Polyhydroxyalkanoate Production from Municipal Waste Streams

INOM TEKNIKOMRÅDET EXAMENSARBETE

TEKNISK KEMI

OCH HUVUDOMRÅDET KEMITEKNIK,

AVANCERAD NIVÅ, 30 HP STOCKHOLM SVERIGE 2020 ,

Polyhydroxyalkanoate

Production from Municipal Waste Streams

ELSA ERIKSSON

KTH

SKOLAN FÖR KEMI, BIOTEKNOLOGI OCH HÄLSA

TRITA TRITA-CBH-GRU-2020:039

www.kth.se

1

Abstract

Polyhydroxyalkanoates (PHAs) are a group of bioplastics, which are produced through microorganisms.

They are accumulated in granules inside bacteria’s cell cytoplasm and serve as an energy reserve.

Moreover, PHAs are completely biodegradable and biocompatible biopolyesters, which make them to suitable materials for medical applications and replace conventional petrochemical plastics. However, it is not economically feasible to produce PHAs yet, as it is four to nine times as expensive as to produce fossil fuel-based plastics. In order to reduce the price, it is possible to use waste streams rich in carbon and mixed cultures as microorganisms, which was applied in this thesis work.

In this study, PHAs were synthesized from a volatile fatty acid (VFA) mixture rich in hexanoic acid, which was produced by anaerobic digestion of waste streams. To be able to obtain the maximum PHA content, the experimental work was separated into a selection phase and a production phase respectively. During the selection step, enrichment of the mixed culture took place during 50 days altering feast/famine cycles. The production phase was then conducted in a fed-batch cultivation to accumulate as much PHAs as possible, while utilizing the enriched mixed culture.

The selection phase was seen as successful since the quantity of synthesized PHA increased with time.

Solely polyhydroxybutyrate (PHB) was formed during this period. The specific consumption rates for

the hexanoic acid and acetic acid were almost the same in this phase (0.10 g hexanoic acid/(g volatile

suspended soilds (VSS),h) and 0.11 g acetic acid/(g VSS,h)), which suggests that the consumption of

these majoritarian fatty acids was simultaneous. However, the determined consumption rate for

butyric acid was approximately solely half of the values for hexanoic acid and acetic acid. The highest

PHA yield obtained in the enrichment phase was 0.26 g PHB/g VFA. In the production phase, the

highest achieved PHA content was 31.4 % of VSS, which was obtained after five hours. Both PHB and

polyhydroxyvalerate (PHV) were formed in this phase, even though the quantity of accumulated PHB

dominated with its approximately 97 weight-%.

2

Sammanfattning

Polyhydroxyalkanoater (PHA:er) är en grupp bioplaster som produceras med hjälp av mikroorganismer. De ackumuleras inuti granulater som finns i bakteriers cellcytoplasma, och används som en energireserv. PHA:er är dessutom fullständigt bionedbrytbara och biokompatibla biopolyestrar, vilket gör dem till lämpliga material att applicera inom medicinska preparat och för att ersätta konventionella petrokemiska plaster. Det är däremot inte ekonomiskt fördelaktigt att producera PHA:er än så länge, då det är fyra till nio gånger dyrare att producera än i jämförelse med att producera plaster från fossila bränslen. Ett tillvägagångssätt för att reducera priset är genom att applicera kolrikt avfall som råmaterial och en blandad kultur av mikroorganismer. Det var detta som tillämpades i detta examensarbete vid PHA produktionen.

I denna studie syntetiserades PHA:er från en blandning av flyktiga fettsyror rik på hexansyra, som framställts av avfall genom anaerobisk digestion. Det experimentella arbetet delades in i två faser: en selektionsfas och en produktionsfas. Detta för att kunna erhålla högsta möjliga PHA innehåll. Den blandade kulturen av bakterier berikades under selektionsfasen genom applicering av alternerande svält/frossa cykler i 50 dagar. Produktionsfasen utfördes därefter i en så kallad ”fed-batch odling” för att ackumulera högsta möjliga kvantitet av PHA, med hjälp av den berikade kultur blandningen.

Selektionsfasen ansåg vara lyckad, då mängden ackumulerad PHA ökade med tiden. Endast polyhydroxibutyrat (PHB) producerades under berikelsefasen. De erhållna specifika konsumptionshastigheterna för hexansyra och ättiksyra var i samma storleksordning (0.10 g/(g flyktiga suspenderade ämnen, h) respektive 0.11 g/(g flyktiga suspenderade ämnen, h)), vilket tyder på att förbrukningen av dessa fettsyror skedde samtidigt. Konsumptionshastigheten för butansyra var däremot endast cirka hälften av hastigheterna för hexansyra samt ättiksyra. Det högsta PHA-utbytet beräknades till 0.26 g PHB/g flyktiga fettsyror. Det högsta PHA-innehållet som erhölls i produktionsfasen var 31.4 % av de flyktiga suspenderade ämnena, vilket uppmättes efter fem timmar.

Både PHB och polyhydroxivalerat (PHV) bildades under denna fas, även om mängden ackumulerad

PHB dominerade med 97 vikt-%.

3

Table of Content

Abstract ... 1

Sammanfattning ... 2

1 Introduction ... 5

1.1 Background ... 5

1.1.1 Strategies for a Fossil Fuel Free Future ... 5

1.1.2 What is Polyhydroxyalkanoate? ... 5

1.1.3 Application Areas of Polyhydroxyalkanoates ... 7

1.1.4 How are Polyhydroxyalkanoates Produced? ... 8

1.1.5 Volatile Fatty Acids ... 9

1.1.6 Polyhydroxyalkanoate Production by Microbial Mixed Cultures from Waste Streams ... 11

1.1.7 Metabolism for Polyhydroxyalkanoate Synthesis from Volatile Fatty Acids ... 14

1.1.8 Factors influencing the Polyhydroxyalkanoate Accumulation ... 16

1.2 Aim... 17

1.2.1 Purpose ... 17

1.2.1 Objectives ... 18

2 Methodology ... 19

2.1 Materials and Methods ... 19

2.1.1 Growth Conditions and Cultivation Medium ... 19

2.1.2 Selection Phase ... 20

2.1.3 Production Phase... 22

2.1.4 Analytical Methods ... 22

2.1.4.1 Ammonium Measurement ... 22

2.1.4.2 Total Suspended Solids and Volatile Suspended Solids ... 22

2.1.4.3 Volatile Fatty Acids Quantification ... 23

2.1.4.4 Polyhydroxyalkanoates Quantification ... 23

2.2 Calculations ... 24

3 Results ... 27

3.1 Selection Phase ... 27

3.2 Production Phase ... 29

4 Discussion ... 32

4.1 Selection Phase ... 32

4.2 Production Phase ... 35

4.3 Future Developments ... 37

5 Conclusions ... 39

4

References... 40

Appendix ... 46

Appendix 1 ... 46

5

1 Introduction

1.1 Background

1.1.1 Strategies for a Fossil Fuel Free Future

A major topic in today’s society is the release of greenhouse gases. The strive towards a decrease of primarily carbon dioxide emissions is essential for a sustainable environment. However, approximately 80 % of the globally usage of energy comes from fossil fuels such as coal, oil and natural gas (National Geographic, 2020). In order to achieve the goal in the Paris Agreement, which is keeping the global temperature increment within 2 °C above pre-industrial levels, a radical change is required (United Nations, n.d.). For this to be accomplished, further technology developments are a necessity, as well as the establishment of a circular economy.

A lot of industries are based on nonrenewable resources. For instance, over 99 % of the plastics we utilize in our everyday life are derived from fossil fuels, and only 9 % of the total consumption are recycled (CIEL, 2015; Parker, 2017). Positive environmental impacts can be obtained by replacing the plastics derived from nonrenewable resources with biodegradable bioplastics. In this study,

“Bioplastics” will be used to refer to biodegradable plastics produced from renewable resources.

Polylactic acid (PLA) and polyhydroxyalkanoates (PHA) are two examples of bioplastics. PHAs are completely microbiological produced materials (Tan et al., 2017). It is a more environmentally friendly option in comparison to synthetic plastic, as it is produced from renewable resources and can contribute to reduce greenhouse gas emissions (Serafim et al., 2008). However, the production process is not yet feasible economically, as it is four to nine times more expensive to accumulate PHAs in comparison to producing fossil based plastics (Serafim et al., 2008). Hence, in order to increase the utilization of PHAs compared to synthetic plastics, the production processes need further optimizations.

1.1.2 What is Polyhydroxyalkanoate?

Polyhydroxyalkanoates (PHAs) are a group of polyesters, which are produced through microorganisms.

They are completely biodegradable and biocompatible biopolyesters (Tan et al., 2017). Hence, some bacteria have the ability to accumulate PHA in granulates inside their cell cytoplasm under specific conditions such as carbon excess and nitrogen or phosphorus limitation (Johnson, 2009).

Microorganisms can accumulate PHA up to 90 % of the cell dry weight, and it acts as a carbon and

energy reserve material. The molecular weight can variate from 5∙10

to 2∙10

g/mol and more than

150 different structures of monomers can be synthesized (Tan et al., 2017). Moreover, PHAs can be

separated into short-chain-length (SCL) PHAs and medium-chain-length PHAs (MCL). Monomers that

constitutes of three to five carbon atoms count as SCL PHAs, while monomers containing six to

fourteen carbon atoms belong to MCL PHAs (Tan et al., 2017). It is possible to form homopolymers and

copolymers of the produced monomers, although copolymers are more commercial due to the

possibility to vary the monomer composition and therefore also change the physical properties (Chen

et. al., 2014). One of the most commonly used PHA is the homopolymer poly(3-hydroxybutyrate)

(PHB). Due to its good barrier properties and its possibility to be produced in a large quantity it may be

6 considered as a suitable material for packaging industry. However, the mechanical properties need to be improved (El-Hadi et al., 2002). Another well-known PHA is poly(3-hydroxybutyrate-co-3- hydroxyvalerate) (PHBHV), which is a copolymer of (R)-3-hydroxybutyrate and (R)-3-hydroxyvalerate (Gracida et al., 2004). PHBHV has lower crystallinity in comparison to PHB, and it possess good barrier properties. Its mechanical and thermal properties can be enhanced by integrating it with other polymers or nanoparticles. Hence, this makes it to useful candidate for the packaging industry as well (Kovalcik et al., 2015).

It has been shown that over 300 different bacteria have the ability to accumulate PHA (Chua et al., 2003). Among others, Cupriavidus necator, formerly referred to as Ralstonia eutropha, is a wild type strain, which frequently has been used in synthesis of PHB. It has the ability to grow to a cell density of 100 g/l in a time period of 72h, which makes it a suitable candidate for large scale production (Chen, 2009). Furthermore, it has been illustrated that it is possible to achieve a PHB content of 80 % of its cell dry weight during fermentation for 60 h in a one cubic meter reactor (Chen, 2009). Another wild type strain appropriate for PHA production is Alcaligenes latus. This strain has been utilized by companies such as Biomers (Germany) and Chemie Linz AG (Austria) for accumulation of PHB and PHBHV, due to its capacity to synthesize a PHB content of 50 % of the cell dry weight during a time period of only 18 h (Chen, 2009). However, this result was obtained in a small scale fermentor with a volume of solely one litre (Chen, 2009). Aeromonas hydrophilia, Pseudomonas oleovorans and Azotobacter vinelandii are other examples of wild type strains that have been investigated and can be applied for PHA production (Chua et al., 2003; Chen, 2009).

It is possible to accumulate PHAs by recombining strains as well. Escherichia coli has been successful in several cases. For example, it has been shown that pathway I in combination with pathway V (see figure 1 in chapter 1.1.4) can produce poly(3-hydroxybutyrate-co-4-hydroxybutyrate) (P(3HB4HB)) from glucose by the use of recombinant E. coli. Recombinant E. coli can form poly(4-hydroxybutyrate) (P4HB) from glucose when encoding succinic semialdehyde dehydrogenase of Clostridium kluyeri and PHB synthase of C. necator while inhibiting native succinate semialdehyde dehydrogenase genes sad and gabD to improve the carbon source availability for P4HB production (Chen et. al., 2014). E. coli has a high growth rate and can therefore obtain a high cell density required to achieve a high production of PHA. It can use diverse types of carbon substrates and it is appropriate for genetic manipulation.

These qualities make it a suitable microorganism for recombination and PHA accumulation (Chen, 2009).

Another way to obtain PHAs is through the use of microbial mixed cultures (MMC). This production process has gained greater importance the last decades as it widens the possibilities regarding the used carbon substrate and the process conditions. For example, organic wastewater streams can be applied as raw material, while activated sludge can be utilized as inoculum (Morgan-Sagastume, 2016).

The use of low or no-cost carbon substrates, like food waste, can, in turn, also lower the production costs. Apart from the food industry, industries such as the forest-, paper- and agricultural industry produce waste streams with can be applied as suitable feedstock for the PHA synthesis, as they all have a high carbon content (Jiang et al., 2012; Venkateswar Reddy and Venkata Mohan, 2012).

Furthermore, utilization of mixed cultures can reduce the energy consumption of the PHA production

process as well, since equipment sterilization is not required (Res Urbis, 2017). However, the use of

mixed cultures often leads to a low PHA content as the cell concentration is low, compared to when

7 pure cultures are used (Albuquerque et al., 2011). For instance, as pure cultures are use the PHA content often varies between 70-80 % of the cell dry weight, while reported PHA contents obtained by the use of mixed cultures range in between 21 % of the cell dry weight and 84 % of the cell dry weight (Chua et al., 2003; Reis et al., 2003; Johnson, 2009).

Hence, multiple variations of monomers and several types of PHAs can be produced depending on the conditions and strains. This, in turn, makes the PHA family a collection of polymers with a diverse range of properties and applications (Tan et al., 2017).

1.1.3 Application Areas of Polyhydroxyalkanoates

Depending of the PHA composition, the polymers can obtain several properties, as mentioned previously. This, in turn, leads to a wide range of application areas for the PHAs, making them a great substitute for the synthetic plastics, due to their biodegradability and biocompatibility. Moreover, due to the PHA polymers’ tunable properties, these types of bioplastics have a wide range of melting points, glass transition temperatures and thermo-degradation temperatures (Chen and Patel, 2011).

These factors make PHAs’ properties similar to fossil derived plastics. PHAs have also shown great gas barrier properties. This, along with their nontoxicity, make them a potential source for packaging application when it comes to goods like food (Tan et al., 2017). They can also be applied in paints, as the polymer binder, and be processed into plastic articles such as adhesives, films and fibers (van der Walle et al., 1999; The Procter & Gamble Company, 2001; Perez Zabaleta, 2019).

Furthermore, PHAs are appealing materials for biomedical application, due to their beneficial biocompatibility. Studies have shown that there are PHAs that are consistent with bone and cartilage, tissue, blood and various cell lines (Misra et al., 2006). This has been investigated in vitro as well as in vivo tests. Furthermore, PHAs are not harmful for implantations and they are biodegradable. Tepha Inc. (USA), a company which produces medical devices, uses monofilament suture of P4HB for instance (Tepha Inc., 2018). However, there are drawbacks with utilizing PHA polymers in biomedical materials.

The mechanical properties are often poor, for instance. To overcome this problematic, PHAs can be combined with bioactive ceramics or glasses to form composites. In this case, the inorganic material will fill or coat the biopolymer. As PHB is one of the most commonly used PHAs, it has been investigated for this purpose. It has been incorporated with an inorganic phase to strengthen its brittleness. Other utilized polymers are PHBHV and poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) (PHBHHx). It has been concluded that the composites are of large potential for biomedical applications, in forms of sutures, wound dressings, tissue engineering scaffolds and cardiovascular patches. However, it has also been shown that the inorganic phase affects the biopolymer’s properties in different way depending on the type of PHA (Misra et al., 2006). Other medical applications for the PHAs are as adhesives in surgical sutures, slings, artificial skin and orthopedic pins for example (Perez Zabaleta, 2019). Before these products reach the market, they have gone through meticulous quality controls.

Their manufacturing includes strict regulations, as the quality and purity of the products are of crucial

importance (Perez Zabaleta, 2019). This makes the use of mixed cultures and waste streams

inappropriate to apply, as both the use of microorganisms as well as the substrates have a major

influence on the quantity, and mixed cultures and waste streams contribute to less process control

(Chua et al., 2003; Perez Zabaleta, 2019). Hence, these factors will lead to uncertainties concerning the

quality of the products. Due to the low control of the types of bacteria in the mixed culture and utilized

8 types of substrates it becomes very difficult to meet the demands that the strict regulations require.

Another suitable application for PHAs is as biofuels, since PHA can be converted into liquid fuels through hydrolysis or thermochemical processes (Snell and Peoples, 2009). Moreover, Deng et al.

(2009) investigated the possibility of using PHB and MCL PHAs as biofuel, where they through esterification produced R-3-hydroxybutyrate methyl ester (3HBME) and MCL hydroxyalkanoate methyl ester (3HAME). It was stated that 3HBME and 3HAME had combustion heat values equivalent with ethanol, as the values were 20 KJ/g and 30 KJ/g respectively, in comparison with combustion heat value for ethanol of 27 KJ/g. In the report it was also concluded that 3HBME and 3HAME could be applied to increase the combustion heat value for ethanol. By adding 10 % of 3HBME the combustion heat value rose to 30 KJ/g. If the 10 % of 3HBME was exchanged against 3HAME, the combustion heat value increased another 5 KJ/g to 35 KJ/g. The methyl esters can therefore be used as biofuel individually or combined with ethanol. However, neither 3HBME nor 3HAME can be used to increase combustion heat values for gasoline, diesel, n-propanol or n-butanol (Zhang et al., 2009). It can also be stated that it is unlikely that 3HBME and 3HAME will be used widely in the near future, due to the expensive production processes. The heat combustion values are inferior to the heat combustion values for gasoline (45 KJ/g) and diesel (44 KJ/g) as well (World Nuclear Association, 2018).

Lastly, PHA can also be utilized as drugs (the monomers can be applied to treat diseases such as Alzheimer disease and Parkinson's), animal feed (the oligomers can be valuable food additives), and in 3D printing materials. It may be possible to apply it in so called smart materials, which is materials that respond to stimuli to change their structural and functional properties, as well (Tan et al, 2017).

1.1.4 How are Polyhydroxyalkanoates Produced?

PHAs are accumulated in granules inside microorganisms’ cell cytoplasm and serve as an energy reserve, as previously mentioned (Johnson, 2009). However, the accumulation is highly dependent on the availability of nutrients and, in order for the bacteria to store PHA, carbon must be present in excess. The transformation of carbon substrate to PHAs is, in turn, related to the access of nitrogen, phosphorus and/or oxygen. Limitation or depletion of nitrogen or phosphorus result in inhibition of cell formation and, therefore, a decreased growth rate. During these conditions the cells consume less energy, which lead to higher levels of NADH. The NADH levels affect the activation of the citric acid cycle (TCA) cycle, where an increased amount of NADH reduces its activity. As the activity is decreased, the flux of carbon in the TCA cycle is also decreased. This, in turn, favors the PHA production as the flux of carbon is increased and can be redirected through the PHA pathway (Perez Zabaleta, 2019). The availability of oxygen influence on the PHA synthesis as well. Some bacteria such as polyphosphate accumulating organisms (PAOs) and glycogen accumulating organisms (GAOs) can only grow under aerobic conditions, as oxygen acts as the final electron acceptor (Johnson, 2009). Thus, if the presence of oxygen is limited these bacteria cannot grow. However, they can still consume carbon substrate, by converting it into PHA and store it for future energy use (Johnson, 2009). Furthermore, it has been shown that there are bacteria that can synthesize PHA without being exposed to nutrient limitation.

Actinobacillus sp., recombinant Azotobacter vinelandii and Azotobacter chroococcum are a few

examples. These microorganisms have the ability to accumulate PHA while growing (Shi, Lee and Ma,

2006).

9 As it is difficult to control the composition of the polymer, research has been set on the monomer variation in order to obtain a wide range of polymers. However, some studies, as e.g. Meng et al.

managed to produce PHA while controlling the composition. By encoding the chromosomes of the bacteria P. putida and P. entomophilia, which use pathway II (figure 9, chapter 1.1.7) for PHA production, they could form homopolymers as well as random and block copolymers with specific compositions (Chen et. al, 2014). Furthermore, the authors stated that PHA synthase plays an essential role concerning the regulation of PHA monomers composition (Chen et. al, 2014).

1.1.5 Volatile Fatty Acids

As seen in chapter 1.1.7, volatile fatty acids (VFAs) can be utilized as a carbon substrate for PHA synthesis (figure 9, pathway II). VFAs can be referred to as low molecular weight organic acids, as they constitute of a carbon chain with less than seven carbon atoms. The chains can either be branched or straight and they are defined as acids due to their carboxylic group (Atasoy et al., 2018). Formic acid is the smallest VFA, constituting of only one carbon atom. As formic acid is antimicrobial active, it does not appear in wastewater treatment plants, where microbial activity can be found (Ammayappan and Jeyakodi Moses, 2009).

Acetic acid, also called ethanoic acid, consisting of only two carbon atoms in the chain, as can be seen in figure 1. It is the most commonly used VFA, with a demand predicted to approximately 18 million tons in year 2020 (Bhatia and Yang, 2017). Moreover, acetic acid is often referred to as a key building block to chemical industries, as it can be involved in production of paint, plastics and fibers for instance, or to food and beverage industry when it comes to production of flavours and preservatives (Atasoy et al., 2018). It has been shown that the microorganisms Acetobacter, Thermoanaerobacter, Acetomicrobium and Acetothermus can be used to obtain a successful fermentation process for production of acetic acid (Bhatia and Yang, 2017).

Propionic acid (figure 2) can to some extent be used to prevent growth of mould and some bacteria.

This makes this organic acid an important element in the food industry as preservative. Apart from this, propionic acid serves a major part in the manufacturing of vitamin E (Atasoy et al., 2018). It has been shown that glucose, lactose and xylose can be utilized to produce propionic acid through fermentation. Several species of Propionibacterium, where diverse strains as P. acidipropionici, P.

freudenreichii, P. shermanii and P. thoenii, can be used in the fermentation process to produce the volatile organic acid (Bhatia and Yang, 2017).

Butyric acid and isobutyric acid are volatile organic acids with four carbon atoms (figure 3 and figure 4). Butyric acid is a meaningful chemical in the biofuel industry, since it can be utilized as biodiesel. It is also of importance for the animal feeding sector, since it is anti-pathogenic (Atasoy et al., 2018).

Hence, it can be applied as antibiotic to prevent disease infections. Anaerobic bacteria such as

Clostridium butyricum, C. kluyveri, C. barkeri and C. thermobutyricum can, for example, be used in the

production process of butyric acid (Atasoy et al., 2018). Like the other VFAs, isobutyric acid is also a

versatile and platform chemical (Bhatia and Yang, 2017). However, no bacterium has yet been

discovered to synthesize significantly quantities of isobutyric acid. Thus, there is no natural pathway

for isobutyric acid production with microorganisms. It is possible to produce it through the Ehrlich

10 pathway, as isobutyricaldehyde is formed as an intermediate during the process. In this synthesis, Lactococcus lactis is applied (Zhang et al., 2011).

Valeric acid has five carbon atoms, formed as a straight line, (figure 5) (Atasoy et al., 2018). Waste streams rich in proteins are favorable as raw material for valeric acid production. This was shown in a study by Shen et al. (2017), where they utilized egg whites and tofu as feedstock. They observed a result, where the valeric acid amount was estimated to 18-25 % of the total VFA production. Solely acetic acid was produced in a larger amount (Shen et al., 2017). Isovaleric acid (figure 6) can be produced by the use of the bacteria Propionibacterium freudenreichii (Bhatia and Yang, 2017). It has been shown by Ahn et al. (1990) and by Xiong et al. (2012) respectively, that it is possible to use recombinant strains in order to produce isovaleric acid as well (Ahn and Hayashida, 1990; Xiong et al., 2012).

Caproic acid (figure 7) is often referred to hexanoic acid and can be applied in a wide range of industries. It is commonly used in the pharmaceutical industry, acts as a flavor component in food, works as an additive in animal feed and can be applied in biofuel production. Like most of the other VFAs, hexanoic acid can be obtained through anaerobic fermentation (Cavalcante et al., 2017). The methyl-branched isocaproic acid (figure 8) is not as commonly referred to in literature as the other VFAs are. However, it can be produced together with the other volatile organic acids through acidification of sugar in wastewater of beet-pulp for example (Alkaya et al., 2009).

VFAs are generally based on fossil fuels, but they can be obtained from waste streams via anaerobic digestion (Atasoy et al., 2018). Anaerobic digestion is a wastewater treatment process, where biological activity is applied in some of the steps (Wang et al., 1999). The process can be divided into four major parts: hydrolysis, acidogenesis, acetogenesis and methanogenesis. In the first step, the hydrolysis, complex organic matter is broken down into soluble molecules such as sugar, amino acids Figure 1: Chemical

structure of acetic acid (Ibis Scientific, 2020)

Figure 2: Chemical structure of propionic

acid (Sigma-Aldrich, 2020)

Figure 3: Chemical structure of butyric acid (Castillo, 2019)

Figure 4: Chemical structure of isobutyric acid (Fisher Scientific,

2020)

Figure 5: Chemical structure of valeric acid (Fisher Scientific,

2020)

Figure 7: Chemical structure of valeric acid (Fisher

Scientific, 2020)

Figure 8: Chemical structure of isovaleric acid

(OÜ Liprafarm, 2017) Figure 6: Chemical structure

of isovaleric acid (OÜ

Liprafarm, 2017)

11 and fatty acids. Lipases, cellulases and proteases are extra cellular enzymes that can be applied in the first phase. The soluble organic materials are then converted into VFAs during the acidogenesis by the use of fermentative microorganisms. This step is followed by the acetogenesis, in where the VFAs are transformed into carbon dioxide, acetate and hydrogen as hydrogen consuming and hydrogen producing acetogenic bacteria are utilized. During the last step in the anaerobic digestion, the methanogenesis, the formed acetate, hydrogen and carbon dioxide are converted into methane.

Hence, the final products constitute of methane and carbon dioxide. Hydrogen- and acetate using methanogens are responsible for final transformation (Xu et al., 2019). Hence, the final products constitute of methane and carbon dioxide. However, due to VFA’s wide range of application areas, such as in the chemical-, food- and pharmaceutical industries, VFA are considered as important carbon substrates (Atasoy et al., 2018). It can therefore be favorable to harvest the VFAs produced during the acidogenesis or to inhibit the acetogenesis and methanogenesis. Hence, if VFAs are the wanted product, it is possible to enhance the production of them during the anaerobic digestion. Some factors that will affect the quantity of produced VFA as intermediates are temperature, pH, reactor type and used carbon substrate. In this study, the applied raw material will consist of 30 % food waste and 70 % primary sludge. It has been stated that is possible to obtain approximately 700 mg chemical oxygen demand (COD)/g volatile solids (VS) when utilizing VFAs produced from co-fermentation of primary sludge from wastewater treatment plant and food waste without pH control, as Bedaso (2019) managed to produce 687 mg COD/g VS. The mixture of produced VFAs mostly consisted of hexanoic acid (50 %) follow by acetic acid (23 %) and butyric acid (20 %) (Bedaso, 2019). Furthermore, it has been shown by Döhler (2020) that an acetic acid pH-value, where hexanoic acid constituted a majority of the feedstock rather than acetic acid, was favorable for a high VFA yield. Döhler obtained a maximum VFA yield of 584 mg COD/g VS in a reactor rich in hexanoic acid with a pH-value of circa 5.4.

However, in a reactor rich in acetic acid with a pH of 10 the maximum VFA yield was solely 317 mg COD/g VS (Döhler, 2020).

1.1.6 Polyhydroxyalkanoate Production by Microbial Mixed Cultures from Waste Streams

As mentioned before, PHAs are four to nine times more expensive to produce in comparison to synthetic plastics, due to high substrate cost, high energy consumption and complex downstream processes (Serafim et al., 2008). Even though there are large scale PHA production plants, the synthesized quantity of bioplastics is not comparable with the amount of fossil fuel based plastics.

Today 1.4 % of the used bioplastics in the packaging industry is covered by PHAs (Perez Zabaleta,

2019). In order to increase the demand and global production of PHAs, the production costs need to

be reduced. One way of decreasing the costs is by using mixed bacterial culture instead of pure culture,

since less process control is required. When utilizing pure cultures, the environment needs to be

sterile, which is costly. If utilizing mixed cultures instead, the culture is less dependent on its

surrounding, which is beneficial as it can be both difficult and expensive to establish and then maintain

very specific conditions during the production process. For instance, PHA synthesis is possible to

achieve in a non-sterile environment when mixed cultures are applied (Chua et. al., 2003). Hence,

mixed cultures make the PHA production process more robust. As a mixed culture contains several

types of microorganisms, it is possible to use a wider range of carbon substrates as well, in comparison

with when a pure culture is used (Albuquerque et al., 2011). It may be difficult to apply waste streams

rich in carbon when using pure cultures for example, as it may be complex to identify all different

12 substrates in the stream, which is required in order to know which type of strain that should be used to accomplish PHA production. This would not be the case if mixed cultures would be utilized, as it is likely that the culture already contains strains with the ability to store PHA. Moreover, it has been shown that maximum PHA contents produced through the use of microbial mixed culture are in the same range as when utilizing pure culture, which also is advantageous (Albuquerque et al., 2011).

However, the volumetric productivity is less with mixed cultures in comparison with pure cultures. The productivity is dependent on the cell concentration and as the cell concentration is less in mixed cultures, the PHA synthesis efficiency is also lower. Earlier research on PHA accumulation by microbial mixed culture has stated a cell concentration of less than 10 g/l of volatile suspended solids. As pure culture is used, a cell concentration of 100 g/l can occur (Albuquerque et al., 2011). From this, it is possible to state that if it is possible to increase the cell concentration in mixed cultures, the use of mixed cultures in the PHA production process would be superior to the use of pure cultures in most aspects.

Another option for cost reduction is usage of waste streams as feedstock. PHA accumulation is dependent on carbon rich raw materials, as PHAs mostly consist of carbon. Hence, waste streams with a high carbon content, that can be fermented and stored inside cells are suitable for PHA production, which makes industries such as the food industry as well as the forest industry interesting. Both these industries produce by-products and wastes that can be applied as feedstock for PHA synthesis. As VFAs can work as precursors to PHAs, it is also favorable if the waste streams can be transferred into VFAs easily. At the moment, research has mostly been focusing on waste streams such as molasses, agricultural wastes, food processing industrial wastewater, olive mill as well as paper mill wastewater and design wastewater so far (Venkateswar Reddy and Venkata Mohan, 2012). Concerning carbon substrate taken from the food industry, it is desired that it does not interfere with nutrition for human and animals as a lot of people would consider it as morally wrong (Koller et al., 2017). Fermented and unfermented food wastes have been conducted as feedstock for PHA accumulation. In a study by Venkateswar Reddy and Venkata Mohan (2012) it was illustrated that fermented food wastes reached a higher maximum PHA content (35.2 %) in comparison with unfermented (32.6 %). It was also shown that the fermented as well as the unfermented food wastes produced both PHB and PHV, although fermented food wastes produced a higher percent of PHV than unfermented (Venkateswar Reddy and Venkata Mohan, 2012). It is also possible to convert food waste to VFAs, by anaerobic fermentation, and, as earlier mentioned, VFAs are suitable carbon substrates for PHA accumulation (Strazzera et al., 2018). If mixed cultures are applied in the production process, complex waste materials can be utilized as well. Hence, it is possible to use waste streams where the composition is undefined, which is advantages as it can be difficult to know the exact content in some waste streams (Serafim et al., 2008).

In this thesis, these two factors, mixed cultures and VFAs from waste streams, were used to reduce the production cost, since this results in a lower energy consumption, less equipment costs due to simpler processes and a decrease in equipment control without sterilization step (Serafim et al., 2008).

In order to conduct a successful PHA production process using microbial mixed cultures, the

enrichment step is essential, where microorganisms with higher capacity to accumulate PHA are

13 selected. There are two different possibilities to perform the enrichment: through altering anaerobic and aerobic conditions or by altering feast and famine conditions (Johnson, 2009).

For the enrichment phase when utilizing alternated anaerobic and aerobic cycles polyphosphate accumulating organisms (PAOs) and glycogen accumulating organisms (GAOs) are suitable for PHA synthesis (Serafim et al., 2008). PAOs and GAOs have the ability to convert the carbon substrate into PHA under anaerobic conditions. Therefore, supplying a carbon source into a reactor in absence of oxygen constitute the first phase in the cycle. The energy consumption for the PHA production is provided from degradation of internally stored glycogen/polyglucose and through hydrolysis of polyphosphate for PAOs. When the carbon source is consumed, oxygen is fed into the reactor. During the aerobic phase, the accumulated PHA is degraded to cover the energy required to refill the storages of glycogen/polyglucose and polyphosphate (Serafim et al., 2008; Johnson, 2009). Sequencing batch reactors are often applied. If using activated sludge, this process may also be referred to as the Enhanced Biological Phosphorous Removal (EBPR) in literature, given that PAOs constitute the mixed culture (Chua et. al., 2003).

Enrichment by the use of altering feast and famine cycles have been shown to be more effective in comparison with the altering anaerobic and aerobic cycles (Johnson, 2009). During this type of selection, the availability of nutrients such as carbon or nitrogen substrates are altered instead of the availability of oxygen. In the feast period, the nutrients are present in excess to allow cell growth. In contrary, during the famine, the nutrient accessibility is limited to boost the presence of PHA producers. Commonly, in the feast phase, the nutrients are fed into a reactor under aerobic conditions.

In this way, the bacteria have the possibility to grow and accumulate PHA. During the famine phase, the bacteria that had the ability to store PHA will survive due to degradation of PHA, while the other microorganisms will starve and finally die. In order to be superior to other bacteria, the microorganisms have to take up as much as possible of the limiting nutrient during feast, since this improves their conditions in the famine phase. Thus, the selection of bacteria is mostly dependent on the bacteria’s ability to consume substrate rather than growth rate. By the end of each cycle, excess of biomass is harvested. The sludge retention time, which is the microorganisms’ average time present in the activated sludge, is determined based on the amount of withdrawn biomass (Johnson, 2009). A long sludge retention time results in a small fraction of harvested biomass each cycle, while a short sludge retention time requires a higher quantity of biomass withdrawal at the end of the cycle.

Moreover, it has been shown that the sludge retention time is an important parameter in the PHA accumulation process. It affects the process performance, sludge production and oxygen requirements (Smith, 2018). In chapter 1.1.8 it is discussed how it influences on the obtained PHA content.

Sequencing batch reactors are suitable for these types of cycles in order to be able to create a continuous process (Johnson, 2009).

The enrichment process is followed by a PHA synthesis step or also called production phase, where the

harvested biomass is utilized. During the production of PHAs, the mixed culture is feed with an excess

of carbon substrate in order to maximize its PHA synthesis. To inhibit the cell growth, and therefore

favor the polymerization process instead, by redirecting the carbon source to the product, it is possible

to limit the availability of nutrients such as the nitrogen or phosphorus, as nutrients are required for

the formation of new biomass (Johnson, 2009).

14

1.1.7 Metabolism for Polyhydroxyalkanoate Synthesis from Volatile Fatty Acids

According to the literature, it is possible to biosynthesize PHAs in 14 different pathways (Chen et. al., 2014). However, there are only three natural PHA production processes (figure 9) (Chen et. al., 2014).

In the first one, referred to as pathway I, sugar is used as a carbon source. In the first step, the sugar is converted to acetyl-CoA through glycolysis, then, transformed to acetoacetyl-CoA by the action of β- ketothiolase (PhaA) and reduced to 3-hydroxybutyryl-CoA by acetoacetyl-CoA reductase (PhaB), before it enters the polymerization process to form PHB by the action of PHA polymerase (PhaC). In pathway III, the formed acetyl-CoA is converted to malonyl-CoA (instead of acetoacetic acid as in pathway I), which in later steps are transformed into 3-ketoacyl-ACP to then lastly form R-3- hydroxyacyl-CoA monomers, as in pathway II (Chen et. al., 2014).

Figure 9: An overview of the three natural pathways for PHA synthesis.

As shown in figure 9 (pathway II), VFAs can be used as carbon substrates for PHA production. They are suitable as they can be converted into precursors for PHA, such as acyl-CoA, as mentioned earlier. The metabolism for PHA synthesis from low molecular weight organic acids can be illustrated, e.g., when using a microbial culture consisting of Cupriavidus necator. At first, the VFAs enter the cell membrane.

If the feedstock content is acetic acid, one molecule of acetyl-CoA is formed for each organic acid. The

acetyl-CoA is then condensed to acetoacetyl-CoA by the action of β-ketothiolase, to later be reduced

to 3-hydroxybutyryl-CoA by acetoacetyl-CoA reductase, as mentioned in chapter 1.1.4. This reduction

is an endothermic reaction and energy is supplied through conversion of NADPH to NADP+. 3-

Hydroxybutyryl-CoA gives rise to the synthesis of PHB by the catalysis action of PhaC. If the feedstock

is changed to propionic acid instead of acetic acid the outcome will differ as propionic acid can result

15 in three different PHAs, depending on the used strain. In the β-oxidation of propionic acid, propionyl- CoA can be formed. If a condensation reaction take place between two propionyl-CoA molecules, 3- hydroxy-2-methylvaleryl-CoA will be produced, which lastly will result in the polymerization of poly-3- hydroxy-2-methylvalerate (P(3H2MV)). However, if acetyl-CoA is present, the combination of acetyl- CoA and propionyl-CoA can lead to formation of poly-3-hydroxyvalerate (PHV) or production of poly- 3-hydroxy-2-methylbutyrate (P(3H2MB)). In order for this to happen, the feedstock needs to consist of a mixture of VFAs. A mixture of acetic acid and propionic acid can also lead to formation of PHBHV for instance. The usage of butyric acid as carbon substrate can result in synthesis of PHB directly. PHV can also be produced from valeric acid directly. However, PHB and PHV from butyric acid respectively valeric acid can also be obtained through β-oxidation (Serafim et. al., 2008).

As SCL PHAs are produced, the metabolism which include β-oxidation (pathway II, figure 9) is not a necessity. However, the SCL PHAs mentioned above are only synthesized from acetic acid, propionic acid and in some cases butyric acid and valeric acid. VFAs with six carbon atoms in the chain are not involved in these pathways. For fatty acids with longer chain length to be able to produce SCL PHAs or MCL PHAs, β-oxidation (pathway II) is required. During β-oxidation, long-chain length fatty acids are broken down (Serafim et. al., 2008). Hence, it is a catabolic process, where acyl-CoA and ATP are utilized to convert the VFA into a compound, which, in turn, will be transformed into acetyl-CoA or propionyl-CoA. Each acetyl-CoA molecule consists of two carbon atoms from the VFAs. Hence, with even number of carbon atoms in the VFA, only acetyl-CoA is formed. If the carbon chain contains uneven number of carbon atoms propionyl-CoA is produced as well, since it includes three of the carbon atoms from the VFA. This process can be accomplished by the use of multiple enzymes and figure 10 shows an overview of the β-oxidation process (Lu et al., 2009; Larsson 2015).

Figure 10: An overview of the β-oxidation in bacteria, where fatty acids are degraded (Larsson,

2015).

16 Figure 11 shows a more detailed pathway of the β-oxidation. Apart from acyl-CoA, enoyl-CoA, (S)-3- hydroxyacyl-CoA and ketoacyl-CoA also act as intermediates. In this process, the enzymes 3-ketoacyl- CoA thiolase (FadA) and 3-hydroxyacyl-CoA dehydrogenase (FadB) play an essential role for the formation of acetyl-CoA (Ouyang et al., 2007; Serafim et al., 2008).

Figure 11: Two possible pathways for PHA synthesis from fatty acids; the regular β-oxidation pathway, where acetyl-CoA is formed, and an alternative pathway, where FadB and FadA are

inhibited (Ouyang et al., 2007).

However, it has been shown that the β-oxidation pathway is less efficient in comparison with pathway I (figure 9) and contributes to a low PHA synthesis (Chen and Jiang, 2017). This, in turn, makes the production process even more expensive. In order to improve the conversion of fatty acids into PHAs, including the β-oxidation pathway, it has been illustrated that inhibition of the enzymes FadA and FadB in the bacteria Pseudomonas putida or P. entomophila leads to enhanced PHA synthesis (Chen and Jiang, 2017). Instead of converting the VFAs into acetyl-CoA, most volatile organic acids were transformed to (R)-3-hydroxyacyl-CoA by removing these enzymes. As it is possible to observe in figure 9 and figure 11, (R)-3-hydroxyacyl-CoA can be applied directly in the PHA production by the use of the enzyme PHA synthase (shown as PhaC in figure 11) (Ouyang et al., 2007; Chen and Jiang, 2017).

1.1.8 Factors influencing the Polyhydroxyalkanoate Accumulation

There are many factors affecting the PHA synthesis such as the pH, temperature, and, as mentioned in

the previous sections, nitrogen-, phosphorus- and carbon substrate availability, the type of used

microorganisms and if pure or mixed cultures are utilized. The influence of temperature, sludge

retention time and nitrogen and carbon availability in the production process of PHAs from waste

streams by the use of mixed microbial cultures have been studied. Johnson (2009) concluded that a

sludge retention time over half of a day was favorable in the enrichment phase for PHA synthesis, as

it was advantageous for the PHA producing microorganisms. A sludge retention time of one day and

four days resulted in a better enrichment of the microbial mixed culture in comparison with a sludge

retention time of solely a half day (Johnson, 2009). Moreover, Johnson stated that carbon limitation

17 was effective during the selection stage, since bacteria that could accumulate PHA had the ability to survive in the famine period. It was also concluded that limiting the nitrogen availability favored high growth rates, as a result of competition for the nitrogen among the microorganisms. However, limiting the carbon availability led to higher PHB contents in comparison with limiting the nitrogen accessibility, as carbon limitation resulted in a high acetate (carbon substrate) uptake rate. Furthermore, it was shown that the temperature inside the reactors affected the cultures’ ability to store PHA significantly (Johnson, 2009). Temperatures lower than room temperature resulted in longer feast periods. This, in turn, led to cell growth rather than accumulation of PHA, as the storage capacity was determined to approximately 35 weight-%. At higher temperatures, shorter feast periods were obtained. Therefore, the PHA synthesis was superior to the growth as well. While operating at a temperature of 30 °C, Johnson obtained a PHB content of 84 weight-% (Johnson, 2009).

Chua et al. (2003) investigated the influence of sludge retention time on the PHA accumulation as well, and, unlike Johnson, Chua et al. utilized altering anaerobic/aerobic conditions during the sludge acclimatization instead of enriching under altering feast/famine cycles. Furthermore, Chua et al.

applied municipal wastewater as carbon source while Johnson’s results were obtained when acetate constituted the carbon substrate (Chua et al., 2003; Johnson, 2009). In Chua’s et al. study it is possible to observe that a sludge retention time of three days led to a higher PHA content in comparison with a retention time of ten days, as circa 31 % of sludge dry weight respectively 21 % of sludge dry weight was received. They also stated that the PHA synthesis rate decrease during the first 25 days of the acclimatization period, before the rate reached steady state and was kept constant (Chua et al., 2003).

Apart from the influence of sludge retention time, Chua et al. analyzed the pH dependence of the PHA production. They concluded that the accumulation and production rate of PHA was not dependent on the pH in the range of seven to eight, since they obtained similar results when operating a reactor at a pH of seven and another reactor at pH eight. However, as they varied the pH-value between six and nine in the same reactor the PHA synthesis was affected. At low pH-values (six to seven) the PHA content did not even reach 5 % of the sludge dry weight. If the pH was increase to eight or nine the PHA accumulation was measured to 25-32 % of the sludge dry weight. Hence, a higher pH-value was shown to be favorable for PHA production in their case (Chua et al., 2003).

1.2 Aim

The aim of this study is to contribute to the development of a sustainable PHA production process from municipal waste streams. The goal is to make the PHA synthesis more economically feasible by the use of mixed cultures and waste streams as carbon substrates, more robust by not controlling the pH or the temperature, and more efficient by producing cells with a high PHA content. Hence, this study will focus on improving PHA accumulation.

1.2.1 Purpose

This study is involved in a project called Carbon Neutral Next Generation Wastewater Treatment

Plants, which, in turn, is a collaboration between the Royal Institute of Technology and IVL Svenska

Miljöinstitutet. The purpose is to investigate and analyze the production process of PHAs using

microbial mixed culture, when utilizing a mixture of VFAs rich in caproic acid as the carbon substrate.

18 As similar experiments have been conducted while using a mixture of VFAs rich in acetic acid as feedstock, a comparison can be made. In this way, it is possible to evaluate the influence of the composition of VFAs on the PHA production. In this project carbon depletion will be combined with nitrogen depletion, which is rare since most research has focus on either carbon depletion or nitrogen depletion.

1.2.1 Objectives

The objectives of this master thesis work are:



Produce PHA from a mixture of VFAs that has a composition according to table 1 (chapter 2.2) by the use of mixed microbial cultures.



Quantification and qualification of the synthesized PHA, quantification of the VFAs, quantifications of the total suspended solids (TSS) and volatile suspended solids (VSS), determination of ammonium content of the mixed culture present in the reactors.



Evaluate these factors’ time dependence within the cycle and throughout the whole selection phase as well as during the production phase.



Make a comparison of the obtained results in study with what has been stated in previous research.

To reach the objectives with the time and resources available, this study has following limitations:



The selection phase will be performed during 50 days, and the production phase will last for 53 hours.



As the reactors have reached steady state, samples for ammonium measurement, VSS, TSS, VFA- and PHA quantification will be taken each Monday, Wednesday and Friday at specific time points in the cycle during the selection phase.



The temperature and pH will be measured in each reactor once a week during the enrichment stage.



Samples for ammonium measurement, VSS, TSS, VFA- and PHA quantification will be taken

for analysis every three to two hours on day time during the production phase.

19

2 Methodology

In this study synthesis of PHAs by microbial mixed culture was performed, using a mixture of VFAs rich in caproic acid taken originally from municipal waste streams as carbon source. The experiment was conducted at a lab set up placed in the wastewater treatment plant Henriksdals Reningsverk. The synthesis process was separated into two major phases: the selection phase and the production phase.

The selection phase corresponded to the enrichment process mentioned in chapter 2.1.2. As it has been illustrated that altering feast and famine cycle are favorable for achieving maximum PHA contents, this method was applied in this study. The enrichment phase was carried out in three equivalent sequencing batch reactors for a time period of 50 days, where each feast/famine cycle was extended over 12 hours. All phases mentioned in chapter 2.1.2 was involved in each cycle. During the enrichment process, samples were taken regularly in order to evaluate the ammonium consumption, the VSS and the TSS concentrations, the PHA content and composition and the VFA composition and consumption. After the enrichment phase, the production phase was performed in the same reactors.

This phase was performed during 53 hours under fed-batch mode. During this period of time, the carbon substrates, hence the mixture of VFAs, was present in excess until maximum PHA contents had been reached, while limiting the nutrients required for cell growth (nitrogen). To be able to analyze this phase, samples were taken multiple times each day for evaluation of the same factors as in the selection phase.

2.1 Materials and Methods

2.1.1 Growth Conditions and Cultivation Medium

The lab set up at Hammarby Sjöstadsverk consist of three parallel reactors. During this study each reactor had three inlet streams: salt medium, VFA solution and air. The salt medium contained: 5.4 g L

K

HPO

(VWR International, Leuven, Belgium, CAS no: 7758-11-4), 2.4 g L

KH

PO

(VWR International, CAS no: 7778-77-0), 10 mg L

thiourea (Sigma-Aldrich, St Louis, MO, CAS no: 62-56-6) to inhibit nitrification, 2 ml L

trace elements (1000x), 2 ml L

MgSO

(1M) (VWR International, CAS no:

10034-99-8) and a variable amount of NH

Cl (Alfa Aesar, Kandel, Germany, CAS no: 12125-02-9), depending on the ammonium content in the VFA effluent in order to have an initial NH

Cl concentration of 5 mM. The trace elements solution consisted in: 0.8 g L

CaCl

·2H

O (Merck, Darmstadt, Germany, CAS no: 10035-04-8), 9.7 g L

FeCl

·6H

O (Merck, CAS no: 10025-77-1), 0.4 g L

ZnSO

·7H

O (Merck, CAS no: 7446-20-0), 0.4 g L

MnSO

·H

O (Merck, CAS no: 10034-96-5), and 10 g L

Na

-EDTA (Merck).

The VFA solution was received from an anaerobic fermentation process, where food waste (30 % v/v) combined with activated sludge (70 % v/v) were used as raw material. Since the VFA solution was obtained in batches, the composition varied. Hence, in order to determine the exact composition of the VFAs, samples from each batch were analyzed by a gas chromatography (GC) (chapter 2.1.4.3 for method), before to used, in either, the enrichment or the production phase. To make the inlet stream of VFA solution consistent and equal throughout the experiment, the solution was adjusted based on the results from the GC. The wanted composition of the VFAs is presented in table 1.

20 Table 1: The concentration (in %) of each compound in the adjusted VFA solution.

Volatile fatty Acid Concentration [%]

Acetic acid 31.0

Butyric acid & Isobutyric acid 19.0

Valeric acid & Isovaleric acid 6.5

Hexanoic acid & n-Heptanoic acid 43.5

Total concentration of adjusted solution: 29.5 g/l

Apart from the acids shown in table 1, there were also small amounts of propionic acid and isocaproic acid in some of the batches. Sodium acetate (CAS no: 127-09-3), sodium butyrate (CAS no: 156-54-7), isovaleric acid (CAS no: 503-74-2) and hexanoic acid (CAS no: 142-62-1) were used for the adjustment.

As the desired concentration of VFAs inside the reactors at the start of each cycle was 2 g L

, 135 ml of the VFA solution was added every 12 hours. The ammonium content was also determined in each batch by the use of ammonium kit (method, chapter 2.1.4.1) to be able to determine the amount of NH

Cl that should be added to the salt medium. The ammonium content in the VFA solutions were approximately 0.52 ± 0.11 g/l. After adjusting the VFA solution, the pH was measured and regulated with a solution of NaOH (CAS no: 1310-73-2) to a pH-value of around 7 (6.8 ± 0.16). Both salt medium and VFA solution were prepared continuously throughout the entire study. Agitation and air flow were supplied on demand to maintain dissolved oxygen above 20%. The temperature and pH in the reactors were under uncontrolled conditions.

Before beginning the selection phase, timers and pumps for all the in- and outlets were programmed according to the cycle described in chapter 2.1.2. Each cycle started at 10:13 am or pm and ran for 12 hours, passing all the different phases, before it entered the next cycle.

2.1.2 Selection Phase

The department of Chemical Engineering at the Royal Institute of Technology had a lab set up for PHA

synthesis placed in the wastewater treatment plant Hammarby Sjöstadsverk, together with IVL

Svenska Miljöinstitutet, as mentioned in the previous chapter. The set up consisted of three

sequencing batch reactors, where each reactor had equally volume of 5 L and a working volume of 2

L. The system was designed with a 12-hours cycle, when it ran during the enrichment process. Each

cycle was divided into seven phases: feast, nitrogen depletion, carbon and nitrogen depletion, biomass

withdrawal, settling, effluent withdrawal and filling. Each cycle started with a feast period (figure 12,

phase 1), where the reactors were exposed to agitation and aeration to have aerobic conditions. The

cycle then eventually entered the nitrogen depletion phase (figure 12, phase 2) when all the NH

had

been consumed, to allow PHA accumulation. In this step the reactors were still exposed to oxygen and

mixing. As the carbon substrates were consumed, the phase switched to the carbon and nitrogen

depletion stage (figure 12, phase 3). During carbon depletion, the accumulated PHA was consumed by

the bacteria, for its survival. After the carbon and nitrogen depletion phase, the biomass was

withdrawal (figure 12, phase 4). In this phase approximately half a liter of the biomass was taken from

21 each reactor during a time period of four minutes. No oxygen addition occurred in order to facilitate the effluent uptake. The agitation and aeration were turned off in the settling phase (figure 12, phase 5), and the conditions remained the same during the effluent withdrawal step. The settling lasted for 30 minutes. In this time period the content inside the reactors formed layers. During the effluent withdrawal (figure 12, phase 6), the volume inside the reactors decreased with circa another half-liter, in a period of time of five minutes. In the last phase, the filling (figure 12, phase 7), approximately one liter of nutrition and carbon substrates were added to the reactors during 13 minutes. The process was continuous, hence the last phase in the cycle was directly followed by the first step in next cycle. In figure 12 the cycle is illustrated. However, the time periods for the feast, nitrogen depletion respectively nitrogen and carbon depletion are not shown as the specific time points where the phases change were unknown.

Figure 12: The cycle during the selection phase of the set up at Henriksdals Reningsverk. The conditions in each step are illustrated with colored horizontal lines.

All three reactors were inoculated with activated sludge taken from the aerobic tank placed at Hammarby Sjöstad’s wastewater treatment plant. The reactors ran for seven days until they had reached steady state and during this period, samples were taken at 9:20 am, 9:54 am, 10:15 am and 2:00 pm every Monday, Wednesday and Friday to confirm the steady state. The samples were taken for ammonium-, VSS- and TSS measurement and VFA quantification in order to evaluate the nitrogen- and VFA consumption but also the VSS/TSS production to see if they were similar each day, which indicated steady state, or not.

When the reactor system had reached steady state, samples were taken at 10:15 am and 3:00 pm

every Monday, Wednesday and Friday. The samples from 10:15 am were used for process control,

ammonium measurement, VFA quantification and occasionally VSS and TSS respectively, while the

samples taken at 3:00 pm were analyzed for ammonium content, VFA and PHA quantification and lastly

VSS and TSS (duplicates). Furthermore, throughout the selection phase, the pH was measured in the

22 reactors, the salt medium and the VFA solution once a week. The oxygen level and temperature in each reactor was checked each week as well. Apart from this, the inflows and outflow were measured every now and then, to ensure the maintenance of steady state. At the end of the enrichment phase the nitrogen consumption was measured by analyzing the ammonium content at several time points during the same day.

2.1.3 Production Phase

During the production phase, the same reactors as in the enrichment step were utilized under fed batch mode. Hence, no outflow took place in this phase. Instead of utilizing 12 hours cycles, the production phase was performed during a time period of 53 hours, starting with a short batch phase of 5 hours and a feed phase of 48 hours. During the feed phase, the nitrogen availability was limited, while the carbon substrate was present in excess, in order to maximize the PHA accumulation.

Therefore, the inflow solely consisted of 60 ml of VFA solution each hour. The fed batch reactors were exposed to agitation and oxygen throughout the whole production phase.

The first day of the fed batch experiment, samples for ammonium measurement, VFA and PHA quantification as well as VSS and TSS were taken at the start (10:00am), 12:00 pm and 3:00 pm. The second and third day the same types of samples were taken every second hour at daytime, starting with the first sampling at 9:00 am each day. Hence, the last samples were taken at 3:00 pm the third day. This experiment was repeated twice for two of the reactors. The second time solely 20 ml of VFA solution was added to each reactor per hour.

2.1.4 Analytical Methods

2.1.4.1 Ammonium Measurement

The ammonium levels were measured by the use of an ammonium kit (Ammonium test Sulpeco, Cat N° 1.14559.0001, range 4.0-80.0 mg/l NH4 - N, 5.2-103.0 mg/l NH4), following the instructions of the maufacturer. At first, the samples from the reactors were filtered through 0.45 μm cellulose filters (VWR sterile syringe filters of acetate membrane) and then through 0.20 μm filters (VWR syringe filters of polypropylene membrane). If the samples had been centrifuged, the samples were not filtered through the 0.45 μm filters. The spectrophotometer (photoLab 6600 UV-VIS) was used to measure the ammonium content. If the samples were too concentrated, the filtered samples were diluted with distilled water.

2.1.4.2 Total Suspended Solids and Volatile Suspended Solids

Filters (CAT No. 1823-055) were prepared at least a day before usage. In the preparation process the filters were put in an oven at 105 °C over the night. The next day the filters were placed in another oven at 550 °C for 30 minutes (60 minutes if the oven was cold initially). Before usage, the filters were cooled at room temperature and weighed. As 5 ml of samples was required for VSS and TSS analysis, a little more than 5 ml (10 ml if duplicate) of the microbial culture was withdrawn from each bioreactor.

The VSS/TSS were then measured according to standard protocols.

23 2.1.4.3 Volatile Fatty Acids Quantification

The samples taken from the three bioreactors were either filtered twice, through 0.45 μm cellulose filters (VWR sterile syringe filters of acetate membrane) and through 0.20 μm filters (VWR syringe filters of polypropylene membrane), like for the ammonium measurement, or centrifuged for 11 minutes at 9700 rpm (Centrifuge 5804 from Eppendorf) and then filtered through the 0.20 μm filters.

500 μl of each filtered sample was pipetted and added to a glass vial along with 100 μl of phosphoric acid (H

PO

, 25 % v/v). The glass vial was then covered with a lid and stored in a freezer until a GC analysis took place. The samples were analyzed in the GC Agilent Intuvo 9000, using distilled water as blank sample, helium as mobile phase for separation of the VFAs, a solution of volatile free acid mix 10 mM (Sigma-Aldrich) as standard and a DB-WAX IU 30m, 0.25mm, 0.25um Intuvo column (part Np:122-7032UI-INT). The injection volume was set to 0.6 µl with a split ratio of 1:15, the carrier gas (helium) to 3 ml/min and the injection temperature to 250 °C. The temperature gradient for the separation was programmed according to table 2.

Table 2: GC program for quantification of VFAs.

Rate [°C/min] Value [°C] Hold time [min] Run time [min]

Initial 70 0.5 0.5

Ramp 10 200 16.5

Lastly, a flame ionized detector (FID) was utilized at a temperature of 280 °C for detection.

2.1.4.4 Polyhydroxyalkanoates Quantification

A sample of 40 ml of mixed culture was taken from each reactor and added to falcon tubes, along with 1.2 ml of formaldehyde (36% v/v). The ratio of formaldehyde and volume sample should be circa 1 % v/v to accomplish inhibition of biomass activity. The falcon tubes were then centrifuged for 11 minutes at 9700 rpm (Centrifuge 5804, Eppendorf). Thereafter, the supernatant was eliminated, except from approximately 5 ml for ammonium measurement and VFA quantification, before putting the falcon tubes in an oven at 105 °C for circa two days without the lids on to dry the samples. As the tubes were removed from the oven, the weight of the samples were constant. The lids were put on the falcon tubes to keep them dry and the samples were stored at room temperature until the digestion process.

Before GC quantification, the dried cells were digested to extract the PHA. In the digestion process, the dried cells were crushed to obtain smaller particles. Approximately 10 mg of each sample was weighed and put into a capped culture tube. 2 ml of chloroform (CAS no: 67-66-3) and 2 ml of a mixture consisted of 30 ml H

SO

(CAS no: 7664-93-9), 970 ml methanol (CAS no: 67-56-1) and 100 mg benzoic acid (CAS no: 65-85-0) as internal standard were added to the tubes. The tubes were then covered with thread seal tape of polytetrafluoroethylene, closed with lids and vortexed for a few seconds to assure that all cells in the tubes were in the liquid before they were incubated for two hours at 100 °C in a thermo-heat block (Merck, Spectroquant TR 320, or Hach, Lange LT 200). To avoid formation of large cell fragments during the incubation, the tubes were vortexed for 10 seconds every 30 minutes.

After the incubation, and after being cooled down to room temperature, 1 ml of deionized water was