Making Plastic from the garbage

A special issue of the Journal of Industrial Ecology, guest edited by Robert Anex [http://www.abe.iastate.edu/faculty/anex.asp] Associate Professor of Agricultural and Biosystems Engineering, Iowa State University, Ames, Iowa, USA. Support for this special issue was provided by the U.S. National Institute of Standards and Technology (NIST) through a grant to Professor Tillman Gerngross of Dartmouth College.

RESEARCH AND ANALYS IS

http://mitpress.mit.edu/jie Journal of Industrial Ecology 63

2004 by the Massachusetts Instituteof Technology and Yale University

Volume 7, Number 34

Making Plastics from GarbageA Novel Process for Poly-L-LactateProduction from Municipal Food Waste

Kenji Sakai, Masayuki Taniguchi, Shigenobu Miura,Hitomi Ohara, Toru Matsumoto, and Yoshihito Shirai

Keywords

biodegradable plasticbiopolymerlactic acid fermentationmunicipal solid waste (MSW)recyclingpoly-L-lactate (PLLA)

Address correspondence to:Dr. Kenji SakaiDepartment of Applied Chemistry,

Faculty of EngineeringOita UniversityOita 870-1192 [email protected]/biochem/biochem

-e.html

Summary

We propose a novel recycling system for municipal foodwastethat combines fermentation and chemical processes to pro-duce high-quality poly-L-lactate (PLLA) biodegradable plastics.The process consists of removal of endogenous D,L-lactic acidfrom minced food waste by a propionibacterium, L-lactic acidfermentation under semisolid conditions, L-lactic acid puri-cation via butyl esterication, and L-lactic acid polymerizationvia LL-lactide. The total design of the process enables a highyield of PLLA with high optical activity (i.e., a high proportionof optical isomers) and novel recycling of all materials pro-duced at each step, with energy savings and minimal emissions.Approximately 50% of the total carbon was removed, mostlyas L-lactic acid, and 100 kg of collected food waste yielded7.0 kg PLLA (about 34% of the total carbon). The physicalproperties of the PLLA yielded in this manner were compa-rable to those of PLLA generated from commercially availableL-lactic acid. Evaluation of the process is also discussed fromthe viewpoints of material and energy balances and environ-mental impact.

RESEARCH AND ANALYS I S

64 Journal of Industrial Ecology

Introduction

The exhaustion of nite fossil resources as asource of energy and chemicals, as well as short-ages of food and feed, are serious problems to beovercome if society is to be sustained. In addi-tion, an environmentally friendly mode of treat-ing municipal refuse is required. Of about 50MMT of waste1 that is generated annually in Ja-pan, around 20% is high-moisture-content refusefrom kitchens and the food industry. Such wastesreadily decompose, generate odors, and some-times cause illness. Municipal solid wastes(MSWs), including food waste, are usually incin-erated or landlled, but these processes generatemany problems. Incineration facilities can bedamaged by temperature uctuations when foodwaste with high water content is burned in asemicontinuous process. In addition, it is difcultto recover energy from such waste incinerationprocesses because the heating value of food wasteis low (Harrison et al. 2000). Also, landll spaceis limited, and uncontrolled fermentation of or-ganic wastes in landlls causes secondary prob-lems, such as methane emissions (Camobreco etal. 1999). Under these circumstances, the food-recycling law was promulgated in Japan in 2001,thus putting pressure on managers of food wasteto improve the process of recycling waste food(Food Recycling Law 2000).

Treatment of biological solid waste2 via mi-crobiological processes improves these wastesand reduces the need for both landll space andfuel used in waste incineration. Direct compost-ing and methane fermentation, which producefertilizers and biogas, are alternative ways to re-use food waste, but these processes have beenapplied only in rural areas. On the other hand,our previous study, which aimed to develop aneconomical means of converting solid domesticand industrial food wastes into valuable products,showed that municipal food waste is a goodsource of natural lactic acid bacteria (Sakai et al.2000). This nding indicated another route offood waste reuse suitable for use in urban areas.

Lactic acid has both hydroxyl and carboxylgroups with one chiral carbon atom, and it iswidely used in the food, pharmaceutical, andgeneral chemical industries (Litcheld 1996). In

addition, lactic acid can be polymerized to formthe biodegradable and recyclable polyester poly-lactic acid, which is considered a potential sub-stitute for plastics manufactured from petroleum(Ohara and Sawa 1994). Although the esterbond of poly-l-lactate (PLLA)3 is susceptible tosome enzymes, including proteinases and lipases,and PLLA has been recognized as a biodegrad-able plastic (Sakai et al. 2001), its biodegrada-tion in soil is rather slow and depends on mor-phology and thickness (Miyazaki and Harano2001). Therefore, PLLAmay better be developedas a chemically recyclable plastic with an appro-priate collection system for the used materialsand not as a single-use plastic. The industrialproduction of PLLA from cornstarch has recentlybegun (Lunt 1998), and its potential use as agreen plastic has been discussed (Gerngross1999; Gerngross and Slater 2000). Such materi-als are expected to come into worldwide use, butPLLA and other plant-derived plastics are costly,thus preventing their widespread application. Inaddition, the process uses cornstarch feedstock,which is also a source of food for humans andother animals.

During the past decade, many authors havedescribed lactic acid production from wastes, in-cluding molasses (Aksu and Kutsal 1986), bakerywaste (Oda et al. 1997), waste waterpaper(Schmidt and Padukone 1997), MSW (Zhou etal. 1996), and sugarcane pressmud (Xavier andLonsane 1994). These studies, however, have op-erated using extracted liquids from the wastes,which contain well-known, uniform compo-nents, and these approaches have never been de-veloped into practical industrial processes. Herewe demonstrate a novel process in which high-quality PLLA, a recyclable plastic, is producedfrom actual (collected) municipal food waste bycombining two fermentation and chemical steps,thus creating a system for total reuse of municipalfood waste. We also discuss the practicability ofthe system in terms of materials and energy bal-ances, as well as environmental impact.

Materials and Methods

Figure 1 shows a schematic diagram of theprocess for production of PLLA from food waste.


Sakai et al., PLLA Production from Municipal Food Waste 65

Figure 1 Process outline of PLLA production from food waste. See experimental protocol. Photosadjacent to the gure show food waste (A), concentrated broth after lactic acid fermentation (B), puriedL-lactic acid (C), fermentation residue (D), and pellets of PLLA (E). Average amounts of intermediates andproducts yielded from 100 kg food waste are also included.

Food Waste

To investigate the amount of food waste gen-erated by the food industry, we sent question-naires to companies with over 20 employees inKitakyushu City, a city in southern Japan with apopulation of approximately 1 million. Of 851mailings, 15.2% of the replies were valid. Wethen calculated the amount of food waste dis-posed in 2001 per employee (basic unit) in eachtype of food business. We estimated from thesebasic units and the numbers of workers in Kita-kyushu the potential amount of food waste dis-posed by each type of food business in Kitakyu-shu. The total amount of food waste available forPLLA production was estimated by summing thefood waste generated by companies that annuallydiscarded over 100 tons.

Food waste samples were collected from vari-ous types of food businesses including hotels,restaurants, and hospitals (20 samples from 15sites). After recording the approximate foodcomposition, the samples (12 to 50 kg,4 30.4 kgaverage) were minced in an equal amount of wa-ter using an extruder. We then analyzed the pro-portions of water, d- and l-lactic acids, total sol-uble sugars, glucose, and CHN elements.

Propionic Acid and Lactic AcidFermentation

The minced food waste was autoclaved in a90 L fermentor (Marubishi Bioengineering Co.,Japan) (121C for 20 min) and then inoculatedwith Propionibacterium freundenreichii. The sus-



pension was gently rotated at 60 revolutions/min(rpm) for 12 hr at pH 5.5 (37C) to consumenaturally produced d- and l-lactic acids. Afterthe rst fermentation, the temperature was in-creased to 50C and glucoamylase (Glucoteme,Yanase Sangyo Co., Japan) (20,000 units/g) wasadded to the suspension to obtain a concentra-tion of 300 ppm. Saccharication proceeded for2 hr, and then Lactobacillus rhamnosus KY-3 wasseeded at an initial cell density of 106 cells/mL.The culture (pH 6.5 automatically maintainedusing 7% ammonia) was incubated at 37C for 3or 4 days to complete fermentation.

Purication of L-Lactic Acid

The fermented suspension was passed througha lter press to separate solid residue from thefermented liquid, which contained an average of60 g/L l-lactic acid. The ltrate was concen-trated using a rotary vacuum evaporator (10 L)to yield an average of 40% ammonium lactate.The condensate was placed in a 20 L reactorequipped with an agitator and a distillation col-umn. The internal diameter of the column was500 mm, and the theoretical stage number was5. After n-butanol (3 mol/mol lactic acid) wasadded, the condensate was heated in an oil bathat 150C to 160C and rotated at 80 rpm. Theazeotropic vapor of water and n-butanol was col-lected by a condenser connected to the top ofthe distillation column and separated into waterand n-butanol phases in a separator. The n-butanol, as the upper phase, was reuxed to thetop of the column, whereas the lower water phasewas extracted. The ammonia gas liberated andvaporized in the reactor was collected using anammonia absorber. The reaction mixture was l-tered to remove solid precipitates, and the ltrate(butyl lactate) was evaporated using a rotaryvacuum evaporator. The butyl lactate puried bydistillation was placed with water in the hydro-lyzing vessel and heated to between 95C and110C. The water phase was reuxed to the topof the column, and the n-butanol phase was sepa-rated from l-lactic acid in the vessel.

Polymerization of L-Lactic Acid

The puried l-lactic acid was heated to135C, 150C, and 160C in a stepwise manner

and totally reuxed under a reduced pressure of10 mmHg. Tin isooctate was added as a catalystat a concentration of 30 grams per liter and theresulting lactide (cyclic dimer of lactic acid) wasreactive distillated at 180C and 5 mmHg. Topurify the lactide, four cycles of crystallization at60C to 90C, partial melting at 65C to 95C,and total melting at 70C to 100C proceededusing a melt crystallizer (Sulzer Chemtech Ltd.,Switzerland) equipped with four pipes (70 mminside diameter and 400 cm in length). Tin isooc-tate (0.5% w/v of sample) was added as a catalystto a 50 L reactor (Kansai Chemical EngineeringCo., Hyogo, Japan) to open the ring of the cyclicll-lactide, and the continuous polymerizationwas done using an extruder (S1-KRC, KurimotoLtd. Osaka, Japan) (200C, 15 min retentiontime).

Chemical Analysis

Short-chain fatty acids (formic, acetic, pro-pionic, lactic, and butyric acids) in the super-natant obtained by centrifugation of fermentedfood waste (3,000 rpm, 20 min) were evaluatedusing an organic acid analysis system (Shimadzu,Kyoto). Optical isomers of lactic acid were ana-lyzed by using d- and l-specic lactate dehydro-genase (Roche Diagnostics Inc., Tokyo), accord-ing to the protocol recommended by themanufacturer. The concentration of total solublesugars was analyzed using phenol-sulfuric acid(Dubois et al. 1956), and that of glucose was en-zymatically determined using a glucose oxidasekit (Glucose Test Wako, Wako Chemicals, To-kyo). The elemental carbon and nitrogen com-position in the dried food waste were analyzedusing a CHN analyzer (PE2400 series 2, PerkinElmer Japan).

Results and Discussion

The Process

Figure 1 shows the four steps required to pro-duce PLLA from food waste and the balance ofmaterials obtained. These steps consisted of(1) removal of endogenous d,l-lactic acids fromminced collected food waste by a propionibacte-rium, (2) l-lactic acid fermentation under semi-solid conditions, (3) purication of l-lactic acid



Table 1 Product yields and carbon balance

Content and yield

(kg/kg wet waste) (kg/kg dry waste)Carbon yield

(%)

Dry materiala 0.215 1. Carbon contenta 0.101 0.470 100Total soluble sugara,b 0.143 0.665 Lactic acid in culture ltratec 0.118 0.549 47Puried l-lactic acid c 0.099 0.459 37PLLAc 0.069 0.320 34Fermentation residued 0.14 0.101 27Esterication residued 0.04 0.038 16

a Average of 20 samples from 15 companies.b Average concentration in saccharied samples.c PLLA was experimentally produced from three representative culture ltrate samples. Average yield was calculatedusing efciencies of each step (puried l-lactic acid from culture ltrate, 78.7%; PLLA from puried l-lactic acid,91.9%).d Representative data. Water contents of fermentation residue and esterication residue were 38% and 6.4%,respectively.

by butyl esterication, and (4) polymerization ofl-lactic acid via ll-lactide. The process was ap-plied to food waste collected from commercialkitchens. The yield and composition of collectedfood waste is discussed rst. The specic pointsof each step are then addressed, followed by anevaluation of the entire process from material,energetic, and environmental perspectives.

Amount and Composition of ActualFood Waste

The questionnaire sent to food-related com-panies indicated that about 350 tons/day of foodwaste was discarded in the year 2000 in Kitakyu-shu. The estimated sum of food waste by com-panies that discarded over 100 tons/yr was 62tons/day, which could be used for PLLA produc-tion. The water content of 20 samples from 15different kitchens averaged 78.5% (table 1). Drymatter consisted of 35% vegetable, 23% meatand sh, and 42% cooked carbohydrates, includ-ing rice, bread and noodles. Although the car-bohydrate concentration varied between 28%and 63%, the total amount of soluble sugar afterglucoamylase digestion was 0.14 kg/kg wet waste.Clearly, considerable potential feedstock for lac-tic acid fermentation is being disposed of on adaily basis.

The collected food waste also contained onaverage 1.6 g/kg wet waste of endogenous d- and

l-lactic acids, produced by native lactic acid bac-teria during transfer and storage of the waste.The presence of d-lactic acid decreases the op-tical activity (i.e., proportion of the quantity ofoptical isomers) of accumulated lactic acid anddecreased PLLA crystallinity.

Removal of Endogenous D(L)-Lactic Acidby Propionibacterium

We found that P. freudenreichii preferentiallyconsumes d- and l-lactic acids before sugars as acarbon source under acidic conditions. Figure 2Ashows that lactic acid in the refuse paste wasquickly degraded and diminished within 10 hr atpH 6.5, whereas glucose began to decrease afterseveral hours of lag and was then consumed grad-ually. This was observed more clearly at pH 5.5(gure 2B). The microorganisms assimilate onlylactic acid, and most of the glucose remainedeven after 24 hr. These characteristics of P. freu-denreichii were useful for selective removal of op-tically inactive lactic acid, with little effect onthe amount of sugar availability for subsequentl-lactic acid fermentation.

L-Lactic Acid Fermentation underSemisolid Conditions

After the optically inactive lactic acid wasconsumed, polysaccharides including starch were



Figure 2 Selective degradationof endogenous lactic acid in foodwaste by P. freudenreichii at pH6.5 (A) and pH 5.5 (B). Foodwaste from a university kitchen(500 g) was minced with anequal amount of water, and thepH was adjusted using 7%ammonia. It was then inoculatedwith P. freudenreichii (10 mLpreculture) and was gentlyshaken (60 rpm) at 37C for12 hr. Symbols: circle, lactic acid;square, glucose; triangle, aceticacid; cross, formic acid.

Figure 3 Representativetransients showing L-lactic acidfermentation of food wastecollected from various sources.Food waste samples from ahospital (42.5 kg), a hotel (36 kg),a college (23.7 kg), and auniversity (50 kg) were used.Each sample was minced with anequal amount of water, whichwas autoclaved and treated with0.3 grams per liter glucoamylaseat 50C for 6 hr. It was theninoculated with L. rhamnosus andincubated at 37C. Symbols:open circle, lactic acid; closedcircle, total sugar ; square, glucose.

solubilized by glucoamylase. L. rhamnosus, whichis an l-forming homo-fermentative strain, wasthen inoculated into the treated refuse paste. Asreported previously (Sakai et al. 2000), nutrient-rich food waste appears to be a superior growthmedium for fastidious lactic acid bacteria thatgenerally require a variety of nutritional ele-ments. The amount of water added was mini-mized to reduce energy input for its distillationfrom the fermented broth; an equal proportionof water to refuse was sufcient to yield the high-est productivity of l-lactic acid. The concentra-tion of total sugar after solubilization averaged74 g/L, meaning that the estimated concentra-tion of carbohydrate in the food waste availablefor lactic acid fermentation was estimated to be

143 g/kg wet waste. Figure 3 shows some exam-ples of lactic acid fermentation using the refusecollected from various sources.

Solubilized sugars including glucose seemed tobe involved in the lactic acid fermentation be-cause over 82% of the total was converted to l-lactic acid (an average of 61 g/L), with an aver-age optical purity of over 98% (97.2% to98.9%).5 The amount of accumulated l-lacticacid was dependent on the C/N ratio of the sam-ple, in addition to the amount of total sugar. Re-fuse with an extremely high or low C/N ratio,such as extracted tea residue or sh residues,yielded rather less l-lactic acid (data not shown).Although the amount of accumulated l-lacticacid varied depending upon the source of the



Figure 4 Yield of L-lactic acidfrom food waste generated fromvarious kinds of commercialkitchens. For experimentalconditions, see the discussion inthe section on materials andmethods.

food waste, an average lactic acid yield of morethan 10% per total weight of refuse was con-rmed by semisolid, two-step fermentation (g-ure 4).

Purication of L-Lactic Acid

The ltered and concentrated broth contain-ing approximately 35% l-lactic acid was esteri-ed with n-butanol (150C to 160C) and dis-tilled in the form of butyl lactate (130C, 98%yield), which was then hydrolyzed between 95Cand 110C. Soluble proteins and salts were pre-cipitated with n-butanol. Esters, such as those ofacetic acid and propionic acid, were separatedduring this part of the process. The optical purityof lactic acid did not change during these puri-cation steps. Ammonia stripped at the esteri-cation step was reused to adjust the culture pHat the fermentation step described above. Con-densed water and n-butanol were recycled forsubsequent esterication. This step consumed arelatively high amount of energy (23.4 MJ/kgPLLA) but yielded l-lactic acid of very high pu-rity. Purication of l-lactic acid by butyl esteri-cation is advantageous in that a wastewatertreatment process is not required. In contrast,wastewater treatment is crucial in a conventionalprecipitation using methanol and calcium sulfate(Filachione and Costello 1952).

Polymerization of L-Lactic Acidvia LL-Lactide

The prepolymer was rst synthesized from l-lactic acid by increasing the reaction tempera-

ture in a stepwise manner (135C to 160C, 10mmHg), and the lactide produced was then reac-tively distilled (180C, 5 mmHg). As the lactideobtained through the reactive distillation con-tained a small amount of linear products that in-hibit active polymerization, the ll-lactide wasfurther puried by four steps of crystallizationwith successive partial and total melting (60Cto 100C). Small amounts of the meso-lactide(melting point 52.0C) were separated from thell-lactide (melting point 97.8C) during thisstep. We also separated a racemic eutectic mix-ture of ll- and dd-lactide, which has a meltingpoint of 124.0C. Consequently, ll-lactide athigh optical purity (98.8%) was obtained (95%yield). On the other hand, our preliminary ex-periments showed that one-step synthesis of pre-polymer decreased the optical purity of lactide to88.0%.

Using tin isooctate as a catalyst (Zhang et al.1994), 97.5% optically pure PLLA was synthe-sized at 95.1% yield. The PLLA had an averagemolecular weight of 200 kilodaltons (kDa), amelting point of 175C, and a glass transitiontemperature of 58C (table 2). These physicalproperties are comparable to those of PLLA pro-duced from commercially available l-lactic acid.

Material and Energy Balances forthe Process

The material balance and energy require-ments of the total process are summarized in ta-bles 1 and 3, respectively. Table 1 shows that theoverall experimental process yielded 68.8 g PLLAfrom 1 kg food waste (1.0 kg PLLA/14.6 kg food



Table 2 Characteristics of PLLA produced fromcollected food waste

Optical purity 97.5%Average molecular weight 200 kDaMelting point 175CGlass transition temperature 58C

waste). This means that 34% of total carbon inthe food waste was recovered as PLLA. Theamount of water added to the food waste wasoptimized to generate the highest yield of l-lacticacid, as well as to minimize the size of the fer-mentation facility and the energy input for waterremoval by evaporation during the subsequentprocessing. Furthermore, the process was de-signed to have low environmental impact. Thefermentation residue is rich in nitrogen (C/N6.5; concentrations of N, P, and K were 75, 2.6,and 0.7 mg/g dry matter, respectively), reducedin weight to 14% of that of the untreated foodwaste, and the precipitated residue produced atthe esterication step contains high concentra-tions of phosphorus and potassium (C/N 7.7;concentrations of N, P, and K were 39, 28, and23 mg/g dry matter, respectively). These stableresidues were conrmed to be useful fertilizers(unpublished data). Condensed water, ammonia,and butanol were reused during the process. Theremoved lactide by-product with low optical pu-rity could also be used in molding applicationswhere no crystallinity of the polymer is required.Consequently, nearly all materials are convertedto valuable resources or recycled in the process.

Production of plastics worldwide consumesaround 270 MMT of fossil fuel each year (Gern-gross and Slater 2000), with around 45% of thisused as feedstock; the balance is used as processenergy (Kurdikar et al. 2000). Because petroleumresources are limited and the combustion of fossilfuels releases greenhouse gases with the potentialto change our environment, the use of fossil en-ergy resources is a global issue. Table 3 shows thatabout 44.4 MJ (12.3 kWh) of process energy wasrequired to yield 1 kg of PLLA, supplied mainlyas electricity in these laboratory-scale experi-ments. For comparison, the rst commercialPLLA plant, operated by Cargill Dow Polymers,6

reportedly requires gross fossil process energy7 of39.5 MJ/kg (Vink et al. 2003). Meanwhile, theprocess energy required for production of bottle-grade polyethylene terephthalate (PET) andhigh-density polyethylene (HDPE) using petro-chemicals is 27 MJ/kg (Boustead 2002) and 23MJ/kg (Boustead 2000), respectively.

Although the process energy required forPLLA production is signicantly larger than thatrequired for some petrochemical polymers, it has

the advantage that the energy content of thefeedstock is entirely renewable, resulting in lowertotal fossil energy use. For example, the fossil en-ergy embodied in the feedstock of PET andHDPE is 39 MJ/kg (Boustead 2002) and 49 MJ/kg (Boustead 2000), respectively. Thus thecradle-to-factory-gate fossil energy required forproduction of PET and HDPE is 77 MJ/kg (Bou-stead 2002) and 80 MJ/kg (Boustead 2000), re-spectively. For comparison, the cradle-to-factory-gate fossil energy requirement of the rstcommercial PLLA plant is 54.1 MJ/kg (Vink etal. 2003). As currently practiced by Cargill Dow,the cradle-to-factory-gate polylactide productionsystem uses 20% to 50% fewer fossil resourcesthan required for competing petrochemical poly-mers (Slater et al. 2003).

The PLLA process proposed here has an en-ergy advantage over even the Cargill Dow poly-lactide process, because the feedstock is a wastestream. In the Cargill Dow process, nearly 30%of gross fossil energy use goes into producing andprocessing corn to provide dextrose to feed thelactic acid fermentation. Because the feedstockto the proposed PLLA process is food waste thatmust otherwise be disposed of, the only upstreamfossil energy allocated to the production of PLLAwould be that required for collection of the sepa-rated waste (approximately 2 MJ/kg in Kitakyu-shu City).

Because polylactide production technology isin its infancy, and the PLLA process proposedhere is still in laboratory development, there aremany opportunities for substantial reductions infossil energy use. Themost energy-intensive stepsin our process are the concentration, esterica-tion, and distillation of l-lactic acid (8.04 kWh/kg PLLA; table 3). These are the key steps re-quired for separating l-lactic acid from the com-plex materials in refuse and are appropriate tar-gets for efciency improvements.



Table 3 Energies required at each step of PLLA production from food waste

StepElectricity consumeda

(kWh/kg PLLA)Energytype

Energy in MJb

(MJ/kg PLLA)Temperature

(C)Alternativeenergy type

Fermentation/ltration 2.54 Electricity 9.1 125 Heat/electricityEvaporation 3.25 Electricity 11.7 80 HeatEsterication 3.08 Electricity 14.1 150160 HeatDeesterication 1.71 Electricity 6.2 95110 HeatPolymerization 1.71 Electricity 6.2 60200 Heat

Total 12.33 44.4

a PLLA yield was 68.8 g/kg wet waste.b Calculated from electricity (kWh/kg PLLA) 3.6 MJ/kWh.

As table 4 shows, the amount of fossil fuelrequired by our process could be improved bysubstituting steam heat for electricity in manystages of the process, because most of the energyrequired is heat below 150C (table 3). For ex-ample, if 80% of the total energy were suppliedby steam rather than electricity, the process en-ergy use would be reduced from 3.53 kg FossilFuel Equivalent (FFE)/kg of PLLA to 1.42 kgFFE/kg.8

As in current commercial operations, it is pro-ducing lactic acid and PLLA that is the mostenergy-intensive part of our process, and thusprocess energy is expected to dominate the en-vironmental impact of the overall scheme. Oneway to reduce this impact is to use a renewableform of process energy. This can be done by gen-erating electricity and heat energy through com-bustion of the nonfood waste portion of theMSW stream. Because over 60 tons/day of refusewould be available for PLLA production in Ki-takyushu City, 18,000 tons/yr of food waste couldbe treated to produce 1.2 103 tons of PLLAper year, requiring 1.5 107 kWh of electricity.9

Using the data of Harrison and colleagues(2000), the rate of electricity generation fromMSW combustion that contains 4.9% food wasteis 0.584 kWh/kg wet waste.10 If the electricity forPLLA production (20% of total energy) is pro-vided by combusting such MSW, then 5,100 tonsof wet waste11 would be required. This gure cor-responds to around 1/50th of the MSW burnedannually in Kitakyushu City (data not shown).The remaining 80% of the energy required forPLLA production could be recovered as wasteheat from the MSW combustion. Thus, the solid

waste stream could easily provide the feedstockas well as the electricity and heat energy requiredfor PLLA production from food waste. In addi-tion, closely coupling PLLA production withwaste heat utilization during MSW combustionwould be benecial because it would provide avaluable use for low-quality waste heat from thecombustion process and reduce total infrastruc-ture requirements.

If process energy were derived from MSWcombustion, it might be that from a greenhousegas perspective PLLA would be preferable to pet-rochemical polymers such as polyethylene (PE)only because of the use of biomass power. A simi-lar result was found for polyhydroxyalkanoate(PHA) produced using biomass power (Kurdikaret al. 2000). This seems quite possible becausePLLA production requires substantially moreprocess energy than many comparable petro-chemical polymers. Such life-cycle comparisonsof petrochemical and biobased plastics are alsofrequently quite sensitive to assumptions regard-ing end-of-life disposition (Patel et al. 2003). Forexample, PE that is landlled sequesters carbon,but if combusted, it releases its fossil feedstockcarbon into the atmosphere. In contrast, PLLAbiodegrades when landlled or composted, butthe quantity and form of this carbon release isboth variable and uncertain. A life-cycle assess-ment of the proposed PLLA process addressingthese issues is now in progress, and results will bereported elsewhere.

In conclusion, the innovative practical pro-cess described here can produce high-qualityPLLA from food waste. The system was designedas a total material recycling process for municipal



Table 4 Fossil fuel required in producing PLLAfrom food waste

Form of energyFossil fuel required(kg FFE/kg PLLA)

100% electricitya 3.5380% heat, 20% electricityb 1.42

a Using the factor of U.S. average for producing 1 kWhelectrical power described by Gerngross (1999): 12.33kWh/kg PLLA 0.272 kg FFE/kWh 3.53 kg FFE/kgPLLA.b Using the factor of heat energy from natural gas citedby Gerngross (1999): (47.2 MJ/kg FFE)/(3.6 MJ/kWh) 0.0763 kg FFE/kWh, 12.33 kWh/kg PLLA(0.0763 0.8 0.272 0.2) 1.42 kg FFE/kgPLLA.

food waste, with minimal environmental emis-sions and energy savings. It also has the potentialto produce from MSW a valuable, renewableproduct that can substitute for currently pro-duced nonrenewable, petrochemical polymers.This system would also be applicable to the pro-duction of PLLA from other waste streams thatcontain fermentable sugars with a high propor-tion of water, such as agricultural wastes.

Acknowledgments

This study was supported by Special Coordi-nation Funds of the Ministry of Education, Cul-ture, Sports, Science, and Technology of the Jap-anese Government.

Notes

1. One metric ton 1 Mg (SI) 1.102 short tons.Unless otherwise noted, all tons in this articlerefer to metric tons.

2. Biological waste in this context refers to the por-tion of MSW composed of food waste, yard waste,and, in some cases, paper and related biodegrad-able materials. It is often labeled organic wastein the United States and biowaste in Europe.

3. Editors note: Poly(lactide) or PLA means apolymer derived from the condensation of lacticacid or by the ring opening polymerization of lac-tide. The terms lactide and lactate are usedinterchangeably. Polylactide or polylactate is acommercial, biodegradable polymer that is usedfor a variety of packaging, medical and other ma-

terials applications. Lactic acid, lactide and poly-lactide all occur as chiral molecules. Chiral mol-ecules are molecules that are so asymmetric thatthey are nonsuperimposable on their mirror im-ages. That is, they have handedness in the waygloves are left- or right-handed. Lactide occurs inthree forms: the two chiral isomers of lactide, l-lactide and d-lactide, and an achiral form knownas meso-lactide. The predominant stereoisomer inthe lactide polymer product is l-lactide. Thus, theterms polylactide or PLA used in the articlein this issue by Gruber (2003) and poly-l-lactate or PLLA used in this article are bothused to describe the same material.

4. One kilogram 2.204 lb.5. The degree of stereochemical (optical) purity is

important for commercial applications becausethe properties of PLA such as melting point, me-chanical strength, and crystallinity, are deter-mined by the different proportions of l-, d-, ormeso-lactide in the polymer (as well as its molec-ular mass). PLA resins with 5093% content ofl-lactide are amorphous (non-crystalline), whileresins containing over 93% l-lactide are semi-crystalline. PLA that is too crystalline however,can be brittle, and so a small number of stereo-defects can improve the polymers mechanicalproperties.

6. Editors note: For a description of the Cargill Dowjoint venture in the Journal of Industrial Ecology,see the corporate prole by Gruber (2003).

7. This gross process energy term includes energyused in the fermentation and polymer plant andalso the energy to run the wastewater treatmentplant and produce and deliver various suppliessuch as nitrogen and water. It also includes theenergy required to produce the process energyused in the PLLA production process. The pro-cess energy accounting for the PLLA process de-scribed in this article is not as complete but doesinclude the largest energy requirements.

8. This calculation is based on the conversion factorof Gerngross (1999), which assumes steam gen-erated by ideal combustion of natural gas with noheat loss.

9. PLLA production and associated electricity re-quirements are calculated as follows: 1.2 103

ton PLLA/yr ([18,000 ton/yr]/[14.6 kg waste/kg PLLA]); 1.5 107 kWh 12.3 kWh/kgPLLA 1,200 tons of PLLA.

10. Note that 0.584 kWh/kg 10,512 (Btu/kg wetwaste)/18,000 (Btu/kWh); 1 Btu 0.00106 MJ.

11. 5,100 tons 0.2 1.5 107(kWh)/0.584(kWh/kg wet waste).



References

Aksu, Z. and T. Kutsal. 1986. Lactic acid productionfrom molasses utilizing Lactobacillus delbruekii andinvertase together. Biotechnological Letters 8: 157160.

Boustead, I. 2000. Eco-proles of the plastics and relatedintermediates (about 55 products). Prepared forthe Association of Plastics Manufacturers in Eu-rope (APME), Brussels. www.apme.org. Ac-cessed October 2003.

Boustead, I. 2002. Eco-proles of the European plasticsindustry: Polyethylene terephthalate. Prepared forthe European Centre for Plastics in the Environ-ment, Brussels. www.apme.org. Accessed Oc-tober 2003.

Camobreco, V., R. Ham, M. Barlaz, E. Repa, M. Felker,C. Rousseau, and J. Rathle. 1999. Life-cycle in-ventory of a modern municipal solid waste land-ll, Waste Management and Research 17, 394408.

Dubois, M., K. A. Gilles, J. A. Hamilton, P. A. Rebers,and F. Smith. 1956. Colorimetric method for de-termination on sugars and related substances.An-alytical Chemistry 28, 350356.

Filachione, E. M. and E. J. Costello. 1952. Lactic estersby reaction of ammonium lactate with alcohols.Industrial and Engineering Chemistry 44, 21892191.

Food Recycling Law. 2000. Ministry of Agriculture,Forestry, and Fisheries of Japan. www.maff.go.jp/eindex.html. Accessed October 2003.

Gerngross, T. U. 1999. Can biotechnology move ustoward a sustainable society? Nature Biotechnology17, 541544.

Gerngross, T. U. and S. C. Slater. 2000. How green aregreen plastics? Scientic American 282(2): 2429.

Gruber, P. R. 2003. Cargill Dow LLC. Journal of In-dustrial Ecology 7(34): 209213.

Harrison, K. W., R. D. Dumas, M. A. Barlaz, and S. R.Nishtala. 2000. A life-cycle inventory model ofmunicipal solid waste combustion. Journal of theAir and Waste Management Association 50, 9931104.

Kurdikar, D., L. Fournet, S. C. Slater, M. Paster, T. U.Gerngross, and R. Coulon. 2000. Greenhouse gasprole of a plastic material derived from a genet-ically modied plant. Journal of Industrial Ecology4(3): 107122.

Litcheld, J. H. 1996. Microbiological production oflactic acid, Advances in Applied Microbiology 42,4595.

Lunt, J. 1998. Large-scale production, properties, andcommercial application of polylactic acid poly-mers. Polymer Degradation and Stability 59, 145152.

Miyazaki, H. and K. Harano. 2001. Degradation test

on biodegradable plastics polymers into compost-ing. In 2000 Annual Research Report of Oita In-dustrial Research Institute. Ono-Kousoku, Oita, Ja-pan: Oita Industrial Research Institute, 154157.

Oda, Y., B. Park, K. Moon, and K. Tonomura. 1997.Recycling of bakery wastes using an amylolyticlactic acid bacterium. Bioresource Technology 60,101106.

Ohara, H. and S. Sawa. 1994. Purication process ofl-lactic acid for bio-degradable poly-l-lactate. InBiodegradable plastics and polymers, edited by Y.Doi and K. Fukuda. Amsterdam: Elsevier.

Patel, M., C. Bastioli, L. Marini, and E. Wurdinger.2003. Life-cycle assessment of bio-based polymersand natural ber composites. In General aspectsand special applications. Biopolymers, Vol. 10, ed-ited by A. Steinbuchel. Weinheim, Germany:Wiley.

Sakai, K., H. Kawano, A. Iwami, M. Nakamura, andM. Moriguchi. 2001. Isolation of a thermophilicpoly-l-lactide degrading bacterium from compostand its enzymatic characterization. Journal of Bio-science and Bioengineering 92, 298300.

Sakai, K., Y. Murata, H. Yamazumi, Y. Tau, M. Mori,M. Moriguchi, and Y. Shirai. 2000. Selective pro-liferation of lactic acid bacteria and accumulationof lactic acid during an open fermentation of foodwaste with intermittent pH adjustment. Food Sci-ence and Technology Research 6, 140145.

Schmidt, S. and N. Padukone. 1997. Production of lac-tic acid from waste waterpaper as a cellulosicfeedstock. Journal of Industrial Microbiology andBiotechnology 18, 1014.

Slater, S., D. Glassner, E. Vink, and T. Gerngross.2003. Evaluating the environmental impact ofbiopolymers. In General aspects and special appli-cations. Biopolymers, Vol. 10, edited by A. Stein-buchel. Weinheim, Germany: Wiley.

Vink, E. T. H., K. R. Rabago, D. A. Glassner, and P. R.Gruber. 2003. Applications of life cycle assess-ment to NatureWorks polylactide (PLA) produc-tion. Polymer Degradation and Stability 80: 403419.

Xavier, S. and B. K. Lonsane. 1994. Sugar-cane press-mud as a novel and inexpensive substrate for pro-duction of lactic acid in a solid-state fermentationsystem. Applied Biochemistry and Biotechnology 41:291295.

Zhang, X., D. A. Macdonald, M. F. A. Goosen, andK. B. Mcauley. 1994. Mechanism of lactide po-lymerization in the presence of stannous octoate:The effect of hydroxy and carboxylic acid sub-



stances. Journal of Polymer Science, Part A: Poly-mer Chemistry 32, 29652970.

Zhou, S. D., T. A. McCaskey, and J. Broder. 1996.Evaluation of nitrogen supplements for biocon-version of municipal solid waste to lactic acid.Applied Biochemistry and Biotechnology 57/58:517524.

About the Authors

Kenji Sakai is an associate professor in the De-partment of Applied Chemistry, Faculty of Engineer-ing, Oita University, Oita, Japan. Masayuki Taniguchi

is a professor in the Department of Materials Scienceand Technology, Faculty of Engineering, Niigata Uni-versity, Niigata, Japan. Shigenobu Miura is a managerin the research and development department of Mu-sashino Chemical Laboratory, LTD, Tokyo, Japan. Hi-tomi Ohara is a group manager in the Toyota Biotech-nology and Afforestation Laboratory, Toyota MotorCorporation, Aichi, Japan. Toru Matsumoto is an as-sociate professor in the Department of EnvironmentalSpace Design, Faculty of Environmental Engineering,University of Kitakyushu, Fukuoka, Japan. YoshihitoShirai is a professor in the Graduate School of LifeScience and Systems Engineering, Kyushu Institute ofTechnology, Fukuoka, Japan.

Documents

Making Plastic from the garbage