# Bioplastics: PLA, PHA, bio-PE and starch blends: PLA and PHA: the new polyesters

Source page: https://www.materialinsights.org/learn/chemicals/bioplastics-pla-pha-biope-starch/p2/

[Understanding Chemicals](https://www.materialinsights.org/learn/chemicals/) / [Module 14: Bioplastics: PLA, PHA, bio-PE and starch blends](https://www.materialinsights.org/learn/chemicals/bioplastics-pla-pha-biope-starch/) Part 2 of 5

## PLA and PHA: the new polyesters

PLA and PHA are both polyesters built from small hydroxy acids, and both start with sugar. They differ in who builds the chain. PLA's monomer is made by fermentation and then polymerised by chemists in a melt reactor. PHA is polymerised by bacteria inside their cells and must be dug out again, which is why the two sit at very different points on the cost curve.

### 2.1 PLA begins as lactic acid from fermented sugar

Lactic acid is a small molecule with an acid group and an alcohol group on neighbouring carbon atoms, so it can link head to tail into a polyester. It comes in two mirror-image forms, L and D, like left and right hands. Chemical synthesis gives a 50:50 mixture, which is useless for a crystalline polymer; fermentation gives either form at close to 100% purity, which is why commercial PLA depends on biology (Kamm 2006, Vol 2 Ch 14).

The lactic acid plant has four areas (figure). In fermentation, bacteria such as Lactobacillus turn dextrose from corn, or sucrose from cane or beet, into lactic acid. The acid would soon stop the microbes, so lime or chalk is added to hold the pH between 5.0 and 6.8; the broth therefore contains calcium lactate. In acidification, sulphuric acid releases the lactic acid, and calcium sulphate (gypsum) crystallises and is filtered off. Purification removes cells, leftover sugar and nutrients, by solvent extraction with an amine, by distillation, or by making an ester and distilling that. Finally the solution is concentrated to 60-70% lactic acid at better than 98% optical purity and piped to the polymer plant (Kamm 2006, Vol 2 Ch 14).

Gypsum is the awkward by-product. By the reaction's arithmetic, two lactic acid molecules (2 x 90 g) come with one of gypsum (172 g), so roughly a tonne of gypsum for every tonne of acid. Producers place it in plasterboard, cement or farming, or avoid it: Cargill developed a microbe that works below pH 4.0, cutting both lime and sulphuric acid (Kamm 2006, Vol 2 Ch 14).

Sugar is the largest single cost, 25-40% of the total cost of PLA made from corn in 2006. That is why virtually every lactic acid plant sits beside a sugar or corn mill (Kamm 2006, Vol 2 Ch 14).

Process flowchart From corn sugar to PLA pellets (Cargill Dow / NatureWorks route) Inputs **Dextrose from corn (or cane or beet sucrose)**25-40% of PLA cost (2006) **Lime or chalk**holds fermenter pH at 5.0-6.8 **Sulphuric acid**frees lactic acid from calcium lactate **Tin octoate catalyst**100-1,000 ppm in polymerisation **Natural gas and coal-fired power**process heat and electricity (Blair, Nebraska) Process steps 1 **Fermentation**

Lactobacillus, Bacillus or Rhizopus convert sugar to calcium lactate; L or D form at near 100% chiral purity
 2 **Acidification**

sulphuric acid added; lactic acid released, gypsum filtered off
 3 **Purification and concentration**

cell removal; amine extraction, distillation or ester distillation; 60-70% lactic acid, >98% optical purity
 4 **Prepolymer**

continuous condensation drives off water to give short PLA chains
 5 **Lactide reactor and vacuum distillation**

tin catalyst cyclises short chains into lactide, a ring of two lactic acid units; lactide purified by distillation
 6 **Ring-opening polymerisation**

in the melt, no solvent; 180-210 C, 2-5 h to 95% conversion
 7 **Devolatilisation and pelletising**

unreacted lactide removed under vacuum and recycled to the start
 Outputs **PLA pellets**1 t, needing about 1.25 t lactic acid (each 90 g unit loses 18 g of water: 90/72) **Gypsum**roughly 1 t per t lactic acid on the calcium route

The Blair, Nebraska plant was rated at 300 million lb of PLA a year, about 136,000 t.
 Source: Kamm (2006), Vol 2 Ch 14, pdf pp. 835-846 (book pp. 382-393), p. 855  In the data

#### Who made lactic acid in 2006

Kamm lists Purac (Netherlands, Spain, Brazil), Galactic (Belgium), ADM (Illinois), and joint ventures in Nebraska and China, mainly for food and industrial markets at 20-100 million lb a year each. Cargill Dow's Nebraska plant was the largest single plant, at a declared 400 million lb a year (about 181,000 t), nearly all used for polymer. Almost every producer was tied to a company supplying its sugar.

- Kamm (2006), Vol 2 Ch 14 Polylactic Acid from Renewable Resources · pdf pp. 834-843, 855 (book pp. 381-390, 402)

### 2.2 Ring-opening polymerisation turned PLA into a commodity

Lactic acid can be polymerised directly by driving off water, but that is an equilibrium: the last traces of water are hard to remove, so the chains stay short. Direct condensation needs 180-200 C, a vacuum below 5 mmHg, long reaction times, a catalyst and a solvent that carries water away to exceed a molecular weight of 100,000 (Kabasci 2014, section 8.4). The route that made PLA cheap goes round the problem. A short prepolymer is heated with a tin catalyst so that its chain ends bite themselves off as lactide, a six-membered ring of two lactic acid units. The lactide is purified by vacuum distillation, then opened and strung together in the melt. Each ring-opening adds two units and releases no water, so chains grow fast. DuPont reached 98% conversion in five minutes at 180 C, but acidic impurities at six times the catalyst level cut it to 12% (Kamm 2006, Vol 2 Ch 14). Lactide purity is everything.

Optical purity is the control knob. Pure L-PLA crystallises; a little D or meso lactide acts like kinks in the chain, lowering the melting point by about 3 C per 1% meso-lactide, and below about 78% optical purity the polymer cannot crystallise at all. Thermoforming grades deliberately carry 4-8% D so the sheet stays amorphous while it is formed (Kamm 2006, Vol 2 Ch 14).

PLA processes on PET-type equipment with two cautions: dry the pellets, because water cuts chains in the melt, and keep temperatures and times low, because thermal breakdown matters above about 250 C (Kabasci 2014, section 8.5; Kamm 2006).

The weak points are brittleness and heat. PLA's Tg is about 58 C, so an amorphous cup softens in a hot car. Remedies include nucleating agents, fibre composites, and stereocomplexes of L and D polymers, whose crystals melt far higher, approaching 230 C in practical resins (Kamm 2006, Vol 2 Ch 14; Kabasci 2014, section 8.12).

Data Typical properties of two NatureWorks PLA grades (2006)

| Property | 2000D (extrusion and thermoforming) | 3010D (injection moulding) |
|---|---|---|
| Specific gravity | 1.25 | 1.21 |
| Tensile strength at break | 53 MPa | 48 MPa |
| Tensile yield strength | 60 MPa | not given |
| Tensile modulus | 3.5 GPa | not given |
| Flexural modulus | not given | 3.8 GPa |
| Elongation at break | 6.0% | 2.5% |
| Clarity | transparent | transparent |

Grades differ in stereochemical purity, molecular weight and additives.
 Source: Kamm (2006), Vol 2 Ch 14, Table 14.4, pdf p. 853 (book p. 400)  Case

#### NatureWorks: scale first, profit later

Cargill and Dow formed Cargill Dow in 1997 to cut the cost of PLA, which had kept it in medical niches at more than $2 per lb. Their solvent-free melt process won the 2002 US Presidential Green Chemistry Award (Kamm 2006). Scaling it was not smooth: Rosenboom et al. (2022) report that optimising one process step at large scale caused billion-dollar losses over several years before the business broke even. By about 2013 NatureWorks, a Cargill company, made some 140,000 t a year (Kabasci 2014, section 14.2).

- Kabasci (2014), Ch 8 Poly(Lactic Acid) · sections 8.4, 8.5, 8.12
- Kamm (2006), Vol 2 Ch 14 Polylactic Acid from Renewable Resources · pdf pp. 844-855 (book pp. 391-402)
- Rosenboom et al. (2022), Bioplastics for a circular economy · pdf p. 14 (journal p. 130)

### 2.3 PHA is a plastic that bacteria store like fat

Polyhydroxyalkanoates are polyesters that many bacteria make as an energy reserve, much as animals store fat. Give a culture plenty of carbon but starve it of nitrogen, phosphate or oxygen, and the cells stop growing and pack the surplus carbon into polymer granules 0.2-0.7 micrometres across. Under good conditions the granules exceed 90% of the cell mass (Kabasci 2014, section 7.1). PHA is the only family of bio-based plastics that living cells both make and break down.

The simplest member, PHB, is a chain of 3-hydroxybutyrate units. It is stiff and crystalline, melting at about 175 C, with properties close to polypropylene's: modulus 2.5-3.5 GPa and tensile strength 20-40 MPa, but only 3-5% elongation, so it is brittle (Kamm 2006, Vol 1 Ch 10). Its processing window is narrow: it should not spend more than five minutes above 160-170 C. Feeding the bacteria a precursor such as propionic acid builds 3-hydroxyvalerate units into the chain; the copolymer PHBV is tougher and melts lower, at 71 C with 40% valerate (Kamm 2006, Vol 1 Ch 10). Medium-chain PHAs behave like rubbers (Kabasci 2014, section 7.4).

The process has two halves (figure). Fermentation is aerobic and fed-batch, meaning sugar is added as the culture consumes it. In the Brazilian design, Ralstonia eutropha grows for 24-28 h, then stores polymer for 30-35 h at 32-34 C and pH 6.8, reaching 150 kg of dry cells per cubic metre, 75% of it PHB. Heat release and oxygen demand set the size of each fermenter. Recovery is the second half: the cells must be broken and the polymer separated from proteins, fats and nucleic acids. Chloroform works but is hazardous; enzymes and detergents leave impurities. The Brazilian process uses isoamyl alcohol instead, a by-product of the mill's own ethanol distillery (Kamm 2006, Vol 1 Ch 10; Kabasci 2014, section 7.3).

Process flowchart PHB from sugar in a Brazilian cane mill Inputs **Medium-grade cane sugar**29% of production cost **Ammonium, phosphorus and salts**nitrogen cut back to trigger polymer storage **Air**vigorous aeration for high oxygen uptake **Steam and power from bagasse**boiler at 60 bar and 450 C **Isoamyl alcohol**from the distillery's fusel oil; recycled Process steps 1 **Growth phase (fed-batch)**

Ralstonia eutropha, 32-34 C, pH 6.8, 24-28 h, low carbon-to-nitrogen ratio
 2 **Storage phase**

high carbon-to-nitrogen ratio, 30-35 h; 150 kg/m3 dry cells, 75% PHB; 1.89 kg PHB per m3 per h
 3 **Inactivation**

heat to 105 C for 5 min so starving cells do not consume their polymer
 4 **Coagulation and concentration**

phosphoric acid, lime and flocculant; settling and centrifuge to 25-35% solids
 5 **Solvent extraction**

countercurrent stirred tanks at the boiling point of the solvent-water mixture; breaks cells and dissolves PHB
 6 **Filtration and precipitation**

diatomaceous earth bed removes debris; cooling precipitates PHB; microfiltration from 1.5% to 4%
 7 **Solvent stripping and drying**

evaporation and steam stripping; dried to <0.3% moisture
 Outputs **PHB powder**99% pure grade; 500-800 kDa **Cell debris cake**composted onto cane fields **Liquid effluents**sprayed on cane fields like vinasse

Design basis: 10,000 t PHB a year over 330 days, beside a mill crushing 12,000 t of cane a day in a 180-day season.
 Source: Kamm (2006), Vol 1 Ch 10, pdf pp. 245-250 (book pp. 217-222)  In the data

#### Salt-loving bacteria skip sterilisation

Sterilising big fermenters costs energy, and a contaminated batch is lost. A salt-loving bacterium, Halomonas TD01, ran for 14 days in an open, unsterile, continuous process at 40 g of cells per litre containing 60% PHB (Kabasci 2014, section 7.7.5). Such strains also burst in fresh water, releasing their granules without solvent.

- Kabasci (2014), Ch 7 Polyhydroxyalkanoates · sections 7.1, 7.3, 7.4, 7.7.5
- Kamm (2006), Vol 1 Ch 10 Sugar-based Biorefinery (PHB, sugar and ethanol) · pdf pp. 237, 242-250 (book pp. 209, 214-222)

### 2.4 PHA still costs several times more than the polyolefins

The Brazilian authors costed a 10,000 t a year PHB plant (table). Installed equipment came to $38 million, split equally between fermentation and recovery: about $3,800 per tonne of annual capacity. Production cost was $2.25-2.75 per kg when integrated with a sugar mill and $2.50-3.00 per kg as a stand-alone plant; sugar was the largest item at 29% (Kamm 2006, Vol 1 Ch 10, mid-2000s figures). Kabasci (2014, section 7.2) puts the carbon substrate at up to half of PHA production cost in general.

Some arithmetic shows why scale is hard. 10,000 t over 330 days is 1,263 kg of PHB an hour. At 1.89 kg per cubic metre per hour, that needs about 670 cubic metres of working fermenter volume before any allowance for cleaning and refilling. The authors note that oxygen transfer limits fermenters to under 500 cubic metres each, so even this small plant needs several vessels. A single polyolefin unit makes several hundred thousand tonnes a year (Kabasci 2014, section 11.2).

Researchers attack cost through waste feedstocks such as whey, crude glycerol and molasses, continuous multistage reactors (2.14 g per litre per hour in a five-stage cascade) and cheaper recovery (Kabasci 2014, section 7.7).

Prices reflect all this. Rosenboom et al. (2022) list P3HB at $3-8 per kg against $1.1 for PP and $1.4-1.6 for HDPE, a premium they put at three to four times. PHA's attractions are real: PHBV was approved for food contact in Europe in 1996, PHAs have good oxygen and carbon dioxide barriers, and P3HB degrades in sea water in one to six months (Kabasci 2014, section 7.6; Rosenboom et al. 2022).

Data Investment and cost structure of a 10,000 t/yr PHB plant (Brazil, mid-2000s)

| Item | Value | Share |
|---|---|---|
| Buildings and civil works | $3.0 million | 7.9% of investment |
| Fermentation unit | $15.0 million | 39.5% |
| Downstream processing unit | $15.0 million | 39.5% |
| Utilities | $5.0 million (see note) | 13.2% |
| Total installed investment | $38 million | about $3,800 per t of annual capacity |
| Sugar substrate |  | 29% of production cost |
| Other raw materials and chemicals |  | 15.5% |
| Salaries |  | 12.3% |
| Depreciation of buildings and equipment |  | 11% |
| Others |  | 8.6% |
| Maintenance |  | 5.1% |
| Bagasse for energy |  | 4.9% |
| Production cost, integrated with mill | $2.25-2.75 per kg |  |
| Production cost, stand-alone | $2.50-3.00 per kg |  |

The source table prints utilities as $5,000; the 13.2% share and $38 million total imply $5 million. The listed cost shares add to 86.4%; the rest is not itemised.
 Source: Kamm (2006), Vol 1 Ch 10, Table 10.3, pdf pp. 250-251 (book pp. 222-223)  Case

#### Metabolix and ADM: a $339 million lesson

In 2010 Metabolix and Archer Daniels Midland opened an Iowa plant to make 55,000 t of PHA a year. Two years later sales had missed forecasts and income did not cover operating costs. ADM wrote off its $339 million investment, about $6,200 per tonne of annual capacity, and Metabolix sold the technology to CheilJedang of South Korea. PHA has since revived, with Danimer Scientific and RWDC Industries scaling against signed packaging contracts and Tepha focusing on high-margin medical uses (Rosenboom et al. 2022).

- Kamm (2006), Vol 1 Ch 10 Sugar-based Biorefinery (PHB, sugar and ethanol) · pdf pp. 246, 250-252 (book pp. 218, 222-224)
- Kabasci (2014), Ch 7 Polyhydroxyalkanoates; Ch 11 · sections 7.2, 7.6, 7.7, 11.2
- Rosenboom et al. (2022), Bioplastics for a circular economy · pdf pp. 8, 14 (journal pp. 124, 130), Table 1

### See it in the table
[H₂O Water](https://www.materialinsights.org/commodities/water/)[NaCl Salt](https://www.materialinsights.org/commodities/salt/)[Cem Cement](https://www.materialinsights.org/commodities/cement/)[Gyp Gypsum](https://www.materialinsights.org/commodities/gypsum/)[CaO Lime](https://www.materialinsights.org/commodities/lime/)[Su Sugar](https://www.materialinsights.org/commodities/sugar/)[H2SO4 Sulphuric acid](https://www.materialinsights.org/chemicals/sulphuric_acid/)[O2 Oxygen](https://www.materialinsights.org/chemicals/oxygen/)[CO2 Carbon dioxide](https://www.materialinsights.org/chemicals/carbon_dioxide/)[PP Polypropylene](https://www.materialinsights.org/chemicals/polypropylene/)

### Check your understanding

Why is lime or chalk added to a lactic acid fermenter?
 Lactic acid would quickly lower the pH and stop the microbes. The base converts it to calcium lactate; sulphuric acid later frees the acid and leaves gypsum.

A PLA producer adds 4-8% D-lactic units to a thermoforming grade. Why?
 D units act like kinks that slow and limit crystallisation. Thermoforming works best on an amorphous sheet, so the grade is designed not to crystallise during melt processing.

In the Brazilian PHB design, which change would most directly cut cost per kilogram?
 Sugar is 29% of production cost, the biggest item. The other options would either kill the bacteria, add hazard and cost, or reduce output per batch.

[← Part 1: What 'bioplastic' actually means](https://www.materialinsights.org/learn/chemicals/bioplastics-pla-pha-biope-starch/p1/)[Mark read · Part 3: Bio-PE, starch blends and the comparison →](https://www.materialinsights.org/learn/chemicals/bioplastics-pla-pha-biope-starch/p3/)

## Other sections

- [Module overview](https://www.materialinsights.org/learn/chemicals/bioplastics-pla-pha-biope-starch/index.md)
- [What 'bioplastic' actually means](https://www.materialinsights.org/learn/chemicals/bioplastics-pla-pha-biope-starch/p1/index.md)
- [Bio-PE, starch blends and the comparison](https://www.materialinsights.org/learn/chemicals/bioplastics-pla-pha-biope-starch/p3/index.md)
- [End of life, climate and land](https://www.materialinsights.org/learn/chemicals/bioplastics-pla-pha-biope-starch/p4/index.md)
- [Market, barriers and choosing the right material](https://www.materialinsights.org/learn/chemicals/bioplastics-pla-pha-biope-starch/p5/index.md)
- [Key terms, quiz and sources](https://www.materialinsights.org/learn/chemicals/bioplastics-pla-pha-biope-starch/review/index.md)