Bacterial species-structure-property relationships of polyhydroxyalkanoate biopolymers produced on simple sugars for thin film applications

Abstract

Background

The bacterial production of polyhydroxyalkanoates (PHAs), a class of non-toxic, biodegradable, and bio-based polymers, has gained increasing attention as a sustainable alternative to petrochemical plastics. Among PHA producers, Cupriavidus necator H16 and Pseudomonas putida KT2440 are used for their ability to synthesise short- and medium-chain-length PHAs, respectively. While PHAs have been produced from simple hexoses like glucose and fructose, there remains a lack of systematic and integrated analysis linking carbon source, strain selection, monomer composition, and polymer crystallinity to blend behavior in ultrathin films.

Results

PHB and mcl-PHA production using Cupriavidus necator H16 and Pseudomonas putida KT2440 on glucose and fructose were compared herein. C. necator accumulated PHB up to 60 wt% on fructose and 45 wt% on glucose, with high molecular weight (0.7–1.3 MDa), while P. putida produced mcl-PHA up to 22 wt% on fructose and 18 wt% on glucose, with lower molecular weight (46–47 kDa) and a C6 – C12 monomer profile. Notably, C. necator exhibited extreme cell elongation (up to 30 μm) during PHB accumulation on fructose. Extracted polymers were systematically solvent-blended at defined ratios (100:0, 80:20, 60:40, 40:60, and 20:80 PHB:mcl-PHA) and cast into ultrathin films (~ 20 μm) with varying composition. Crystallinity was modelled using a Gaussian fitting approach on FTIR spectra via custom MATLAB code, enabling localised phase analysis and offering a rapid alternative to DSC for thin film crystallinity estimation. While film blends exhibited tunable crystallinity and multiple melting transitions, elongation at break was consistent across compositions, with increases observed at higher mcl-PHA content.

Conclusions

This study provides a systematic comparison of PHAs from C. necator H16 and P. putida KT2440 grown on common hexoses, with full characterisation of monomer composition, molecular weight, and thermal behaviour to guide thin film bioplastic design. Blending PHB and mcl-PHA in ultrathin films revealed reduced melting points and crystallinity, likely due to reduced crystal size from film thickness constraints. This work offers a comparative reference for microbial PHA production and presents a strategy to design bioplastics with tunable properties for temperature-responsive packaging and drug delivery applications.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12934-025-02833-7.

Background

Plastic waste stands as one of the key environmental challenges of the 21st century, with the prevalent use of synthetic petroleum-derived plastics leading to an alarming accumulation of non-biodegradable materials in ecosystems worldwide [1]. One promising avenue is the utilisation of bioplastics, with polyhydroxyalkanoates (PHAs) emerging as a bio-based, non-toxic, marine biodegradable, and home compostable bioplastic in the fight against plastic pollution [2, 3]. These biopolymers offer a solution to the problems posed by conventional plastics as a material replacement, and their production from renewable resources has garnered attention in recent years since poly(3-hydroxybutyrate) (PHB) was produced commercially by Biomer and Zeneca in the early 1990’s [4].

PHAs are a family of biopolymers synthesised by over 300 microorganisms, including bacteria, archaea, and fungi, under conditions of nutrient imbalance [5]. In nature, PHAs are accumulated inside microorganisms as insoluble particles for reserve energy in the cytoplasm, and do not influence the cellular osmotic pressure even when accumulated in high concentrations [6]. Under deficiency of micronutrients such as nitrogen, phosphorus, sulphur, or oxygen, and in the presence of excess carbon, PHAs with high molecular weight (20,000 to 3,000,000 Da, 0.2–0.5 μm granule size and 5–13 particles per cell) are formed by the intracellular enzymatic system through conversion of the carbon source [6, 7]. Material properties of PHA biopolymers are dependent on the type and distribution of the monomeric building blocks and their sidechain length (hydroxyalkanoates (HA)) as observed in Fig. 1. Major PHAs include the short-chain-length homopolyester PHB with a methyl group sidechain, the copolyesters poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV), poly(3-hydroxybutyrate-co-4-hydroxybutyrate) (P(HB-co-4HB)), poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) (P(HB-co-3HHx)) and medium-chain-length PHA (mcl-PHA) copolyesters containing C5–C13 monomers [4].

Fig. 1.

General polyhydroxyalkanoate structure

These biopolymers can be utilised as alternatives to conventional plastics due to their comparable properties to fossil-based polymers, with PHB exhibiting characteristics similar to those of polypropylene, while mcl-PHA’s demonstrate properties akin to elastomers or flexible plastics. Blends of PHAs have been previously investigated to improve the properties of PHB for film applications, including with other biodegradable polyesters such as medium-chain-length polyhydroxyalkanoates, poly(ε-caprolactone), poly(lactic acid) or polysaccharides and their derivatives [8]. PHB and PHBV blends are miscible, with PHB/PHBV(6% HV) cast film blends exhibiting only one glass transition temperature (T_g) and a linear dependence of T_g and melting temperature (T_m) on the concentration of PHBV in the blend, however with > 15% HV units in the PHBV causing immiscibility [9]. PHB and pol(3-hydroxyoctanoate) (PHO) blends were also investigated, finding that increasing PHO content increased elongation at break in the blend, whilst decreasing the modulus of elasticity (Young’s modulus) — a measure of stiffness that indicates how resistant a material is to elastic deformation [10]. Cross-linking in a 30% PHO/70% PHB blend improved the elongation at break and Young’s modulus [11]. However, investigations have mainly focused on film thickness > 100 μm, which are not relevant for commercial applications as a replacement for fossil based plastics, where a majority of films used for packaging are between 10 and 50 μm.

Cupriavidus necator H16 and Pseudomonas putida KT2440 are gram-negative soil-dwelling bacteria that stand out as robust candidates for PHA production. Due to their efficient PHA synthesis capabilities, well-characterised genetic and metabolic pathways, ability to grow in a variety of carbon sources, and environmental stress tolerances and scalability, they are ideal candidates for industrial PHA production [5, 12]. Cupriavidus necator H16 (formerly Wautersia eutropha, Ralstonia eutropha and Alcaligenes eutropha) is considered the most productive organism for PHA production, due to its ability to accumulate PHA at up to 71% PHB of the cell dry weight on fructose in a shake flask [13, 14], over 90% PHB on waste rapeseed frying oil in a batch fermenter [15], and more recently over 50% PHB in a fed batch reactor on fructose, making it an attractive candidate for bioplastic manufacturing [16]. PHB exhibits a high degree of crystallinity (30–90%), high melting temperature (T_m = 175–180 °C) and close degradation temperature (~ 220 °C), leading to a material that has high brittleness but a poor elongation at break of 2–8% [2, 17]. Pseudomonas Putida KT2440 is able to synthesise various types of mcl-PHAs with longer side chains, with a lower degree of crystallinity (20–40%), lower melting temperature (T_m = 30 –80 °C) and glass transition temperature (-25 to -49 °C) but similar degradation temperature (~ 160–180 °C), leading to a material that is more elastic and amorphous, with a higher elongation at break (300–500%) [2, 18].

To produce these PHA’s cheaply, a low cost substrate is required, and this is where waste sugars offer a key alternative. Glucose and fructose are simple six carbon sugars that are abundant and can be derived from renewable sources including food and agricultural waste by breaking down starches. Both bacterial species follow distinct biochemical pathways to convert sugar substrates into PHAs [19, 20]. Cupriavidus necator H16 primarily employs the Entner-Doudoroff (ED) and butanoate pathways for sugar metabolism [21, 22]. Cupriavidus necator H16 can only use fructose as a feedstock and cannot assimilate glucose [23]. However, a glucose-positive phenotype can be isolated after 70 h when grown on media with high glucose concentration [23]. P. putida KT2440 also primarily utilises the ED pathway for sugar metabolism to incorporate sugars into various PHA monomers; however, it accumulates medium-chain-length polymers due to its different PHA synthases [24, 25].

The growing interest in biodegradable polymers has driven research into polyhydroxyalkanoates due to their sustainable production and potential to replace conventional plastics. While the metabolic pathways underlying PHA production from hexoses in C. necator and P. putida have been previously studied, there remains limited research that systematically integrates polymer yield, monomer composition, pH behavior during cultivation, and extracted polymer thermal and chemical properties, particularly in the context of blending these PHAs into ultrathin films. In this study, we investigate the production of PHB from Cupriavidus necator H16 and mcl-PHA from Pseudomonas putida KT2440 using fructose and glucose, thoroughly characterise the extracted polymers and blend them to create thin films with tunable thermal characteristics. By examining the T_m and crystallinity, we explored how the polymer blend ratio influences thermal properties. To support this, crystallinity was also modelled from FTIR spectra using custom MATLAB code, enabling quantitative analysis of polymer phase structure based on spectral features in the carbonyl region (1720–1740 cm⁻¹).

Methods

Media preparation

Nutrient Broth (NB) and Miller Lysogeny Broth (LB) were prepared for initial bacteria growth. NB was prepared with per litre; 3 g beef extract (Westlab, BL062) and 5 g peptone (Westlab, PL097). LB was prepared with per litre; 10 g tryptone (Sigma-Aldrich, 16922), 5 g yeast extract (Sigma-Aldrich, 70161) and 10 g NaCl (Sigma-Aldrich, S7653).

For the preparation of one litre of minimal salt media (MSM) for Cupriavidus necator H16 glucose adaptation medium (C: N:P ratio), chemicals utilised include 20 g glucose (Sigma-Aldrich, G7021), 2.3 g KH₂PO4 (Sigma-Aldrich, P5655), 2.9 g Na₂HPO₄.2H₂O (Sigma-Aldrich, 567550), 1 g NH₄Cl (Sigma-Aldrich, A9434), 0.5 g MgSO₄.7H₂O (Sigma-Aldrich, M2773), 0.5 g NaHCO₃ (Sigma-Aldrich, S5761), 0.01 g CaCl₂.2H₂O (Sigma-Aldrich, C3306), 0.05 FeSO₄.7H₂O (Sigma-Aldrich, F8633) and 5 mL of trace element solution. The trace element solution contained per litre: 0.1 g ZnSO₄.7H₂O (Sigma-Aldrich, Z0251), 0.03 g MnCl₂.4H₂O (Sigma-Aldrich, M5005), 0.3 g H₃BO₃ (Sigma-Aldrich, B6768), 0.2 g CoCl₂.6H₂O (Sigma-Aldrich, C8661), 0.01 g CuCl₂.2H₂O (Sigma-Aldrich, C3279), 0.02 g NiCl₂.6H₂O (Sigma-Aldrich, N6136) and 0.03 g Na₂MoO₄.2H₂O (Sigma-Aldrich, M1651).

For the preparation of one litre of MSM for Cupriavidus necator H16 PHB accumulation, chemicals utilised included 10 g glucose (Sigma-Aldrich, G7021) or fructose (Sigma-Aldrich, F3510), 1 g (NH₄)₂SO₄ (Sigma-Aldrich, A4418), 1.5 g KH₂PO₄ (Sigma-Aldrich, P5655), 4.47 g Na₂HPO₄.2H₂O (Sigma-Aldrich, 567550), 0.2 g MgSO₄.7H₂O (Sigma-Aldrich, M2773), 0.002 g CaCl₂.2H₂O (Sigma-Aldrich, C3306) and 1 mL trace element solution. The trace element solution contained per litre: 10 g FeSO₄.7H₂O (Sigma-Aldrich, F8633), 0.351 g MnSO₄.H₂O (Sigma-Aldrich, M7899), 1 g CuSO₄.5H₂O (Sigma-Aldrich, C8027), 2.25 g ZnSO₄.7H₂O (Sigma-Aldrich, Z0251), 0.23 g Na₂B₄O₇.10H₂O (Sigma-Aldrich, S9640), 0.1 g (NH₄)₆Mo₂O₂₄ (Sigma-Aldrich, M1019), 10 mL 37% HCl (Sigma-Aldrich, 258148) and made up to 1 L with deionised water.

For the preparation of one litre of MSM for Pseudomonas putida KT2440 mcl-PHA accumulation, chemicals utilised included 10 g glucose (Sigma-Aldrich, G7021) or fructose (Sigma-Aldrich, F3510), 5.61 g (NH₄)₂SO₄ (Sigma-Aldrich, A4418), 3 g KH₂PO₄ (Sigma-Aldrich, P5655), 0.8 g EDTA (Sigma-Aldrich, E6758), 1 g MgSO₄.7H₂O (Sigma-Aldrich, M2773), 0.03 g CaCl₂.2H₂O (Sigma-Aldrich, C3306), 0.32 g FeSO₄.7H₂O (Sigma-Aldrich, F8633) and 0.5 mL trace element solution. The trace element solution contained per litre: 40 g MnCl₂.4H₂O (Sigma-Aldrich, M5005), 4 g CoCl₂.6H₂O (Sigma-Aldrich, C8661), 40 g CuCl₂.2H₂O (Sigma-Aldrich, C3279), 20 g ZnSO₄.7H₂O (Sigma-Aldrich, Z0251), 20 g Na₂MoO₄.2H₂O (Sigma-Aldrich, M1651) dissolved in 1 L of 1 M HCl (Sigma-Aldrich, 258148).

Bacterial strains

Cupriavidus necator H16 (DSM 428) and Pseudomonas putida KT2440 were purchased from ATCC (Virginia, USA). Bacteria were initially inoculated in 50 mL falcon tubes, using NB for Cupriavidus necator H16 and LB for Pseudomonas putida KT2440, and placed into an incubator shaker at 200 rev. min^− 1 and 26 °C and 37 °C, respectively, until stationary phase was reached. After 24 h, the bacterial cultures were put into glycerol stock (50:50 mix of 35% glycerol stock solution and bacterial culture) in a -80 °C freezer for long term storage.

Phenotype isolation

Cupriavidus necator H16 glucose phenotype (GP) was isolated using the glucose adaptation medium. Shake-flask cultivations were carried out in triplicate in 250 mL flasks, containing 100 mL of the medium at 30 °C, in an incubator shaker system (Ratek Instruments Pty. Ltd., Australia) at 200 rev. min^− 1. Inoculum was then added at 10% v/v and optical density (OD) was measured every 6 h until the bacteria started to grow spontaneously after passing a lag period of 137 h (Figure S1). The longer adaptation time was likely due to the lower air-to-medium ratio used in this experiment (1.5:1) compared to literature (4:1) [23]. The bacteria were removed from the vessel at 260 h and put into glycerol stock in a -80 °C freezer for long term storage.

Shake flask experiments

Cupriavidus necator H16 GP and Pseudomonas putida KT2440 PHA accumulation utilised separate minimal salt media due to differing metabolic preferences and regulatory mechanisms [26, 27], with the cultivation plan to enhance the weight of cells first by supplying sufficient nutrient sources, and then limiting the nutrients excluding carbon substrate, with the carbon source selected being either glucose or fructose. Sterile conditions were maintained throughout experiments and all the stock solutions used were either sterile filtered using 0.2 μm syringe filters (Thermo Fisher, UK) in a laminar flow hood or autoclaved.

Fermentation was carried out in 2 L Erlenmeyer flasks containing 1 L of media, in an incubator shaker system (Ratek Instruments Pty. Ltd., Australia) at 200 rev. min^− 1, 30 °C and initial pH of 7. Cultures were initially reactivated in 10 mL of their respective rich medium (NB or LB) in 50 mL falcon tubes, before inoculating into 100 mL of rich medium in 250 mL flasks. When OD reached 1, inoculum was then added at 10% v/v to 1 L minimal salt media in 2 L flasks. All experiments were conducted in triplicate, with a control medium for each condition to ensure contamination safety. Optical density measurements were undertaken at 600 nm on a DiluPhotometer™ OD600 (Implen, München, Germany). OD graphs were fitted with a Boltzmann function to produce a sigmoidal curve, reflective of lag, growth and stationary phases of bacteria growth. pH was measured every 24 h using a pH meter (Mettler Toledo, Switzerland). pH graphs were fitted with an exponential decay function. The experiments were run for approximately 120 h and the cell mass was harvested at 24 and 72 h for characterisation.

Microscopy

Nile red staining was carried out as described previously [28]. A Nile red solution was prepared in a covered bottle containing 10 µg mL^− 1 Nile red in DMSO. Samples were taken after 24 h in the rich medium, as well as 24 h and 96 h in the minimal salt media. For each sample, 1 mL of culture was harvested by centrifugation (90 s, 10,000 x g) and the supernatant was discarded. The pellet was resuspended in the remaining ~ 30 µL medium that congregated from the tube walls and 4 µL of cells were then added to 1 µL of Nile red solution in an Eppendorf tube. 1 µL of the stained cell suspension was dropped onto a cleaned cover slide and let dry for a few seconds. Samples were imaged at 100x magnification (NA = 1.45) on an IX83 Inverted Microscope (Olympus, Japan) equipped with an ORCA-Flash4.0 V3 Digital CMOS camera (Hamamatsu, Japan) in red fluorescence (Nile red excitation: 562/40 nm, emission 594, CY3 filter) and differential interference contrast (DIC) imaging modes. Images were analysed with ImageJ Fiji. A representative region was selected from each condition and cell measurements were averaged over 50 bacteria. A cell length, diameter and L/D range were quoted for all conditions.

PHA extraction

Cultures were harvested by centrifugation at 3500 g for 15 min (Heraeus Multifuge 3 S-R Centrifuge, Thermo Fisher, UK). The supernatant was removed and cell pellets were washed with water and centrifuged again, before being freeze dried (Alpha 1–4 LSCbasic, Martin Christ, Germany) for 24 h and cell dry weights determined.

Dried biomass was vortex mixed with methanol at 50 mg in 1 mL for 1 min at room temperature to make the PHA more accessible and soluble for the extraction step [29]. The methanol treated biomass was then centrifuged for 10 min at 3500 rev. min^− 1 and dried at 40 °C overnight. Soxhlet extraction was undertaken with chloroform at 70 °C for 2 h at a maximum biomass: chloroform ratio of 1 g : 250 mL. The solution was then concentrated by rotary evaporation at 60 °C. For Cupriavidus necator H16 GP samples, the solution was precipitated in cold methanol and filtered with a Bucher funnel under vacuum and filter paper, before being dried at 40 °C overnight. For Pseudomonas putida KT2440, the solution was filtered with a 0.45 μm PTFE filter to remove cell debris and precipitated in cold methanol (1:10 v/v). The polymer was collected by centrifugation (3000 × g for 15 min) and dried at room temperature under vacuum.

Cast film production

90 ± 5 mg of PHA sample was dissolved in 20 mL of chloroform and heated to 40 °C for 5 min while vigorously stirring to ensure dissolution. Solutions were cast onto a 9 cm diameter petri dish and evaporated in a fume hood, covered with foil, for 8 h. Films were produced 20 ± 5 μm in thickness. Due to the viscous nature of the mcl-PHA, the final mcl-PHA film (20% PHB, 80% mcl-PHA) had to be recast into a 5 cm diameter petri dish to prevent the formation of holes in the sample, however the thickness was still 20 ± 5 μm.

Characterisation

Gas chromatography mass spectroscopy (GCMS)

Dried samples were methylated as described by Juengert et al., 2018 [28], and analysed using gas chromatography-mass spectrometry (GC System 7820 A, MSD 5977B; Agilent) equipped with a Fatwax UI capillary column (30 m × 320 μm × 0.25 μm; Agilent). The analysis conditions were set as follows: an injection volume of 1 µL, an injection temperature of 250 °C, a helium flow rate of 1.5 mL min^− 1, and a split inlet mode with a ratio of 1:25. Gradient temperature was programmed from 60 °C to 250 °C rising at 15 °C per minute for the separation of the methyl esters. The mass spectrometer was configured to detect masses within the range of 30 to 600 amu. Peak identification and quantification were achieved using the MassHunter software. The identities of the PHA monomers were established based on their retention times and mass spectra, which were compared against the NIST Mass Spectral Database (NIST MS Search 2.3). The mole fraction of each monomer unit was calculated based on the area percentage obtained from the gas chromatography analysis (Table S3). If a monomer concentration was less than 1% it was not included in the calculation.

Gel permeation chromatography (GPC)

PHA polymers were dissolved in chloroform at a concentration of 10 mg mL^− 1, which was selected to provide sufficient refractive index detector (RID) response and reproducibility for reliable molecular weight determination across all samples. To ensure complete solubilisation, PHB samples were heated gently at 40 °C. C. Necator fructose (72 h) underwent minimal sonication (bath sonicator) for 15 min to ensure complete solubilization. Gel permeation chromatography measurements were carried out on a HPLC system (1220 Infinity LC; Agilent, California, USA) using chloroform as the mobile phase and an injection volume of 20 µl. The analysis was carried out at a flow rate of 1 mL min^− 1 using a gel permeation chromatography column (PLgel 10 μm MIXED-B 300 × 7.5 mm; Agilent, California, USA) with an oven temperature of 25 °C and a refractive index (RI) detector at 35 °C. Samples were run in triplicate with a stop time of 15 min per run. Polystyrene molecular mass standards (PS-H EasiVial 2 mL; Agilent, California, USA) ranging from 162 to 6,545,000 Da were used for calibration curve and molecular weight determination, and are summarised in Table S4, Figure S3 and S4.

Fourier transform infrared (FTIR)

FTIR was undertaken on a Spectrum Two FT-IR (PerkinElmer, Inc., Massachusetts, USA), equipped with UATR and in the frequency range of 400–4000 cm^− 1. The spectra were obtained with a resolution of 4 cm^− 1, with an average of 16 scans and averaged to obtain a good signal-to-noise ratio. The data fitting method and MATLAB code are included in Supporting Information S10.

Differential scanning calorimetry (DSC)

Thermal properties were studied by differential scanning calorimetry (DSC) using a DSC 2500 (TA Instruments, USA), equipped with an autosampler. Two separate methods were used for PHB and mcl-PHA polymers. For PHB polymers from Cupriavidus necator H16 GP, an aluminium pan containing sample (5 ± 1 mg) was heated from 30 to 200 °C at a scanning rate of 10 °C min^− 1, cooled to -50 °C and reheated to 200 °C before cooling to room temperature, in a nitrogen atmosphere. Cooling to -50 °C, which is below the expected glass transition temperature (T_g) range (approximately − 10–10 °C), was chosen to ensure a stable and reproducible baseline and complete relaxation of polymer chains, eliminating residual thermal history. For PHA polymers from Pseudomonas putida KT2440, an aluminium pan containing sample (5 ± 1 mg) was heated from room temperature to 70 °C at a scanning rate of 10 °C min^− 1 and annealed at 70 °C for one hour. The sample was then cooled to 20 °C at a scanning rate of 10 °C min^− 1 and annealed at 20 °C for one hour, before cooling to -65 °C, heating to 70 °C and then cooling back to -65 °C. Glass transition temperature (T_g) was determined as the inflection point of the heat capacity change, whereas melting temperature (T_m) and cold crystallisation (T_cc) were established as the maximum peak heights. Melting enthalpies (ΔH_m) were calculated by integrating the endothermic peaks. Overall theoretical degree of crystallinity X_c was calculated based on the melting enthalpy according to the following equation [30]:

1

Where ΔH^°_PHB is the melting enthalpy of pure PHB crystals, i.e. 146 J g^− 1 [31].

Polarised bright field (BF) imaging

1 × 1 cm cast film samples were placed on a cover slide and held with tape. Cast film samples were imaged at 20x magnification on a BX51 Upright Microscope equipped with a LC35 camera (Olympus, Japan). A BX-POL system (Olympus, Japan) was utilised for polarised light imaging, consisting of an analyzer (U-ANT) and polarizer (U-POT).

Results

Bacteria optical density, pH, cell dry weight and polyhydroxyalkanoate accumulation over 120 h are firstly summarised. Following this, polymer composition and molecular weight of the extracted PHAs is examined, followed by the thermal properties. Subsequently, the blending of PHB and mcl-PHA thin films is discussed, followed by modeling of the infrared spectral data to quantify.

Bacteria growth and PHA accumulation

Figure 2(a) shows the optical density, pH, cell dry weight, and PHB accumulation over 120 h. C. necator H16 was able to utilise glucose for PHB accumulation when a phenotype was isolated; however, growth was observed to be slower than fructose, reaching a maximum OD600 of 10.5. C. necator H16 on fructose media reached stationary phases at approximately 24 h, whilst stationary phase was not reached until 72 h for glucose. The pH for both sugars dropped to 6.4 within 72 h for both samples, however this occurred within 24 h for fructose and correlated to increased cell growth, as the drop in pH was associated with the production of acidic metabolites. The final CDW obtained was 4.127 g L^− 1 for fructose media and 3.213 g L^− 1 for glucose media. PHB accumulation increased over the 120 h, accumulating 2.485 g L^− 1 of PHA (60% w/w of the biomass) for fructose and 1.452 g L^− 1 of PHA (45% w/w of the biomass) for glucose.

Fig. 2.

(a) C. necator H16 GP growth rates based on the optical density at 600 nm (OD600), pH, cell dry weight and dry weight of PHA over time. (b) Combined overlaid DIC & CY3 fluorescence images of C. necator H16 GP grown (i) on nutrient media after 24 h, (ii) for 24 h on Fructose MSM, (iii) for 24 h on Glucose MSM, (iv) for 96 h on Fructose MSM, (v) for 96 h on Glucose MSM. Error bars, derived from triplicate samples, are included for all data points; where not visible, the associated error is smaller than the symbol dimensions

Figure 2(b) shows cells elongating due to PHA accumulation. Fluorescent inclusions within overlayed DIC images represent accumulated PHA structures, while ‘dark’ inclusions that are not stained with nile red represent other inclusion bodies/structures (e.g. polyphosphate granules) [28]. C. necator H16 after 24 h in nutrient media (i) grew in size to 6.7 μm, however PHA granules were only visible in a few cells and did not fully occupy the length of these elongated cells, leaving some intracellular space uncovered. Cell width was approximately 0.9 μm, and this did not vary widely for C. necator H16 throughout imaging. At 24 h, the cell size reduced for both media at 1.0–4.5 μm for fructose (ii) and 1.1–2.0 μm for glucose (iii) and PHA granules were visible in some cells. At 96 h, the cell length for fructose media had increased from 2.0 to 30 μm (iv), with many longer bacteria visible in the sample. PHA granules were visible in a majority of samples, and granules completely filled the cells, occupying the full length. For glucose media (v), the cell length stayed consistent at 1.0–2.0 μm, with PHA granules visible in most cells. The L/D ratio increased for C. necator H16 grown in fructose media after 96 h (2.11–30.53) compared to the NB growth media (1.06–7.88). The L/D ratio for C. necator in glucose media at 96 h was lower (1.5–3.1) compared to fructose media at the same time point. These images link with the CDW and PHA accumulation observed, with images at 24 h supporting a mid-late growth curve stage where CDW is increasing and PHA content is low, while images at 96 h correlate with a mid-stationary phase where PHA was accumulated by the cells.

P. putida KT2440 growth and PHA accumulation are shown in Fig. 3(a), with exponential phase occurring in the first 24 h, followed by stationary phase. Optical density reached a maximum of 8.9 for fructose and 6.7 for glucose media. pH levels dropped to 4.5 for both media within 72 h, likely due to the production of gluconate and 2-ketogluconate, which are mild organic acids [32]. Although a pH decrease was observed during all shake flask experiments, cultivation under pH-controlled bioreactor conditions resulted in comparable cell growth and PHA accumulation (see S3 & S4). CDW was greatest for P. putida KT2440 in fructose at 1.74 g L^− 1 at 48 h, while it reached a maximum of 1.36 g L^− 1 for glucose at 124 h. P. putida utilised both sugars in a similar fashion as shown in the growth data in Fig. 3(a), producing 22% and 18% mcl-PHA after 124 h for fructose and glucose respectively.

Fig. 3.

(a) P. putida KT2440 growth rates, pH, cell dry weight and dry weight of PHA over time at 2 L shake flask. Error bars represent deviation in CDW and PHA content in triplicate samples. (b) Combined overlaid DIC & CY3 fluorescence images of P. putida KT2440 grown (i) on nutrient media after 24 h, (ii) for 24 h on Fructose MSM, (iii) for 24 h on Glucose MSM, (iv) for 96 h on Fructose MSM, (v) for 96 h on Glucose MSM. Error bars, derived from triplicate samples, are included for all data points; where not visible, the associated error is smaller than the symbol dimensions

Figure 3(b) displays cell morphology changes for LB growth media and minimal media at exponential and stationary phases. In LB (i) bacteria length ranged from 1.1 to 3.3 μm, with little PHA granules visible in the cell. Length increased for both (ii, iv) fructose and (iii, v) glucose media at stationary phases reaching 1.4–3.3 μm and 1.4–2.4 μm respectively. Generally fructose media bacteria were observed to have longer cell sizes, and some granules were observed in both media, though their abundance was notably higher in C. necator H16 cultures. The L/D ratios for P. putida KT2440 did not show dramatic changes, implying a less pronounced change in cell morphology. Fructose media displayed a higher L/D range at 96 h (1.75 to 4.12) compared to LB growth media (1.38 to 4.12). The L/D ratio grown on glucose was similar to the range observed for cells grown on fructose, indicating a more consistent morphology regardless of the sugar used.

Polymer composition and molecular properties

The monomer chain length of PHAs accumulated can be found in Fig. 4(a). C. necator H16 was found to produce 100% PHB, independent of time or sugar source. P. Putida KT2440 was observed to produce predominantly decanoate (C10) PHAs, at ~ 60% of the accumulated PHA, with dodecanoate (C12) and octanoate (C8) observed as the other predominant monomeric subunits (< 1–1.9% C6, 8–14% C8, 58–64% C10 and 23–29% C12). Both the type of sugar substrate provided and the duration of fermentation did not play substantial roles in the composition of PHAs. Gel permeation chromatography elution peaks can be observed in Fig. 4(b). PHB produced by C. necator H16 from fructose media was found to have a weight average molecular weight (M_w) of 1723 ± 36.59 kDa, whilst on glucose media it was observed a M_w of 1117 ± 45.76 kDa (Table 1 & Table S5). P. putida KT2440 produced mcl-PHA’s were found to have a lower M_w of 74 ± 18.48 kDa for fructose media and 67.3 ± 15.5 kDa for glucose media.

Fig. 4.

(a) Ratio of 3-hydroxy, methyl PHA groups located in Cupriavidus necator H16 GP and Pseudomonas Putida KT2440 samples and (b) normalised gel permeation chromatography peaks for extracted PHA’s at 72 h

Table 1.

Thermal and molecular weight data for PHA species

Sample		Mw (kDa)	Tg (°C)	Tcc (°C)	Tm (°C)	∆Hm (J/g)	Χc (%)	FTIR Fit Χc (%)
Cupriavidus necator H16 GP	Fructose	1723 ± 36.6	2	41	175	61.1	42	46
	Glucose	1117 ± 45.8	0	44	172	60.9	42	40
Pseudomonas putida KT2440	Fructose	74 ± 18.5	-48	N/D	42	2.8	2	11
	Glucose	67.3 ± 15.5	-49	N/D	43	1.7	1	16

Film Sample	Tg (°C)	Tcc (°C)	Tm1 (°C)	Tm2 (°C)	∆Hm2 (J/g)	Χc (%)	FTIR Fit Χc (%)
100% PHB	5	-	-	175	83.5	57	52
80% PHB, 20% mcl-PHA	3	-	44	176	63.1	43	43
60% PHB, 40% mcl-PHA	2	49	43	174	58.1	40	41
40% PHB, 60% mcl-PHA	2	43	49	170	32.2	22	38
20% PHB, 80% mcl-PHA	1	50	48	171	26.2	15	26

Where; M_w is the weight average molecular weight of the polymer distribution curve for samples at 72 h, T_g is the glass transition temperature, T_cc is the cold crystallisation temperature, T_m is the melting temperature for extracted PHA’s, ∆H_m is the enthalpy of melting, X_c is the crystallinity in percent, T_m1 is the lower melting temperature for thin film samples, T_m2 is the higher melting temperature for thin film samples and ∆H_m2 is the enthalpy of melting for the higher melting temperature for thin film samples

FTIR of bacterial biomass and extracted PHA’s at 72 h is shown in Fig. 5(a). Key bands observed include amide 1 (1650–1660 cm^− 1) and amide 2 bands (1540–1550 cm^− 1), which denote C = O and N-H amide vibrations of proteins in the bacteria respectively. These amide stretches were observed clearly for both bacterial biomass. Bands at 1720–1740 cm^− 1 indicate the C = O ester carbonyl peak present in PHA, which can also provide information on the degree of crystallinity of the polyesters [34]. The 1740 cm^− 1 band is attributed to the amorphous phase, while the 1727 cm^− 1 corresponds to the crystalline phase [35]. P. putida KT2440 extracted PHA polymers were found to have broad carbonyl stretch at 1738 cm^− 1, denoting a more amorphous polymer structure, whilst C. necator extracted PHB had a sharp peak at 1721–23 cm^− 1, denoting a more crystalline structure. This large peak was also present in the C. necator biomass, however only a small peak could be observed in the P. putida biomass. Peaks at 2852–2856 cm^− 1, 2922–2930 cm^− 1 and 2976–2978 cm^− 1 regions represent CH₂–CH₃, CH₂–CH₂ and CH₃ bonding stretches respectively, with clear peaks found for both P. putida PHA samples due to the chain length of the PHA’s isolated.

Fig. 5.

FTIR of (a) bacteria species on different media after 72 h and isolated PHA and (b) cast films at different PHB/mcl-PHA blends. Regions of 1648–1658 cm^− 1 and 1539–1546 cm^− 1 represent the Amide 1 and 2 bands from the cell mass respectively. The 1720–1740 cm^− 1 region represents the ester carbonyl peak (C = O) dominant for PHA, with 1720 denoting a more crystalline region, while 1740 denotes a more amorphous region [33]. 2852–2856 cm^− 1, 2922–2930 cm^− 1 and 2976–2978 cm^− 1 regions represent CH₂–CH₃, CH₂–CH₂ and CH₃ bonding stretches respectively, predominantly found in longer chain length PHA structures

Thermochemical behaviour of pha’s

Thermal data based on Differential Scanning Calorimetry (Figure S5) analysis is summarised in Table 1. A T_g was observed at 0–2 °C for PHB produced by C. necator H16 and at -48 – -49 °C for mcl-PHA produced by P. putida KT2440, whilst a T_m was observed at 172–175 °C for PHB produced by C. necator H16 and 42–43 °C for mcl-PHA produced by P. putida KT2440. Degree of crystallinity was obtained using the melting enthalpy divided by the melting enthalpy of pure PHB crystals (Eq. 1) and was found to be 42% for PHB (crystalline), whilst mcl-PHA was ~1–2% (amorphous). The FTIR spectra Gaussian fit of the carbonyl region (1740–1725 cm^− 1) and second derivative fits produced crystallinities of 40–46% for PHB and 11–16% for mcl-PHA, which was slightly higher than crystallinities calculated from DSC (Table S8).

Cast films were prepared using different ratios of PHB (from C. necator on fructose) and mcl-PHA (from P. putida on glucose, with 11% C8, 63% C10 and 23% C12 monomers), and are shown in Fig. 6. Films decreased in transparency up until the 60% PHB, and increased in transparency as PHB content decreased. As mcl-PHA content increased, the adhesiveness of the films also increased, to where the films could no longer be removed from the petri dish without damaging them at 20% PHB content. Elongation at break analysis was undertaken for all films except 20:80 PHB:mcl-PHA, as this was too adhesive, which prevented reliable sample handling and clamping. An elongation at break of range 9–11% was obtained, which increased with the mcl-PHA content in the blend (S8). Larger domains were visible under bright field microscopy as mcl-PHA content increased. A grainy texture was observed for the 100% PHB cast film likely representing crystalline regions typical of pure PHB, which is highly crystalline. Increasing mcl-PHA increased the size of the structures visible in the images, however phase separation was visible. For all samples, the birefringence was low and did not change as the sample decreased in PHB content. In Fig. 5, FTIR analysis of the PHB and mcl-PHA cast film blends can be observed, showing a shift from 1720 cm^− 1 to 1738 cm^− 1 as mcl-PHA content increased, indicating lower crystallinity, while peaks associated with longer PHA chain length CH₃ and CH₂ groups with antisymmetric and symmetric stretches at 2854–2859 cm^− 1 and 2924 cm^− 1 also increased in size. Two melting peaks were observed in DSC and assigned for the two polymers in Table 1. Melting point 1 (T_m1) was seen to increase as mcl-PHA increased in the blend, while melting point 2 (T_m2) decreased (Figure S6). However, glass transitions were only observed for the PHB polymer in the films. As the percentage of mcl-PHA increased in the film, the crystallinity decreased and a shift in glass transition and melting points was observed (Figure S9). FTIR peak fitting via a Gaussian applied to the secondary derivative (Table S9) produced a similar crystallinity for high PHB content, however as mcl-PHA content increased above 40%, the fit suggested greater crystallinity than that calculated from the DSC enthalpy (Eq. 1).

Fig. 6.

Images of cast films made from PHA’s extracted from bacteria cells and microscope images (20x lens, BF) of the films (middle column) and cross polarised (final column)

Discussion

In this study, C. Necator H16 accumulated a maximum of 60% PHB of the cell dry weight after 120 h grown on fructose, while a maximum of 22% mcl-PHA was achieved by P. putida KT2440 on the same sugar. Both bacteria species were able to accumulate PHA’s on fructose and glucose, however, the maximum was achieved on fructose. It has been previously observed that the concentration and ratio of carbon, nitrogen and phosphorus play a key role for PHA accumulation, with high C:N ratios of 80:1 maximising PHA accumulation while reducing biomass production, with the ideal ratio being found to be 20:1 for both biomass production and PHB accumulation [36, 37]. The carbon, nitrogen and phosphorus ratio of C. necator hexose media in this study was 18.7:1:5.2, close to conditions favouring biomass and PHB production. In previous studies, when grown on glucose after isolating the glucose utilising phenotype, C. necator H16 was able to accumulate 45 wt% PHB after 120 h (C:N:P ratio 31:1:4) [23]. When grown in three stages of shake flask fermentation with fructose and avocado oil, C. necator H16 was able to accumulate 77% PHA [38]. The PHA accumulation achieved by fructose and glucose reached 60 wt% and 45 wt% respectively in this study; in agreement with literature.

P. putida KT2440 accumulated more PHA when grown on fructose than when grown on glucose, however the monomer composition was similar independent of sugar or time, accumulating predominantly C10 and C12 monomers in the polymers, with a media containing a C:N:P ratio of 3.3:1:0.57. P. putida KT2440 concentrations and PHA monomeric composition agree with literature, where glucose MSM at 72 h with C:N:P ratio of 15.3:1:4.9 was found to accumulate 4 g L^− 1 CDW containing 20% mcl-PHA, made up of 3% PHHx, 22% PHO, 52% PHD and 23% PHDd [35]. Though using the same bacteria species, temperature and time, the monomer composition varied, however C10 and C12 polymers were still the most produced. Utilising glucose media with a C:N:P ratio of 30.5:1:7.4, another study using P. putida KT2440 obtained ~ 3.8 gL^− 1 CDW after 60 h, containing 24% mcl-PHA consisting of 11.8 ± 0.2% PHO, 73.0 ± 0.1% PHD, 4.7 ± 0.2% PHDd, 9.4 ± 0.7 PH5Dd and 0.8 ± 0.3 PHTd. This higher ratio of C:N achieved similar PHA content to this study, however produced predominantly C10 and C8 monomers. P. putida KT2442, the rifampicin-resistant variant of KT2440 with different mcl-PHA nitrogen limitation and growth rate performance [39], was grown on fructose MSM with a large C:N:P ratio of 68.2:1:11.5 to accumulate 24.5% mcl-PHA [40]. This mcl-PHA had a similar monomer ratio, with 0.5% PHHx, 12.6% PHO, 70.8% PHD, 8.5% PH5Dd, 5.7% PHDd, 1.6% PH7Td and 0.3% PHTd [40].

The reduction in pH for P. putida KT2440 bacteria on glucose media to 4.5 ± 0.1 after 72 h is likely due to the production of gluconate and 2-ketogluconate, which are mild organic acids excreted into the medium during glucose consumption [32]. This pathway involves phosphorylation to glucose 6-phosphate (G6P), followed by 6P-gluconate, however glucose can also diffuse into the periplasm and be converted to gluconate and 2-ketogluconate [41]. P. putida KT2440 lacks the enzymatic machinery to convert fructose into gluconate. Unlike glucose, which is oxidized in the periplasm to gluconate via glucose dehydrogenase and then enters the Entner–Doudoroff (ED) pathway, fructose is metabolized through a distinct route involving its conversion to fructose-1-phosphate and subsequent cleavage into triose phosphates (DHAP and G3P) [42]. Consequently, the observed drop in pH during fructose metabolism may not be due to gluconate accumulation, but more likely the result of the secretion of organic acids such as acetate or pyruvate under high carbon flux or nutrient-limited conditions. Interestingly, growth rate, CDW and PHA accumulation were not affected by this extra metabolic step. This is further supported by literature which observes P. putida KT2440 metabolises these sugars through different metabolic routes, with glucose being processed via the ED pathway, while fructose can be processed through the Embden–Meyerhof–Parnas pathway (EMP), with 34% of the upper carbon flow going into glycolysis and 52%, going through the ED pathway [41]. It is hypothesised that P. putida could use the ED route to rapidly convert glucose for gluconate and sequester available carbon away from its competitors [41].

The presence of granules within the bacteria corresponded to the percentage of PHB accumulated in each cell, however the morphological differences for these species suggest effects of external stimuli such as nitrogen limitation and sugar utilisation, with glucose media taking longer for C. necator H16 to produce PHB at high levels in previous work [23]. Bacteria cell length varied during different stages of growth for C. necator H16 (Fig. 1b), while P. putida KT2440 stayed relatively similar (Fig. 2b), agreeing with literature [25], however the width stayed constant for both species. When C. necator H16 was placed onto fructose or glucose minimal salt media, the average cell size decreased, likely due to nitrogen limitation response indicating that the cells utilised their intracellular PHA storage. At 96 h, larger changes were observed, with C. necator H16 fructose media producing long bacteria up to 10 μm in length, containing multiple PHB granules. One sample and image found a PHB bacteria ~ 30 μm in length containing ~ 54 PHB granules, a size and granule number not reported previously. Glucose media did not produce such large bacteria, with a maximum size of 2 μm observed, however 1–2 granules were visually observed in most structures. P. putida KT2440 morphology was similar for all samples, with cell size staying constant around 1–3 μm. LB media showed little granule formation after 24 h, while both glucose and fructose media had 1–2 granules of mcl-PHA observed in most cells. Cell size is controlled by external factors and not PHB accumulation, as PHB non-producing mutants also increase in cell size, suggesting there is no direct control mechanism regulating the prolongation of cells in response to the presence of PHA granules [43].

Deconvolution and Gaussian curve fitting were undertaken to estimate crystallinity percentage via FTIR (summarised in Supplementary Information). The crystallinity percentages from PHB produced on fructose, glucose, and mcl-PHA on fructose and glucose were determined to be 46%, 40%, 11% and 16%, respectively. When compared to the DSC calculated crystallinity, values of 42%, 42%, 2% and 1% were obtained respectively. Similar data was also observed for PHB with a crystallinity of 46% for C. necator [44]. Lower crystallinity was observed via the DSC method for mcl-PHA. Values found in literature via DSC method reported crystallinity (~ 13%) and T_m values (54 °C) for mcl-PHA (10–18% PHO, 72–78% PHD and 8–12% PHDd) produced on glucose by Pseudomonas putida KT2442 [45]. Mcl-PHA produced from P. resinovorans (12% PHHx, 48% PHO, 31% PHD, 8% PHDd, ~ 1% PHTd) had a very low crystallinity value of 6.0 ± 0.2% [46]. This is due to the steric hindrance of mcl-PHA’s, with the longer carbon chains introducing more freedom of movement and rotational flexibility, preventing the polymer chains from packing tightly into ordered crystalline structures. DSC analysis determined lower glass transition and melting temperatures for mcl-PHAs, indicating an amorphous structure, likely due to random assembly of constituent monomers. The mcl-PHA isolated in this study had a larger component of longer chain PHDd, likely contributing to their low crystallinity. Intracellular PHB is often embedded within a matrix of proteins and other cellular components, which can act as plasticizers and interfere with crystalline packing, thereby reducing overall crystallinity. Additionally, since PHB granules remain dispersed within the unextracted biomass, local variations in granule density across the film surface may contribute to spatial heterogeneity in the measured crystallinity.

Though the PHB crystallinity percentages are similar between methods, a large difference is observed between techniques for the mcl-PHA samples. This is likely due to FTIR-derived crystallinity values being calculated based on localized measurements of peak intensities in the carbonyl region (~ 1720–1740 cm⁻¹), which are sensitive to the local molecular environment and polymer phase distribution. In contrast, DSC provides a bulk measurement. Therefore, FTIR alone can be used as an indicator for crystallinity and provide much faster results than DSC or XRD. Biomass samples of C. necator H16 grown on fructose also had their FTIR fit via the Gaussian, producing a crystallinity of 23%; half the 46% crystallinity of the extracted sample. While this difference may be partly attributed to cellular impurities and extraction effects, it may also reflect reduced crystallinity of the polymer in situ [47].

Very thin cast films with a thickness of 20 ± 5 μm were produced to evaluate the thermal and nanoscale interactions, as well as the morphological characteristics. Gaussian fits showed crystallinity of 60% for 100% PHB, 43% for 80% PHB, 41% for 60% PHB, 38% for 40% PHB and 26% for 20% PHB films (Table 1). These agreed with DSC crystallinity calculations at higher crystallinities, however DSC crystallinity was lower for 40% and 20% PHB films, likely due to the use of the melting enthalpy of pure PHB crystals. DSC data followed a similar trend, with PHB melt temperature decreasing from 176 °C to 170 °C as PHB ratio reduced in the blend, while T_g and the T_m1 of mcl-PHAs in the blend increased from 44 °C to 49 °C (Table 1). The dual melting points of the films offer interesting properties as a smart, multifunctional material, which softens or becomes permeable at moderate heat (~ 45 °C), ideal for controlled release or thermal responsiveness, and retains structural integrity up to ~ 170 °C, allowing robust handling or processing. This property could be related to the structure of the PHB in the blend, with a decrease in PHB crystallite size hypothesised as mcl-PHA ratio increased in the blend, as PHB spherulites grown in the presence of a PHB melt contain PHO, PHD and PHDd domains as the dispersed phase [10]. This phenomenon was visible under the microscope, with a change in the structure occurring as PHB content decreased, forming globules of mcl-PHAs. This was most apparent for 20% PHB films, where mcl-PHAs form visible globules within the PHB matrix, suggesting a phase-separated structure. This biphasic morphology, characterised by partial miscibility during DSC, supports a system where PHB and mcl-PHAs form distinct domains, which was similarly observed in thicker films (~ 500 μm) of PHB/PHO (C8) blends [10]. At low PHB concentrations, PHO formed the matrix and the mechanical properties of the material were governed by the theory of rubber elasticity, due to the orientation of the amorphous rubbery chains of the PHO matrix. Inversely, at high PHB content, PHB formed the matrix and the mechanical behaviour of the material was from enthalpic origin [10]. Researchers investigating PHB blends with mcl-PHAs (27 mol% PHO, 24 mol% PHN, 17% PHHp, 16 mol% PHD) observed that multiple melting peaks were present during DSC on PHB due to α and β crystal forms produced, however no film thickness was recorded [48]. Increasing the mcl-PHA reduced the degree of crystallinity compared to PHB, and small amounts of mcl-PHA restricted crystallization of PHB but higher ones promoted it [48]. Notably, the dimensions of individual PHB crystallites reached up to 30 μm, larger than the thickness of the thin films developed in this work.

Elongation at break analysis was undertaken on the samples to understand the effect on mechanical properties, finding that this slightly increased from 9 to 11% as mcl-PHA content increased. Previous work on co-culturing Cupriavidus necator DSM 428 and Pseudomonas citronellolis NRRL B-2504 grown on apple pulp waste to produce 48% PHB/52% mcl-PHA films of 100 μm thickness observed improved elongation at break of 338 ± 19%, however DSC suggested phase separation due to multiple melting and glass transition temperatures [49]. The thicker films supported increased elongation at break, however phase separation was observed, similar to this study. Blends of PHBV and mcl-PHA (41.6 mol% POH, 35.9 mol% P3HD and 22.5 mol% P3HDD) from Pseudomonas putida KT2440 grown on lauric acid films of 200–400 μm thick, varying from 5 to 40% mcl-PHA, showed that as mcl-PHA content increased in the blend elongation at break stayed similar around 5–15% [50]. Young’s modulus, tensile strength and stiffness all decreased, as well as melting temperature and crystallinity, indicating polymer interaction, which is generally characteristic of compatible systems [50]. To compare elongation at break data to film thickness across literature, a ratio of elongation at break to film thickness was plotted in Figure S3. This suggests that thicker PHA films (> 100 μm) are able to perform well even as a biphasic system due to the effects of phase distribution, interfacial tension, as well as the extra volume enabling more structured phase separation, creating a microstructure where soft mcl-PHA phases act as a continuous or semi-continuous phase around PHB. In thin films, the confined geometry and rapid film formation may restrict the development of biphasic morphologies, causing the softening effects of mcl-PHA to be less pronounced, and large PHB crystals may directly disrupt the matrix, limiting improvements in elongation at break even at high mcl-PHA content. This has been previously noted for 80% PHB / 20% PLLA ultrathin films, where 13 nm thickness reduced PHB crystallisation with low molecular weight PLLA (M_w = 6.9 kDa), while 30 nm films showed phase separation [51]. When developing films for industrial applications, this interaction of PHA’s is important to understand.

A large visual birefringence was not observed in these samples. PHB, a well ordered polymer, was hypothesised to produce birefringence due to the crystallinity of the sample, while mcl-PHA’s were expected to reduce this due to having more flexible and amorphous structures because of their longer side chains, reducing the degree of molecular alignment and crystallisation. The muted birefringence across all films, even in pure PHB, suggests limited long-range crystalline order, possibly due to the thinness of the films. This indicates that at 20 μm thickness, PHB may not achieve extensive crystallisation, leading to smaller, less organised crystalline regions that diminish birefringence. To our knowledge, this is the first study to systematically produce and characterize solvent-cast PHA blend ultrathin films with controlled PHB/mcl-PHA ratios (100:0 to 20:80) at a thickness of 20 ± 5 μm, enabling direct evaluation of composition-dependent structure–property relationships in ultrathin matrices.

Conclusion

PHA was successfully synthesised and extracted from glucose and fructose sugar substrates by Cupriavidus necator H16 and Pseudomonas putida KT2440. Fructose produced the highest %PHA in Cupriavidus necator H16 and Pseudomonas putida KT2440 at 60% and 22% respectively. pH reduction during growth on hexoses in Pseudomonas putida KT2440 was linked to organic acid production, though growth and PHA yields remained unaffected by this pH drop. A crystalline PHB was produced by Cupriavidus necator H16, while Pseudomonas putida KT2440 produced an amorphous mcl-PHA consisting of 1.1–2.8% C4, < 1–1.9% C6, 8–14% C8, 58–64% C10 and 23–29% C12 monomers. Systematic blending of these materials to produce thin films formed a biphasic system, where film thickness limited PHB crystal structure and formation, with tunable crystallisation of the material depending on PHB and mcl-PHA content. FTIR analysis and peak fitting at the carbonyl region (1720–1740 cm⁻¹) showed similar crystallinity results to DSC analysis, enabling a fast method to indicate film crystallinity. The findings from this study suggest that blending PHAs from these two bacterial species at thin films (20 μm) could be a promising approach to develop bioplastics with a desirable balance of thermal and mechanical properties suitable for a range of applications, particularly for biomedical applications, smart thermal sensitive packaging or temperature-sensitive drug delivery systems.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

mcl-PHA

medium- chain-length poly(3-hydroxyalkanoate)

PHB

poly(3-hydroxybutyrate)

PHBV

poly(3-hydroxybutyrate-co-3-hydroxyvalerate)

PHHx

poly(3-hydroxyhexanoate)

PHO

poly(3-hydroxyoctanoate)

PHD

poly(3-hydroxydecanoate)

PHDd

poly(3-hydroxydodecanoate)

PH5DD

poly(3-hydroxy-5-cis-dodecanoate)

PHTd

poly(3-hydroxy tetradecanoate)

PH7Td

poly(3-hydroxy-7-cis-tetradecenoate)

T_c

Crystallisation temperature

T_cc

Cold crystallisation temperature

T_g

Glass transition temperature

T_m

Melting temperature

Author contributions

EA: Conceptualization, Formal analysis (all bacterial and film work), Investigation, Validation, Writing – original draft. FYP: Formal analysis (microscopy imaging), Investigation. RN: Design and interpretation of Microscopy work, Writing – review & editing. MMBH: Image analysis and polymer characterization supervision and conceptualization, Writing – review & editing, Supervision. LvH: Writing – review & editing, Supervision, Funding acquisition, Conceptualization of the study, all data analysis and processing.

Funding

This project was supported by Great Wrap, with funding from the Australian Government Cooperative Research Centres Projects (CRC-P) Grants Round 10.

Data availability

The authors declare that the data supporting the findings of this study are available within the paper and its Supplementary Information files. Should any raw data files be needed in another format they are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.