PHA from olive oil mill effluents and is currently being addressed in our group using sugar cane molasses. The process consists of three stages: a prefermentation step, where the organic waste is converted in a continuous anaerobic reactor to a mixture of organic acids; SBR operation under feast and famine conditions to enrich and produce sludge with a high storage capacity; and a third stage where a batch reactor is fed with a high substrate concentration in order to increase the polymer content in the biomass, and where the excess sludge produced in the second step is introduced. The polymer production rates for OME and sugar cane molasses were 0.57 g PHA · g−1 cell dry weight · h−1 and 0.14 g PHA · g−1 cell dry weight · h−1, respectively.
A similar concept, but comprised of only two stages, was proposed by Chua et al.47 In the first step, an anaerobic-aerobic SBR fed with real municipal wastewater was enriched in PHA-accumulating organisms through manipulation of the operational conditions (such as SRT, pH, and substrate concentration). In the second stage, the enriched sludge was fed with acetate in a batch aerobic reactor. A maximum PHA content of 30% and a specific productivity of 0.050 g PHA · g−1 cell dry weight · h−1 were obtained in the second stage of this process.
The economic viability of the PHA production process by mixed cultures is highly dependent on the ability to select a microbial population with both high cell growth rate (to achieve high cell concentration) and high PHA storage rate. Only then are volumetric productivities sufficiently competitive for large-scale production. Following the concept proposed by Dionisi et al.,25 the ideal configuration would be a small SBR for the enrichment of organisms with a high cell growth rate and high PHA storage capacity. The biomass produced per cycle should match the biomass required to inoculate a larger bioreactor for production purposes. The latter reactor should be operated in fed-batch mode, in a similar fashion to pure cultures, for example, with a cell growth phase followed by a PHB production phase with cell growth limitation. The operational strategy is obviously different in these two reactors. The best strategy for sludge enrichment seems to be the adoption of a feast and famine strategy with ammonia in excess. The excess of ammonia increases the selective pressure of organisms with a high cell growth rate, and, when combined with the feast and famine regimen, it also favors the organisms with a high PHA storage capacity. However, it is unclear the impact of the physiological adaptation previously discussed in Section 2.1 and how it can be incorporated in this process scheme. This is an important issue still open to discussion.
3.3. Operational Parameters
For the same kind of substrate, the polymer yield and productivity depend on the reactor operating conditions such as the substrate concentration, organic loading rate (OLR), carbon to nitrogen ratio (C/N), SRT, pH, oxygen concentration, and temperature.
The amount of substrate supplied can determine the quantity of polymer produced. A linear relationship between the amount of polymer produced and the substrate consumed, in the range of 0.90–2.70 g · L−1 of acetate, was observed in a feast and famine process.14 Similar behavior was also observed in a different range of acetate concentration by Dionisi et al.22 (1–8 g · L−1) and Beun et al.48 (0.18–1.44 g · L−1).
The results of Serafim et al.14 were obtained in batch experiments using sludge from a SBR operated at constant organic loading rate (1.8 g substrate · L−1 · h−1). The effect of the organic loading rate (ranging from 7.5 to 27.6 g substrate · L−1 · h−1) was evaluated by Dionisi et al.26 in a feast and famine SBR. The maximum polymer concentration and storage rate occurred at 17.6 g substrate · L−1 · h−1 and decreased sharply for higher OLR values. The length of the feast and famine phases were kept constant (about 2 h) in all assays and therefore the increase in the OLR led to a relative decrease in the famine phase duration in such a way that the selective pressure that favored PHA-storing microorganisms was progressively reduced. This gave rise to the enrichment of fast-growing microorganisms with low PHA storage capacity. In fact, ammonia was never limiting along the entire feast and famine cycle in these experiments and the cell growth on external substrate continued until its exhaustion. These results illustrate well the importance of having a starvation period for the selection of organisms with a high storage capacity.
The SRT can also have an impact in the storage yield, for a given OLR. It can be anticipated that the shorter the SRT the higher the cell growth rate and the less the substrate is used for storage. This behavior was observed for a pure culture of Paracoccus pantotrophus under dynamic feeding.29 These authors showed that higher cell growth rates resulted in a lower PHB content. Indeed, higher PHB yields were produced at longer SRT when the cells were growing more slowly (average µ less than 0.1 h−1 for SRT higher than 0.5 d). These results are in accordance with those obtained by Beun et al.49 Using a metabolic model, Beun et al.48 estimated that, at a specific acetate uptake rate of at least 0.40 g acetate · g−1 cell dry weight · h−1, where the SRT was higher than 2 d, the yield of PHB from acetate was constant at 0.43 g PHA · g−1 substrate, was independent of the SRT, and, thus, was independent of the specific growth rate. Below this threshold value of 2 d, the PHB storage yield and productivity increased sharply with the SRT. These authors confirmed experimentally through varying the SRT from 3.8 to 19.8 d that the PHB yield per substrate and specific productivity were almost constant. Dionisi et al.22 obtained a relatively constant storage yield in a SRT range of 0.37–3 d.
On the other hand, Chua et al.47 showed that sludge at a SRT of 3 d accumulated more PHA (PHA content 31%) than at a SRT of 10 d (PHA content 21%). However, the biomass concentration used in the experiment with the SRT of 10 d was almost four times higher than that with the SRT of 3 d (2 500 mg MLSS · L−1 vs. 700 mg MLSS · L−1). Therefore, since the same amount of substrate was supplied in both experiments, it would be rational that the amount of PHA accumulated per unit of biomass would be less for the SRT of 10 d.
As explained in Section 2.1, the mechanisms of PHA storage in the feast and famine process occur without the need for external nutrient limitation. This is the reason why in almost all of the studies SBR were operated without ammonia or phosphorus limitation. Exceptions are the work of Serafim et al.14 and Lemos et al.,4 where ammonia limitation was used to control the fraction of substrate driven toward cell growth. It was shown in batch experiments that, in the range of 0–40 mg N · L−1, the growth yield increased proportionally with ammonia concentration while the storage yield decreased. On the other hand, Dionisi et al.25 compared one batch reactor fed with an excess of ammonia with another reactor fed without a nitrogen source and observed no significant impact on the efficiency of PHA storage. The difference observed in both studies can be attributed to differences in the method of selecting the PHA-accumulating bacteria. In the studies of Serafim et al.14 and Lemos et al.,4 ammonia was only available at the beginning of the feast phase, whereas in the studies of Dionisi et al.22,25,37 and Beun et al.48,49 ammonia was always present along the feast and famine cycles. The first operational procedure could have enriched the reactor with two different populations: one that is able to simultaneously accumulate PHA and grow, while the other is able to accumulate PHA with less cell growth. The second procedure may have only enriched for the first microbial population.
Accumulation of PHA at different C/N mass ratios (from 20 to 140) was studied using real wastewater (by adding NH4Cl).40 The polymer yield on substrate increased linearly up to a C/N of 100 and reached a maximum of 0.11 g PHA · g−1 substrate. The polymer content increased with the C/N ratio to a maximum of 38%.
In the majority of the studies performed with mixed cultures under feast and famine conditions, the pH is controlled.26,37,42,48,49 If not controlled, pH changes along the SBR cycle as shown in Figure 2. The pH tends to increase during the feast phase and then it stabilizes. The effect of pH on the PHA content was studied by Chua et al.,47 using acetate as the substrate. They found that, through controlling the pH at 6 or 7, the PHA content (less than 5%) was lower than at pH 8 or 9 (25–32%). This is in accordance with the results obtained by Serafim et al.,14 in which the polymer yield per substrate and the intracellular PHB content were higher at pH 8 than at pH 7, and increased sharply when the pH was not controlled (it varied from 8 to 9.5). The effect of pH on the degree of acetic acid dissociation (diffusion of undissociated acetic acid into the bacterial cells) and consequent inhibition of the bacteria at low pH, was given as the hypothesis for the decrease in PHA content.47 Since in the noncontrolled pH reactors, the pH varied between 8 and 9.5, inhibition by the substrate was unlikely to occur. Nevertheless, from the point of view of the operation simplicity, it is more advantageous to have a non-pH-controlled system.
The effect of dissolved oxygen concentration (DO) on PHA production was studied by Third et al.50 The yield of PHA on acetate was 0.49 g PHA · g−1 substrate when oxygen was limiting and 0.34 g PHA · g−1 substrate under excess oxygen. In the latter case a higher fraction of acetate was used for biomass growth. A metabolic model showed that at low DO concentrations, the limited availability of ATP prevented significant biomass growth and that most ATP was used for acetate transport into the cell. In contrast, high DO supply rates resulted in a surplus of ATP and then higher cell growth rates, thereby decreasing the PHB yield.
The effect of temperature on PHA production has also been evaluated. The increase of temperature from 15 to 35 °C led to a decrease in the yield of PHB on acetate from 0.43 to 0.072 g PHA · g−1 substrate and a decrease in the specific productivity from 0.12 to 0.060 g PHA · g−1 cell dry weight · h−1.17 The yield of biomass also decreased with temperature increase. Low temperatures (between 15 and 20 °C) allow for a less costly process (through savings in energy) and favors polymer produced in the cells, thus increasing the PHA volumetric productivity.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
4. Process Modeling and Control
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
4.1. Mathematical Modeling
Models are essential to provide estimations of important but otherwise inaccessible process/physiological parameters, or to support the design of control strategies. The Activated Sludge Model (ASM) number 3 is one of the most referenced models for the quantitative analysis of activated sludge processes.52 It also provides a sound basis for studying PHA production by mixed microbial systems. The full ASM3 model accounts for the existence of two groups of organisms (heterotrophic and autotrophic organisms) and attempts to simultaneously describe the sludge production, nitrification and denitrification processes, as well as the storage of organic substrates (PHA, glycogen), either through the aerobic or anoxic storage of COD. The ASM3 can be simplified to aerobic heterotrophic storage of acetate in the form of PHB by eliminating the autotrophic organisms. In the original ASM3 model, the COD is first stored in the form of PHB and only then can PHB be metabolized for cell growth. This reaction mechanism may be consistent in stoichiometric terms, but is unrealistic in kinetic terms. Modifications were introduced to allow for the simultaneous occurrence of cell growth and PHB storage on COD.17,45
A number of detailed metabolic models describing catabolism, biomass synthesis, and PHB storage have been proposed in the literature. Examples are given by van Aalst-van Leeuwen et al.29 for pure cultures of Paracoccus pantotrophus and by Beun et al.48,49, Third et al.,50 and Dias et al.53 for mixed microbial cultures. Tables 2 and 3 compile the information of a full metabolic model for aerobic PHB storage and degradation. These models are based on seven metabolic reactions (see Table 2): acetate uptake, respiration, oxidative phosphorylation, biomass precursor synthesis, biomass precursor polymerization, aerobic PHB formation, and aerobic PHB degradation.48,49 The corresponding theoretical yield coefficients, which are derived from mass balances on acetyl-CoA, biomass precursors, ATP, NADH2, and from the overall redox balance, are compiled in Table 3. Dias et al.53 compared the theoretical yield values obtained from the equations in Table 3 with experimental values calculated from a mixed microbial culture subjected to a feast and famine feeding regimen with a low C/N ratio. The main conclusion drawn was that the culture was able to store PHB at a maximum theoretical efficiency (YPHB/S = 0.75 C mol · C mol−1 or 0.54 g PHB · g−1 substrate and δ = 3 mol ATP · mol−1 NADH2), whereas the biomass growth on acetate showed a lower efficiency than expected (YX/S = 0.42 C mol · C mol−1 or 0.32 g X · g−1 substrate). This result highlights the fundamental effect of the feeding strategy on culture selection. The reaction kinetics compiled in Table 3 are essentially based on previous studies by Serafim et al.14 and Dias et al.53 This kinetic model assumes explicit feast and famine periods, and was therefore validated under such conditions with the data from five calibration batch tests and three prediction batch tests. An illustrative example of two batch experiments, with the acetate and ammonia fed pulsewise, is given in Figure 3 using the parameters of Table 4. It is worth noting that in both experiments, ammonia eventually limits cell growth and the transition between the cell growth and nongrowth periods is well described by the model.
Table 2. Metabolic model for PHB production by mixed microbial cultures. Process description Reaction
Acetate uptake 0.5 · HAc + 0.5 · CoA + 1 · ATP → 0.5 · AcCoA + 0.5 · H2O
Biomass precursors synthesis 0.6335 · AcCoA + 0.2 · NH3 + 0.66 · ATP + 0.301 · H2O → 1 · CH1.4N0.2O0.4 + 0.267 · CO2 + 0.534 · NADH2
Biomass precursors polymerization
Respiration 0.25 · AcCoA + 0.75 · H2O → 0.5 · CO2 + 1 · NADH2
Oxidative phosphorylation 1 · NADH2 + 0.5 · O2 → 1 · H2O + δ · ATP
Aerobic PHB storage 0.5 · AcCoA + 0.25 · NADH2 → 1 · CH1.5O0.5
Aerobic PHB consumption 1 · CH1.5O0.5 + 0.25 · ATP → 0.5 · AcCoA + 0.25 · NADH2
Table 3. Theoretical yield coefficients and reaction kinetics. Acetatea) NH3 X PHB O2 CO2 Kinetics
a)S, Acetate; PHB, polyhydroxybutyrate; N, ammonia; O, oxygen; C, carbon dioxide; X, active biomass; ATP, adenosine triphosphate; δ, efficiency of oxidative phosphorylation (mol ATP · mol−1 NADH2); µj- specific growth rate on component j (C-mol · (C-mol · h)−1); µj,max, maximum specific growth rate on component j (C-mol · (C-mol · h)−1); Sj, concentration of component j in the liquid phase (mmol · L−1); Kj, half-saturation constant for component j uptake (mmol · L−1); Kj,k, half-saturation constant for component j in the k phase (N mmol · L−1); qj, specific reaction rate for component j (C mol · (C mol · h)−1); qj,max, maximum specific reaction rate for component j (C mol · (C mol · h)−1); fPHB, intracellular PHB fraction (C mol · C mol−1); fPHB,max, maximum intracellular PHB fraction (C mol · C mol−1); , PHB production saturation order constant (dimensionless); Y, yield of component i on component j in the k phase (mol · mol−1); mj, maintenance coefficient on component j (mol · (mol · h)−1); mj,max, maximum maintenance coefficient on component j (mol · (mol · h)−1); KI, acetate inhibition constant (C mmol · L−1); k, kinetic constant for PHB degradation (h−1); n, reaction order of PHB degradation for maintenance (dimensionless).
Feast phase
Cell growth 1 –
PHB storage – – 1
Maintenance on S −1 – – – −1 1
Famine phase
Cell growth – 1
Maintenance on PHB – – – −1 −1 1.125
Figure 3. Mathematical model simulation (Tables 2, 3, and 4) and experimental data of two batch experiments. Batch one (ˆ), batch two (▴).
Download figure to PowerPoint
Table 4. Parameter values. a)Adapted from Third et al.50
Theoretical yields (δ = 2 mol ATP · mol−1 NADH2)48,49
mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1
Theoretical yields (δ = 3 mol ATP · mol−1 NADH2)56
mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1, mol · C mol−1
Experimental yields56
mol · C mol−1, mol · C mol−1, mol · C mol−1
Kinetic parameters56
µS,max = 0.11 C mol · (C mol · h)−1, KS = 0.062 C mmol · L−1, KN,OS = 2.17 N mmol · L−1, KO = 0.019 mmol · L−1a), qPHB,max = 0.56 C mol · (C mol · h)−1, fPHB,max = 2.47 C mol · C mol−1, = 3.85 (dimensionless), qS,max = 0.74 C mol · (C mol · h)−1, mATP = 0.02 mmol ATP · (C mmol · h)−1, µPHB,max = 0.081 C mol · (C mol · h)−1, KPHB = 0.0001 C mol · (C mol · h)−1, KN,OP = 0.72 N mmol · L−1, KI = 0.062 C mmol · L−1, k = 0.067 C mol · (C mol · h)−1, n = 1.94 (dimensionless)
The model does not consider population dynamics, which may be an important source of variability in open microbial systems. The results obtained, however, support the approximation of a homogeneous culture with an average metabolism, since the same model with the same set of parameters could predict a wide range of experiments performed under different feeding conditions. This also substantiates the effectiveness of the feast and famine feeding regimen in selecting the heterotrophic organisms with a high storage capacity in long-term operation. Fundamental challenges are still the modeling of PHA production with mixed substrates and the incorporation of population dynamics. This latter point is crucial to the design of novel transient feeding strategies for improved organism selection. Also, a more mechanistic description of substrate modulation for cell growth and PHA storage is lacking. An attempt was made by van Loosdrecht and Heijnen54 by introducing a pool of enzymes in the model, but experimental validation is still lacking.
4.2. Process Control
Process monitoring and control may be key factors for achieving high productivity, robustness, and reproducibility in the PHA production process by mixed microbial cultures. Closed-loop pH, DO, and feeding control are well documented for PHA production with pure cultures,55–61 but only a few applications have been reported so far for PHA production by mixed microbial cultures.14,53
In Section 3.2, the two main reactor operational objectives were discussed, each one presenting specific control challenges. One is the control of population dynamics in the SBR for the enrichment of heterotrophic organisms in the sludge with a high cell growth rate and with a high PHA storage capacity. The other is the simultaneous maximization of the final intracellular PHA content and the volumetric productivity in the production reactor. The main control variables that may be used to fulfill the aforementioned objectives are, for the former case, the SRT, the SBR cycle duration, DO control, and the feeding strategies of carbon and nitrogen sources in the feast and famine phases. For the production reactor, they are mainly the control of DO and of the feeding strategies for the carbon and nitrogen sources. Among these, the feeding of carbon and nitrogen sources are the most important, and also the most challenging. The on-line measurement of VFA and ammonia, although possible, is not yet widespread. This section focus on closed-loop control strategies based on indirect but easily accessible on-line measurements, namely DO and pH. At the end of the section, prospects for advanced monitoring and control strategies are briefly discussed.
4.2.1. Bioreactor Dynamics
Bioreactor dynamics can be established from transient material balance equations, normally assuming that the liquid and gas phases are well mixed. Table 5 compiles the material balance equations of biomass, substrate, ammonia, oxygen, carbon dioxide, and PHA content for fed-batch operation, again assuming a homogeneous culture with an averaged metabolic activity. These equations establish a dynamical relationship between the most important operating (control) parameters with process performance, and are thus essential to quantitatively study bioreactor optimization and control.
Table 5. Bioreactor dynamics for fed-batch operation with acetate and ammonia feeding. XH, Active biomass concentration; D, dilution rate; Sj0, initial concentration of component j; OUR, oxygen uptake rate; OTR, oxygen transfer rate; CPR, carbon dioxide production rate; CTR, carbon dioxide transfer rate; V, volume; HPR, hydrogen production rate; pKAc, pKNH4, pK1CO2, pK2CO2, log of ionization equilibrium constants.
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Equations adapted or taken from Pratt et al. (2004)67:
(10)
(11)
(12)
(13)
4.2.2. DO-Based Feeding Control
DO-stat carbon source feeding control is often a necessity for PHA production with pure cultures.55,56,62 In this control strategy, the DO is kept constant at a desired set point by continuously adjusting the feeding rate of the carbon source. Cell densities higher than 100 g cell dry weight · L−1 (active cells plus intracellular polymer) are likely to result in oxygen transfer limitation. The DO concentration becomes highly sensitive to the feeding of the carbon source, so enabling a DO-stat feeding strategy. In the case of mixed cultures, a DO-stat feeding strategy has not been previously reported due to the much lower cell densities (typically < 7 g cell dry weight · L−1). The kLa value required to maintain the DO concentration above 20% of saturation is expected to be in the range of 150–260 h−1 during the cell growth phase and 120–220 h−1 during the PHA production phase for an oxidative phosphorylation efficiency of 3 and 2 mol ATP/mol NADH2, respectively. These values were calculated using the model of Table 3 and 4 assuming an active biomass concentration of 10 g cell dry weight · L−1. Such high kLa values, combined with relatively high oxygen limitation levels due to the formation of cellular flocs, are likely to potentiate the implementation of a DO-stat feeding controller.
When the cell densities are low, an on/off DO-feeding controller is a possible alternative to the continuously fed DO-stat. Serafim et al.14 implemented such an automatic DO-feeding control system in a SBR, which proved to be easy to operate and very reliable. In this control strategy, the carbon source is fed pulsewise. A new pulse is fed whenever a sharp increase in the DO concentration is detected. In this way it is possible to extend the feast phase (by chosing the number of pulses) whereas the duration of the famine phase is mantained constant. The acetate depletion was detected on-line by the sudden increase in the DO concentration. A recursive least-squares algorithm calculated the slope of the DO signal on-line. Whenever the slope increased above a threshold value, a pulse of acetate was fed to the reactor. This strategy allowed extention of the feast phase while avoiding acetate limitation. The PHB accumulation was shown to be essentially unaffected by the periodic depletion of acetate. A PHB content of 65% g PHB · g−1 cell dry weight was obtained using this control strategy. This control loop was implemented in a SBR automation software developed in our group in Labview™. The software is based on a supervisory controller that schedules all SBR cycle events and associated control loops including the aforementioned DO-feeding controller in the feast phase.
4.2.3. pH-stat Feeding Control
Depending on the interference of the buffer systems, the feeding of organic acids can be controlled on the basis of pH measurements.55–57,63–65 There are three main processes that affect the pH in a mixed microbial culture for PHA production:66,67 the uptake of the substrate (an organic acid), the NH3 assimilation, and the CO2 production and transfer. The uptake of the organic acid will cause a decrease in the concentration of the hydrogen ion in solution. On the contrary, the assimilation of NH3 increases the concentration of the hydrogen ion in solution since the H+ of the NH is not assimilated by the cells. Finally, the net accumulation of CO2 in the liquid phase increases the H+ concentration through the bicarbonate and carbonate acid/base systems. For a fed-batch system, not only the reaction terms but also the feeding and dilution terms must be taken into account. The rate of H+ production, HPR, is given by Equation 10 to 13 (adapted from Pratt et. al.,67 for fed-batch operation) in Table 5. The m, p, and n parameters represent the contribution of S, NH, and CO2 dynamics to the H+ production through the respective acid/base equilibrium systems.
If left uncontrolled, the pH tends to increase in batch cultures due to the uptake of the organic acid, resulting in a negative HPR. Thus, the feeding of organic acids on the basis of pH measurements is possible. An example of this application is provided by Sugimoto et al.57 for PHA production by pure cultures of R. eutropha. The authors managed to control the acetate concentration at approximately 1 g · L−1 by implementing a pH-stat with a feed of CH3COOH/CH3COONH4/KH2PO4. Other successful applications were reported by Tsuge et al.63,64 for the feeding of acetate and by Kobayashi et al.65 for the feeding of propionate. In Choi and Lee,61 a mixture of glucose, propionate, and acetate was fed based on a pH-stat in an E. coli culture.
In our own process, a linear relationship between the acetate concentration and pH is observed with a characteristic slope that is dependent of the C/N ratio in the feeding (see Figure 4a). This supports the possibility of implementing a pH-stat feeding controller. Furthermore, the acetate-pH slope correlates linearly with the C/N ratio in the feeding in several experiments (see Figure 4b). The depletion of ammonia results in an increase of this slope, and may be detected on-line (Figure 4a). The pH signal is thus very informative and has the potential to be used for on-line monitoring and control of both carbon source and ammonia.
Figure 4. (a) Acetate concentration over pH for two experiments with different C/N ratios. Symbols are measurements; full lines are linear fits. (b) Characteristic acetate-pH slope as a function of C/N ratio in the feeding for eight different batches. Symbols are the values obtained for each batch; full line is the corresponding linear fit; dashed lines are 95% confidence intervals.
Download figure to PowerPoint
4.2.4. Prospects for Advanced Model-Based Control
The availability of a reliable dynamic model introduces the possibility of performing advanced control studies. Fundamental challenges are the design of optimal feeding strategies (mainly VFAs, ammonia, and oxygen) for sludge enrichment in the SBR, and also for obtaining high volumetric productivities in the production reactor. The former problem requires a segregated population modelling framework and appears more challenging.
The main bottleneck in bioprocess closed-loop control applications is still the availability of accurate, robust, and, very importantly, integrated measurement systems of the process state. Unfortunately, this factor is quite critical for mixed microbial cultures due to the formation of cellular aggregates, which hinder the application of in situ spectroscopic techniques, such as UV/vis spectroscopy, and also due to the typically highly complex medium composition especially when using waste sources. There are, however, a number of optical methods using in situ fiber-optic probes that have been applied to wastewater processes and that have an equally high potential to be applied to mixed microbial cultures for PHA production. Possible candidates are single-wavelength or multi-wavelength spectrofluorometry, Fourier transform infrared spectroscopy (FTIR) and dielectric spectroscopy.68 The ability to measure on-line active biomass and PHA content would open the possibility to evaluate the process performance in quasi realtime and to implement model-predictive control strategies.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
5. Polymer Recovery and Characterization
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
5.1. Polymer Recovery
Although large effort has been devoted to the understanding and optimization of PHA production by mixed cultures, the same cannot be said regarding the downstream process. Currently, a low-cost, highly efficient, and environmentally friendly PHA recovery process is not generally accepted or implemented.
Nevertheless, a large number of studies have been reported in the literature about PHA extraction methods for pure cultures. In relation to studies focused on the recovery of PHA produced by mixed cultures, references are omitted.
The choice of a PHA recovery method must be based on the evaluation of green strategies involving low-cost processes, which minimize or nullify polymer degradation, while achieving high extraction yield and polymer purity. The recovery methods developed for pure cultures are based on two different principles: the polymer solubility in appropriate solvents, and disruption of the cell membrane. Some of the more well-known extraction methods are either not environmentally friendly or promote polymer degradation.
Organic solvent extraction using mainly chloroform, 1,2-propylene carbonate, methylene chloride, 1,2-dichloroethane, tetrahydrofuran methyl cyanide, and ethyl cyanide have been widely reported.69–73 The chloroform extraction method allows for high purity without polymer fractionation, and it is widely used in PHA extraction for analytical purposes at the laboratory scale.
The methods of cell disruption can be divided into chemical (by using hypochlorite, acids, alkalis, or surfactants),74–76 enzymatic,77 and mechanical processes.78 All of these methods were developed with the purpose to dissolve or destroy the nonpolymeric cell materials (NPCMs). Some digestion processes are nonselective, while polymer degradation occurs as well. In order to overcome this problem, combined methods involving cell disruption and polymer solubility were developed.30,79,80 Efficiencies of recovery and purity up to 90% and 97%, respectively, have been reported.
The search for environmentally friendly and cost-effective extraction processes led recently to the development of two methods: supercritical CO2 extraction81,82 and non-PHA-selective cell mass dissolution by protons, with PHA crystallization.83 The main advantage of the supercritical CO2 extraction is the removal of lipid impurities, thus resulting in higher final purities and a maximum reported value of 89% efficiency in polymer recovery (an indication of polymer purity was not reported by the referred authors). The second method, with recovery efficiency and purity of 95 and 97%, respectively, appears to be more cost effective when compared to conventional chemical treatment methods.
The reported methods, which were developed for the extraction of PHA obtained by pure cultures, may also be applied to the recovery of PHA produced by mixed cultures. Mixed microbial systems are mostly gram-negative bacteria, as are most of the pure cultures used for PHA production. Thus, the difficulties encountered for PHA recovery from pure cultures are likely to be similar for mixed microbial systems.
5.2. Polymer Characterization
As discussed in Section 3.1, PHA composition and chain length depend on the VFA fed as carbon source. Short-chain fatty acids tend to yield short-chain-length polymers (scl-PHA) such as homopolymers of PHB or PHV, copolymers of HB and HV units, or terpolymers of HB/HV/HMV. When longer-chain fatty acids are used, polymers with a more elastomeric behavior (known as medium-chain-length polyhydroxyalkanoates; mcl-PHA) are produced.
The homopolymer PHB is the most well-known scl-PHA. It has high crystallinity, hardness, a brittle behavior, a high processing temperature that is very close to the degradation temperature (melting point, Tm, around 178 °C for the most crystalline PHB), glass transition temperatures, Tg, ranging from −5° to 5 °C, mechanical properties similar to polypropylene, and an average molecular weight as high as 3 × 106.84 The incorporation of monomeric units other than HB in PHA copolymers increases their flexibility and processability due to the decrease in crystallinity and the lowering of the transition temperatures Tg and Tm.
In the literature, existing information about physicochemical characterization of PHA obtained from mixed cultures is very scarce, although the properties of the bacterial polyesters obtained by pure cultures are widely studied. The mechanical properties of PHA from mixed cultures have not yet been determined. Table 6 summarizes some physical and chemical characteristics of PHA obtained from mixed cultures.
Table 6. Characteristics of PHA obtained using mixed cultures. Process Feeding regime Substrate HB:HV Tg Tm Cryst Reference
mol % °C °C %
EBPR, Enhanced biological phosphorus removal; AAS, aerobic activated sludge; MMA/AE, microaerophilic/aerobic; SBR, sequencing batch reactor; Acet., acetate; Prop., propionate; But., butyrate; Val., valerate; , weight-average molecular weight; , polydispersity; , number-average molecular weight; Tg, glass transition temperature; Tm, melting temperature; Cryst., degree of crystallinity; nd, not determined.
EBPR Batch Acet. 75:25 6.5 1.9 85
Prop. 28:72 6.0 2.1
But. 60:40 4.0 2.9
Acet. + Prop. + But. 55:45 5.8 2.9
AAS SBR But. :Val. (% g:g) 86
100:0 100:0 178
80:20 88:12 144
60:40 70:30 133
40:60 65:35 127
20:80 49:51 109
0:100 46:54 99
MMA/AE(with P limitation) SBR Acet. 42:58 3.23 75 12
Feast/famine SBR Acet. 100:0 3.5 1.2 −19 145 30 88
SBR (day 1) 92:8 14.0 1.8 87
SBR (1.5 months) 97:3 19.3 2.0 87
SBR (6 months) 100:0 21.8 2.3 87
SBR (12 months) 100:0 14.8 2.1 87
SBR (18 months 100:0 21.8 2.1 7.1 168 30 87
SBR (28 months) 100:0 29.9 2.3 nd 170 49 87
Pulse-fed 100:0 22.7 2.2 3.6 175 56 87
SBR Prop. 28:72 3.5 1.4 −30 99 7 87
Pulse-fed Acet. + Prop 90:10 4.0 3.3 1.6 168 34 88
Continuous 75:25 17.0 2.5 56 139 4 88
Pulse-fed 70:30 18.0 2.1 42 141 4 88
To our knowledge, the first characterization study of PHA obtained by mixed cultures was reported by Lemos et al.,85 concerning microbial polyesters obtained in an enhanced biological phosphorous removal process. The polymers obtained were copolymers of HB/HV, with a variable molar ratio of HB:HV, depending on the type and concentration of carbon source used (Table 6). The majority of these copolymers had average molecular weights close to 6 × 105, which is the value of commercially available PHB obtained from Pseudomonas strains.
The effect of the polymer composition (namely, the fraction of HV/HA produced by aerobic sludge fed with different ratios of VFA) on the melting temperature of PHA was evaluated by different authors.12,86–88 Figure 5 presents Tm versus the fraction of HV units in the copolymers HB/HV obtained by both mixed and pure cultures. A clear decrease in Tm with an increase in HV units of the copolymer produced by both mixed and pure cultures is shown.
Figure 5. Melting temperature of poly(hydroxybutyrate-co-hydroxyvalerate) obtained by mixed and pure cultures versus copolymer HV content. Dark symbols, mixed cultures; open symbols, pure culture.
Download figure to PowerPoint
From the literature, the value of for homopolymers and copolymers obtained from various pure cultures varied between 1.7 × 105 and 4.5 × 106 (reviewed by Serafim et al.).89 Overall, the polymers obtained by mixed cultures (Table 6) have average molecular weights, polydispersities, glass transition temperatures, and crystallinity within the same range as those obtained from pure cultures.88
The stability of the PHA in terms of was recently evaluated over more than 2 years in a feast and famine process.87 The polymer had a stable throughout this period, around 2 × 106, which is also in the range reported for pure cultures.
Crystallinity and transition temperatures were also measured by Serafim et al.87 and Reis et al.88 (Table 6). Although in most of the cases the results agree well with those from studies of pure cultures,88 two unexpectedly high values of Tg, 56 and 42 °C, were obtained for copolymers HB/HV with similar co-monomer composition, respectively, 25 and 30% HV units. All the polymers studied in Reis et al.88 were not submitted to an exhaustive purification process. The analysis of the size exclusion chromatograms showed for these copolymers the presence of an impurity of around 1 500. This substance was probably a lipid soluble in chloroform, which after solvent evaporation remained in the polymer. If a strong interaction between the polymer and the lipid substance exists, higher energy is needed for the first movements of the polymer side groups, increasing the Tg values. Near the melting temperature, where the energy state of the polymeric system is higher, this interaction must be negligible, otherwise Tm should also increase, and the values obtained, 139 and 141 °C, respectively, agreed with the tendency of Tm to decrease with the incorporation of HV units in the copolymer.
The thermal stability of PHA obtained from activated sludge processes was determined, using dynamic thermogravimetry (TG), by Carrasco et al.90 These authors analyzed the behavior of the decomposition reaction of two HB/HV copolymers, both with a 20 mol-% HV content, through a kinetic study. The decomposition temperatures were 266.7 and 273.3 °C. These values were comparable with those of the commercial biopolymers produced by pure cultures, PHB and a HB/HV copolymer with 10.4% HV units, whose decomposition temperatures were 267.3 and 282.0 °C, respectively. The HB/HV copolymers obtained by mixed cultures presented a high thermal stability, as needed for their processability and commercial end use.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
6. Conclusion
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
This review shows that the mixed-culture process has high potential for PHA production, but there are still many research challenges that need to be tackled. There are strong economic motivations that justify looking to the mixed-culture process with awareness. Compared to the pure cultures of native bacteria, or to genetically modified organisms, the mixed-culture process may allow the use of cheaper substrates and cheaper nonsterile equipment. These two factors may have a significant impact on the process economics, namely on the costs of raw materials, utilities, and on the related capital costs. However, in order to have a clear idea of the relative advantages and disadvantages of each process, a deep benchmarking of both technologies is necessary.
The strategy for the enrichment of heterotrophic organisms in the sludge is a critical factor for the process productivity. The feast and famine feeding regimen was shown to be an effective strategy for selecting the heterotrophic organisms with high PHA storage capacity. The results of many studies demonstrate undeniably that mixed cultures submitted to the feast and famine regimen have high PHA storage capacity, high PHA yields, and high specific PHA productivity. Moreover, cell growth and PHA storage were shown to be concurrent cellular processes, unlike in many pure cultures, which is an important factor for attaining competitive volumetric productivities.
The maximum intracellular PHB content reported so far (65% on cell dry weight)14 is comparable with that of some pure cultures, but it is lower than that obtained in recombinant E. coli (about 90% on cell dry weight). This parameter is important for the downstream economics since high intracellular polymer contents at end of the PHA accumulation step may simplify the extraction and purification steps. The development of selective, economical, and environmentally friendly downstream processes that could be applied to microbial systems with low intracellular PHA contents would be a major technological achievement.
The polymer yield (0.56 g PHB · g−1 substrate)14 and the specific productivity (0.77 g PHB · g−1 cell dry weight · h−1) obtained so far in the mixed-culture process are very promising. The specific productivity of mixed cultures is approximately tenfold that of recombinant E. coli cultures. Even so, the volumetric PHA productivity reported for mixed cultures are still lower than the volumetric productivity of most pure cultures. The reason for this fact is the apparent difficulty in reaching high biomass concentrations in the mixed-culture process. Indeed, the main challenge regarding the bioreactor operation and control is the development of culture selection strategies of fast-growing organisms that have, at the same time, a high PHA storage capacity. The maximum specific growth rate reported so far was 0.1 h-1 and the maximum cell density was 7.0 g cell dry weight · L−1. A threshold cell density, for competing with pure cultures, would be around 10–12 g cell dry weight · L−1 since the specific PHB productivity is higher in mixed cultures. This would imply working with more diluted solutions and with shorter batch cycles. Working with diluted solutions is advantageous in terms of mixing and oxygen transfer but it could add costs to the extraction and purification steps. Fortunately, activated sludge systems form cellular aggregates that settle very easily. A settling operation at the end of the reaction would thus allow preconcentration of the sludge prior to the extraction step.
Process monitoring and control are important factors for achieving high productivity, robustness, and reproducibility. The main bottleneck in bioprocess control applications is still the availability of accurate, robust, and, very importantly, integrated measurement systems of the process state. As discussed previously, spectral techniques such as spectrofluorometry, capacitance, and FTIR may have high potential for monitoring the PHA production process by mixed cultures, but these are still topics open to investigation.
Finally, in terms of polymer quality, it has been demonstrated that a wide range of variation in monomeric compositions can be manipulated by using different substrate composition. Polymer with comparable monomeric composition produced by mixed cultures have average molecular weights, polydispersities, glass transition temperatures, and crystallinities within the same range as those obtained from pure cultures.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Acknowledgements
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
This work was supported by Fundação para a Ciência e a Tecnologia (FCT) through the project POCI/BIO/55789/2004. J. Dias, M. Albuquerque, and L. Serafim acknowledge FCT for grants SFRH/BD/13714/2003, SFRH/BD/17141/2004, and SFRH/BPD/14663/2003.
1Environmental Protection Agency, “Environmental fact sheet municipal solid waste generation, recycling and disposal in the United States: facts and figures for 1998”, EPA530-F-00-024, April 2000. 2G. Braunegg, G. Lefebre, K. F. Genser, J. Biotechnol. 1998, 65, 127. CrossRef,PubMed,CAS,Web of Science® Times Cited: 1293A. Steinbuchel, H. E. Valentin, FEMS Microbiol. Lett. 1995, 128, 219. CrossRef,Web of Science® Times Cited: 4424P. C. Lemos, L. S. Serafim, M. A. M. Reis, J. Biotechnol. 2006, 122, 226. CrossRef,PubMed,CAS,Web of Science® Times Cited: 305K. Shilpi, A. K. Srivastava, Process. Biochem. 2005, 40, 607. CrossRef,CAS,Web of Science® Times Cited: 656P. Vandamme, T. Coenye, Int. J. Syst. Evol. Microbiol. 2004, 54, 2285. CrossRef,PubMed,Web of Science® Times Cited: 917H. Pereira, P.C. Lemos, M. A. M. Reis, J. P. S. G. Crespo, M. J. T. Carrondo, H. Santos, Water Res. 1996, 30, 2128. CrossRef,CAS,Web of Science® Times Cited: 838H. Satoh, W. D. Ramey, F. A. Koch, W. K. Oldham, T. Mino, T. Matsuo, Water Sci. Technol. 1996, 34, 9. CAS,Web of Science® Times Cited: 479H. Takabatake, H. Satoh, T. Mino, T. Matsuo, Water Sci. Technol. 2000, 42, 351. CAS,Web of Science® Times Cited: 2310M. Majone, P. Massanisso, R. Ramadori, Water Sci. Technol. 1998, 38, 77. CAS,Web of Science® Times Cited: 2811H. Satoh, Y. Iwamoto, T. Mino, T. Matsuo, Water Sci. Technol. 1998, 37, 579. CrossRef,CAS,Web of Science® Times Cited: 1912W. Punrattanasim, “The Utilization of Activated Sludge Polyhydroxyalkanoates for the Production of Biodegradable Plastics”, PhD Thesis, Virginia Polytechnic Institute and State University, Blacksburg, EUA, 2001. 13M. Majone, P. Masanisso, A. Carucci, K. Lindrea, V. Tandoi, Water Sci. Technol. 1996, 34, 223. CAS,Web of Science® Times Cited: 3614L. S. Serafim, P. C. Lemos, R. F. Oliveira, M. A. M. Reis, Biotechnol. Bioeng. 2004, 87, 145. Direct Link:AbstractPDF(435K)References15G. T. Daiger, P. L. Grady, Water Res. 1982, 16, 365. CrossRef,Web of Science® Times Cited: 7216A. J. Anderson, E. A. Dawes, Microbiol. Rev. 1990, 54, 450. PubMed,CAS,Web of Science® Times Cited: 129817C. Krishna, M. C. M. van Loosdrecht, Water Res. 1999, 33, 2374. CrossRef,CAS,Web of Science® Times Cited: 4718D. Dionisi, M. Majone, A. Miccheli, C. Puccetti, C. Sinisi, Biotechnol. Bioeng. 2004, 86, 842. Direct Link:AbstractPDF(352K)References19E. Grothe, M. Moo-Young, Y. Chisti, Enzyme Microb. Technol. 1999, 25, 132. CrossRef,CAS,Web of Science® Times Cited: 3520W. Page, A. Tindale, M. Chandra, E. Kwon, Microbiology 2001, 147, 483. PubMed,CAS,Web of Science® Times Cited: 1421K. S. Kim, S. Y. Lee, H. N. Chang, Biotechnol. Bioeng. 1996, 49, 495. Direct Link:AbstractPDF(819K)References22D. Dionisi, M. Majone, V. Tandoi, M. Beccari, Ind. Eng. Chem. Res. 2001, 40, 5110. CrossRef,CAS,Web of Science® Times Cited: 2423K. Crabtree, W. Boyle, E. McCoy, G. A. Rohlich, J. WPCF 1966, 38, 1968. PubMed,CAS,Web of Science® Times Cited: 5024S. Y. Lee, Tibtech. 1996, 14, 431. CAS,Web of Science® Times Cited: 17425D. Dionisi, M. Beccari, S. Di Gregorio, M. Majone, M. P. Papini, G. Vallini, J. Chem. Technol. Biotechnol. 2005, 80, 1306. Direct Link:AbstractFull Article (HTML)PDF(235K)References26D. Dionisi, M. Majone, G. Vallini, S. Di Gregorio, M. Beccari, Biotechnol. Bioeng. 2006, 93, 76. Direct Link:AbstractPDF(220K)References27L. S. Serafim, P. C. Lemos, S. Rossetti, C. Levantesi, V. Tandoi, M. A. M. Reis, Water Sci. Technol. 2006, 54, 183. PubMed,CAS,Web of Science® Times Cited: 928F. Aulenta, D. Dionisi, M. Majone, A. Parisi, R. Ramadori, V. Tandoi, Water Res. 2003, 37, 2764. CrossRef,PubMed,CAS,Web of Science® Times Cited: 929M. A. van Aalst-van Leeuven, M. A. Pot, M. C. M. van Loosdrecht, J. J. Heijnen, Biotechnol. Bioeng. 1997, 55, 773. Direct Link:AbstractPDF(204K)References30J. Choi, S. Y. Lee, Bioprocess Eng. 1997, 17, 335. CrossRef,CAS,Web of Science® Times Cited: 8431B. S. Kim, Enzyme Microb. Technol. 2000, 27, 774. CrossRef,PubMed,CAS,Web of Science® Times Cited: 3932P. A. L. Wong, H. Chua, W. Lo, H. G. Lawford, P. H. Yu, Appl. Biochem. Biotechnol. 2002, 98–100, 655. CrossRef,PubMed,Web of Science® Times Cited: 433K. Dircks, J. J. Beun, M. C. M. van Loosdrecht, J. J. Heijnen, M. Henze, Biotechnol. Bioeng. 2001, 73, 85. Direct Link:AbstractPDF(160K)References34O. Karahan, M. C. M. van Loosdrecht, D. Orhon, Biotechnol. Bioeng. 2006, 94, 43. Direct Link:AbstractPDF(207K)References35B. S. Kim, S. C. Lee, S. Y. Lee, H. N. Chang, Y. K. Chang, S. I. Woo, J. Biotechnol. 1994, 67, 113. 36T. Yamane, M. Fukunaga, Y. W. Lee, Biotechnol. Bioeng. 1996, 50, 197. Direct Link:AbstractPDF(586K)References37D. Dionisi, M. Majone, V. Papa, M. Beccari, Biotechnol. Bioeng. 2004, 85, 569. Direct Link:AbstractPDF(401K)References38Y. Doi, M. Kunioka, Y. Nakamura, K. Soga, Macromolecules 1987, 20, 2988. CrossRef,CAS,Web of Science® Times Cited: 101,ADS39M. Beccari, D. Dionisi, A. Giuliani, M. Majone, R. Ramadori, Water Sci. Technol. 2002, 45, 157. PubMed,CAS,Web of Science® Times Cited: 2640H. Chua, P. H. F. Yu, C. H. Ma, Appl. Biochem. Biotechnol. 1999, 27, 389. CrossRef41M. Beccari, M. Majone, P. Massanisso, R. Ramadori, Water Res. 1998, 32, 3403. CrossRef,CAS,Web of Science® Times Cited: 4342F. Carta, J. J. Beun, M. C. M. van Loosdrecht, J. J. Heijnen, Water Res. 2001, 35, 2693. CrossRef,PubMed,CAS,Web of Science® Times Cited: 2843D. Dionisi, G. Carucci, M. Petrangeli Papinia, C. Riccardi, M. Majone, F. Carrasco, Water Res. 2005, 39, 2076. CrossRef,PubMed,CAS,Web of Science® Times Cited: 2444D. H. Rhu., W. H. Lee, J. Y. Kim, E. Choi., Water Sci. Technol. 2003, 48, 221. PubMed,CAS,Web of Science® Times Cited: 1245A. Carucci, D. Dionisi, M. Majone, E. Rolle, P. Smurra, Water Res. 2001, 35, 3833. CrossRef,PubMed,CAS,Web of Science® Times Cited: 2746G. Sin, A. Guisasola, D. J. W. De Pauw, J. A. Baeza, J. Carrera, P. A. Vanrolleghem, Biotechnol. Bioeng. 2005, 92, 600. Direct Link:AbstractPDF(287K)References47A. S. M. Chua, H. Takabatake, H. Satoh, T. Mino, Water Res. 2003, 37, 3602. CrossRef,PubMed,CAS,Web of Science® Times Cited: 2748J. J. Beun, K. Dircks, M. C. M. van Loosdrecht, J. J. Heijnen, Water Res. 2002, 36, 1167. CrossRef,PubMed,CAS,Web of Science® Times Cited: 5049J. J. Beun, F. Paletta, M. C. M. van Loosdrecht, J. J. Heijnen, Biotechnol. Bioeng. 2000, 67, 379. Direct Link:AbstractPDF(137K)References50K. A. Third, M. Newland, R. Cord-Rowisch, Biotechnol. Bioeng. 2003, 82, 238. Direct Link:AbstractPDF(253K)References51A. M. P. Martins, J. J. Heijnen, M. C. M. van Loosdrecht, Water Res. 2003, 37, 2555. CrossRef,PubMed,CAS,Web of Science® Times Cited: 3052W. Gujer, M. Henze, T. Mino, M. C. M. van Loosdrecht, Water Sci. Technol. 1999, 39, 183. CAS,Web of Science® Times Cited: 26253J. M. L. Dias, L. S. Serafim, P. C. Lemos, M. A. M. Reis, R. Oliveira, Biotechnol. Bioeng. 2005, 92, 209. Direct Link:AbstractPDF(215K)References54M. C. M. van Loosdrecht, J. J. Heijnen, Water Sci. Technol. 2002, 45, 13. PubMed,CAS,Web of Science® Times Cited: 1855S. H. Lee, D. H. Oh, W. S. Ahn, Y. Lee, J. Choi, S. Y. Lee, Biotechnol. Bioeng. 2000, 67, 240. Direct Link:AbstractPDF(80K)References56S. Y. Lee, H. H. Wong, J. Choi, S. H. Lee, S. C. Lee, C. S. Han, Biotechnol. Bioeng. 2000, 68, 466. Direct Link:AbstractPDF(131K)References57T. Sugimoto, T. Tsuge, K. Tanaka, A. Ishizaki, Biotechnol. Bioeng. 1999, 62, 625. Direct Link:AbstractPDF(179K)References58E. Grothe, Y. Chisti, Bioprocess Eng. 2000, 22, 441. CrossRef,CAS,Web of Science® Times Cited: 1659W. S. Ahn, S. J. Park, S. Y. Lee, Appl. Environ. Microbiol. 2000, 66, 3624. CrossRef,PubMed,CAS,Web of Science® Times Cited: 3860G.-Q. Chen, W. J. Page, Biotechnol. Technol. 1997, 11, 347. CrossRef,PubMed,CAS,Web of Science® Times Cited: 3461J. Choi, S. Y. Lee, Appl. Environ. Microbiol. 1999, 65, 4363. PubMed,CAS,Web of Science® Times Cited: 3162S. J. Park, W. S. Ahn, P. R. Green, S. Y. Lee, Biomacromolecules 2001, 2, 248. CrossRef,PubMed,CAS,Web of Science® Times Cited: 1863T. Tsuge, K. Tanaka, M. Shimoda, A. Ishizaki, J. Biosci. Bioeng. 1999, 88, 404. CrossRef,PubMed,CAS,Web of Science® Times Cited: 864T. Tsuge, K. Tanaka, A. Ishizaki, J. Biosci. Bioeng. 2001, 91, 545. CrossRef,PubMed,CAS,Web of Science® Times Cited: 465G. Kobayashi, K. Tanaka, H. Itoh, T. Tsuge, K. Sonomoto, A. Ishizaki, Biotechnol. Lett. 2000, 22, 1067. CrossRef,CAS,Web of Science® Times Cited: 1066S. Pratt, Z. Yuan, D. Gapes, M. Dorigo, R. J. Zeng, J. Keller, Biotechnol. Bioeng. 2003, 81, 482. Direct Link:AbstractPDF(195K)References67S. Pratt, Z. Yuan, J. Keller, Biotechnol. Bioeng. 2004, 88, 135. Direct Link:AbstractPDF(1096K)References68M. N. Pons, S. Le Bonte, O. Potier, J. Biotechnol. 2004, 113, 211. CrossRef,PubMed,CAS,Web of Science® Times Cited: 2169J. A. Ramsay, E. Berger, R. Voyer, C. Chavarie, B. A. Ramsay, Biotechnol. Tech. 1994, 8, 589. CrossRef,CAS,Web of Science® Times Cited: 3370USA 3044 942 (1962), invs.: J. N. Baptist; Chem. Abstr. 1962, 57, 12725d 71H. Brandl, R. A. Gross, R. W. Lenz, R. C. Fuller, Adv. Biochem. Eng. Biotechnol. 1990, 41, 77. PubMed,CAS72USA 4910145 (1990), invs.: P. A. Holmes, G. B. Lim; Chem. Abstr. 1990, 103, 121704m 73USA 4140741 (1979), invs.: R. M. Lafferty, E. Heinzle; Chem. Abstr. 1979, 87, 119162r 74S. Y. Hahn, Y. K. Chang, B. S. Kim, H. N. Chang, Biotechnol. Bioeng. 1994, 44, 256. Direct Link:AbstractPDF(533K)References75J. Choi, S. Y. Lee, Biotechnol. Bioeng. 1999, 62, 546. Direct Link:AbstractPDF(223K)References76M. Kim, K. S. Cho, H. W. Ryu, E. G. Lee, Y. K. Chang, Biotechnol. Lett. 2003, 25, 55. CrossRef,PubMed,CAS,Web of Science® Times Cited: 377F. M. Kapritchkoff, A. P. Viotti, R. C. Alli, M. Zuccolo, J. C. Pradella, A. E. Maiorano, E. A. Miranda, A. Bononi, J. Biotechnol. 2006, 122, 453. CrossRef,PubMed,CAS,Web of Science® Times Cited: 778M. S. Ghatnekar, J. S. Pai, M. Ganesh, J. Chem. Technol. Biotechnol. 2002, 77, 444. Direct Link:AbstractFull Article (HTML)PDF(112K)References79J. A. Ramsay, E. Berger, R. Voyer, C. Chavarie, B. A. Ramsay, Biotechnol. Technol. 1994, 8, 589. CrossRef,CAS,Web of Science® Times Cited: 3380S. Y. Lee, J. Choi, Polym. Degrad. Stab. 1998, 59, 387. CrossRef,CAS,Web of Science® Times Cited: 2481P. Hejazi, E. Vasheghani-Farahani, Y. Yamini, Biotechnol. Progr. 2003, 19, 1519. Direct Link:AbstractFull Article (HTML)PDF(66K)References82K. Khrosravi-Darani, E. Vasheghani-Farahani, S. A. Shojaosadati, Y. Yamini, Biotechnol. Progr. 2004, 20, 1757. Direct Link:AbstractFull Article (HTML)PDF(220K)References83J. Yu, L. Chen, Biotechnol. Progr. 2006, 22, 547. Direct Link:AbstractFull Article (HTML)PDF(150K)References84K. Sudesh, H. Abe, Y. Doi, Prog. Polym. Sci., 2000, 25, 1503. CrossRef,CAS,Web of Science® Times Cited: 36485P. C. Lemos, C. Viana, E. N. Salgueiro, A. M. Ramos, J. P. S. G. Crespo, M. A. M. Reis, Enzyme Microbiol. Technol. 1998, 22, 662. CrossRef,CAS,Web of Science® Times Cited: 4186H. Chua, P. H. F. Yu, C. K. Ma, Appl. Biochem. Biotech. 1999, 77–79, 389. CrossRef,Web of Science® Times Cited: 887L. S. Serafim, P. C. Lemos, C. Torres, M. A. M. Reis, A. M. Ramos, Biores. Technol. (submitted). 88M. A. M. Reis, L. S. Serafim, P. C. Lemos, A. R. Ramos, F. R. Aguiar, M. van Loosdrecht, Bioprocess. Biosyst. Eng. 2003, 23, 377. CrossRef,CAS,Web of Science® Times Cited: 3089L. S. Serafim, P. C. Lemos, A. M. Ramos, J. P. S. G. Crespo, M. A. M. Reis, “Polyhydroxyalkanoates Production by Activated Sludge”, in: Biorelated Polymers: Sustainable Polymer Science and Technology, E.Chiellini, M. H.Mendes Gil, G.Braunegg, J.Buchert, P.Gatenholm, M.van der Zee, Eds., Kluwer Academic Plenum Publishing, Dordrecht 2001, p. 147. 90F. Carrasco, D. Dionisi, A. Martinelli, M. Majone, J. Appl. Polym. Sci., 2006, 100, 2111. Direct Link:AbstractFull Article (HTML)PDF(154K)ReferencesJump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Biographical Information
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
João Dias received his chemical engineering degree in 2003 at FCT/Universidade Nova de Lisboa. Since 2003 he has a PhD scholarship. His PhD work involves the modelling, optimization, monitoring and control of polyhydroxyalkanoates production process by mixed cultures in the Chemistry Department at the same university. In 2005 his PhD work was done at the Institute of Technical Chemistry of the University of Hanover.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Biographical Information
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Paulo Lemos received his PhD in Biological Engineering (2001) from the Universidade Nova de Lisboa, Portugal, the master in Biotechnology from the Universidade Técnica de Lisboa and the Biology degree from Universidade de Lisboa His PhD work involved interdisciplinary research in biological phosphorus removal from wastewater (kinetics, metabolism, microbiology, process optimization). From 2001 till 2005 he had a joint Pos-Doc at ITQB and IBET, working in biopolymer production from renewable resources. Presently he has a research position at the CQFB/REQUIMTE pursuing research work in the field of Environmental Biotechnology.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Biographical Information
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Luísa Serafim received her degree in Chemical Engineering (Biotechnology specialization) in 1997 at Universidade Técnica de Lisboa. In 2004 she was awarded her PhD in the Chemistry Department, Universidade Nova de Lisboa, for her work on biopolymer production by mixed cultures. Since August 2004 she has a Pos Doc position at the CQFB – REQUIMTE where she pursues her work on kinetics, control and microbiology of biopolymers production process by mixed cultures.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Biographical Information
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Cristina Oliveira holds a 5-year degree in Chemical Engineering and a MSc in Environmental Engineering at the Universidade do Porto (FEUP). She is currently finishing her Ph.D. thesis in crystallization and precipitation processes in the chemical/biochemical industries.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Biographical Information
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Marta Eiroa received the MSc degree in Chemistry (2000) and the PhD (2004) at the University of A Coruña. The objective of her Ph.D. work was the biological removal of different organic and nitrogen compounds present in wastewaters. She spent one year as postdoctoral student in the Chemistry Department of the Universidade Nova de Lisboa (2005–2006). Her research focused on the production of polyhydroxyalkanoates by mixed microbial cultures using sugar cane molasses as substrate.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Biographical Information
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Maria Albuquerque received her Biological Engineering degree in 2002 and her M.Sc. in Biotechnology in 2004 at the Universidade Técnica de Lisboa. She is a PhD scholar, conducting her thesis on Polyhydroxyalkanoates production from sugar molasses by mixed microbial cultures at the Chemistry Department, Universidade Nova de Lisboa.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Biographical Information
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Ana Ramos obtained her PhD on polymerisation reaction engineering at Universidade Nova de Lisboa in 1993, where she works as Assistant Professor. Among her main research interests is polymer characterisation, more related with the determination of average molecular weights and correspondent distributions and thermophysical properties and their correlation with polymer structure and polymerisation conditions.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Biographical Information
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Rui Oliveira holds a 5-year degree in Chemical Engineering and a PhD in Biochemical Engineering at the Universidade do Porto (FEUP) and he is currently Assistant Professor at the Chemistry Department, FCT/Universidade Nova de Lisboa. His main research interests are bioprocess modelling, monitoring and control. Over the last five years he has developed research on the field of adaptive estimation and control of bioprocesses and dynamic programming based on novel hybrid modelling methodologies.
Jump to…Top of pageAbstract1. Introduction2. Mechanisms of Aerobic Dynamic Feeding3. Process Operation4. Process Modeling and Control5. Polymer Recovery and Characterization6. ConclusionAcknowledgementsBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical InformationBiographical Information
Biographical Information
Top of page
Abstract
1. Introduction
2. Mechanisms of Aerobic Dynamic Feeding
3. Process Operation
4. Process Modeling and Control
5. Polymer Recovery and Characterization
6. Conclusion
Acknowledgements
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Biographical Information
Maria Reis obtained her PhD in Biochemical Engineering at the Universidade Nova de Lisboa (UNL), Portugal, in 1991. At present she is Associate Professor at the Chemistry Department of the FCT/UNL. Her current research areas of interest are on the water and wastewater treatment and on the bioproduction of biopolymers from renewable sources, with focus on the microbiological aspects and on process optimization.
Get PDF (299K)
我的社区
签到
赛默飞实验室产品焕新计划有奖调研！
【七月征文】不一YOUNG实验“猿”
报名开启：ICS2024第十三届光谱网络会议！
推荐收藏！盘点中药材及饮片检测解决方案
