What moves from packaging into cocoa and chocolate, and how much packaging matters to chocolate's environmental footprint. Several of these studies are old, and the materials they tested may no longer be in use.
13 studies1989 to 2025Each checked against its published abstract
What these studies add up to
Mineral oil reaches cocoa before it reaches a factory. The oil used to soften jute fibre was found in bagged foods in 1991, and in 2025 beans shipped in contaminated jute gained tenfold in saturated hydrocarbons, some of which carried through into cocoa mass, butter and powder.
Recycled cardboard is a second source, and the film between board and food decides how much arrives: polyethylene hardly slowed it, polyester stopped it.
Substances printed on the outside of a pack can end up in the food. Plasticisers from wrapper ink and an ink component from cartonboard were both measured in chocolate products, in studies from 1989 and 2003.
Chocolate picks up taint: tasters detected styrene from polystyrene held against chocolate flakes.
Life cycle assessments agree that ingredients and manufacture dominate chocolate's footprint, with packaging behind them. Transport's share varied fivefold with the route.
Most multilayer flexible films are not recycled with existing infrastructure, and a biodegradable alternative scored worse on life cycle impact while scoring better on marine litter.
Plasticisers from the ink on the outside of a wrapper turned up in the confectionery
Can components of ink printed on the outside of a film reach the food inside?
How
Experimental storage of a tightly wound reel of specially printed polypropylene film, storage trials of wrapped products, and a small retail survey (47 samples) of confectionery, snack products and biscuits wrapped in printed polypropylene film.
Finding
On the reel, transfer of dicyclohexyl phthalate from the printed outer surface to the food contact surface amounted to 6% of the plasticiser available in the ink. In a chocolate-coated confectionery product, dibutyl phthalate increased from 0.2 to 6.7 mg/kg over 0 to 180 days of storage. In the survey, dibutyl phthalate was found at 0.02 to 14.1 mg/kg.
Does not establish: It does not describe inks or wrappers in use now, and the abstract makes no comparison with a safety limit.
Limits: Products and ink formulations of the 1980s, which may not represent inks in use now. The abstract does not name the country.
Can polystyrene packaging taint cocoa powder drinks and chocolate flakes at levels people can taste?
How
Sensory difference and styrene recognition tests on cocoa powder for drinks and on plain and milk chocolate flakes exposed to polystyrene, with measurement of styrene in the foods. Panel size is not given.
Finding
No differences were observed in cocoa powder for drinks and plain chocolate flakes treated with 0.5 dm2 polystyrene of 1 mm thickness. Differences were detected in milk chocolate flakes and plain chocolate flakes in contact with a larger area or thicker polystyrene. Styrene ranged from 7 to 132 ppb in cocoa drinks and from 414 to 1447 ppb in chocolate flakes.
Does not establish: It does not show that polystyrene packs on sale today taint chocolate, and it is a flavour finding, not a safety assessment.
Limits: An experimental exposure using polystyrene with about 320 ppm residual styrene; an old study, and material specifications may have changed.
Jute sacks treated with mineral oil contaminated the foods carried in them
Grob and colleagues, 1991. Journal - Association of Official Analytical Chemists. doi:10.1093/jaoac/74.3.506
Question
Does the mineral oil used to soften jute fibres reach foods transported in jute sacks, and how can it be measured?
How
Method paper describing automated on-line LC-GC for hydrocarbons in foods, with results for nuts, coffee, cocoa products and rice. The abstract gives no sample numbers.
Finding
Jute fibres are treated with about 5-7% of a high boiling mineral oil fraction, called batching oil. Foods transported in jute bags were contaminated by this oil, and contamination ranged between about 5 and 500 ppm.
Does not establish: It does not describe contamination levels in cocoa shipped today, and it does not separate saturated from aromatic hydrocarbons.
Limits: Early method paper with unstated sample numbers. It predates later changes to jute bag specifications and modern MOSH and MOAH analysis.
Batching oil from jute and sisal bags reached cocoa beans, hazelnuts and almonds
Grob and colleagues, 1993. Zeitschrift für Lebensmittel-Untersuchung und -Forschung. doi:10.1007/bf01242063
Question
Are hazelnuts and chocolate contaminated by mineral oil from jute and sisal bags?
How
Analytical survey; the English abstract gives no sample numbers or methods.
Finding
Jute and sisal fibres are treated before spinning with a batching oil commonly consisting of a raw mineral oil fraction, which is transferred to foods packed in such bags, primarily cocoa beans, hazelnuts, coffee, almonds, oil seeds and rice. Concentrations in the foods easily exceed 100 mg/kg.
Does not establish: It does not show present-day levels, and the abstract does not report measured values in finished chocolate.
Limits: A short abstract from 1993 with no sample numbers; the article is in German.
How much benzophenone from UV-cured inks and recycled fibre is present in cartonboard packaging, and what governs its migration into food?
How
Method development and survey of 350 retail samples in printed cartonboard packaging. 71 samples were selected at random from the 143 packaging samples that contained benzophenone, and the food was analysed by GC-MS.
Finding
A total of 207 (59%) packaging samples had no significant benzophenone. Benzophenone was detected in 51 (72%) of the 71 foods analysed, and three (4%) exceeded 5 mg/kg. The highest level was 7.3 mg/kg, in a high-fat chocolate confectionery product packed in direct contact with cartonboard, stored at room temperature and with a high contact area to food mass ratio. Indirect contact gave a sixfold reduction in migration compared with direct contact.
Does not establish: It does not show current benzophenone levels in chocolate packaging, and the abstract makes no comparison with a legal limit.
Limits: A retail survey published in 2003, which may not represent current ink use. Foods were analysed only where the packaging contained benzophenone.
How much mineral oil, photoinitiator and plasticiser migrates from recycled paperboard into dry foods, and which inner films stop it?
How
Controlled storage study: six test foods industrially packed in five configurations, four with internal plastic films, monitored after 2, 4 and 9 months.
Finding
After 9 months, migration of mineral oil saturated hydrocarbons into foods packed directly in the paperboard amounted to 30-52 mg/kg, and migrated aromatic hydrocarbons ranged from 5.5 to 9.4 mg/kg. More than half of this migration occurred in the first 2 months. With oriented polypropylene the highest migration of saturated hydrocarbons after 9 months was 2.3 mg/kg. Polyethylene terephthalate was a tight barrier.
Does not establish: It does not measure migration into chocolate, and barrier performance of a film type will vary with thickness and grade.
Limits: The six test foods were dry foods; the abstract does not say that chocolate was among them. One paperboard was used.
What are the life cycle environmental impacts of chocolate made and consumed in the United Kingdom, and where do they arise?
How
Life cycle assessment using the ReCiPe method, plus water footprint, of three product types described as occupying 90% of the market: moulded chocolate, chocolate countlines and chocolates in bag.
Finding
Global warming potential ranged between 2.91 and 4.15 kg CO2 eq. per kilogram of chocolate. Raw materials were the major hotspot across all impact categories for all three product types, followed by the chocolate production process and packaging. Including land-use change associated with cocoa increased the global warming impact by three to four times.
Does not establish: It does not show that packaging choices are unimportant for litter or recycling, which this kind of assessment does not measure, and it gives no packaging percentage in the abstract.
Limits: A modelled assessment that depends on its data sources and assumptions; the abstract does not give the percentage share due to packaging.
Can long-distance transport cancel the lower footprint of organically grown cacao?
How
Cradle-to-retailer life cycle assessment of 1 kg of dark chocolate made with Ecuadorian cacao, comparing three farm management systems and three supply chain scenarios with different transport weights.
Finding
The carbon footprint was estimated at between 2.04 and 4.66 kg CO2-eq per kg. The relative weight of transportation in total greenhouse gas emissions ranged between 8.9% and 51.1%. Chocolate from organic agroforestry cacao had a footprint 22.7%-34.2% lower than that from conventional monoculture cacao manufactured and transported under the same conditions.
Does not establish: It does not give a general transport share for chocolate, and the abstract does not report packaging separately.
Limits: Scenario modelling for one origin; the transport share depends heavily on the modes and distances assumed.
What stands in the way of making multilayer flexible food packaging recyclable?
How
Narrative review of multilayer flexible food packaging, recycling infrastructure and design-for-recycling guidelines. No systematic search is described in the abstract.
Finding
Most multilayer films are not sorted and recycled with the currently available infrastructure, which is based on mechanical recycling in most countries. Recyclable designs are expected to be based on polyolefins with a few barrier layers such as aluminium oxide and silicon oxide. In the worst case, restricting material choice could result in a higher environmental burden through shortened food shelf life and increased packaging weights.
Does not establish: It gives no figure for how much flexible confectionery packaging is recycled, and it does not test any chocolate wrapper.
Limits: A narrative review; it reports no recycling rate, and the worst-case trade-off is the authors' assessment, not a measurement.
How much mineral oil and polyolefin oligomer contamination is in retail cocoa powder, and does the packaging contribute?
How
Validation of a rapid, high-sensitivity method for cocoa powder, applied to products bought in Italy and North Macedonia, with analysis of the related primary and secondary packaging. The abstract does not give the number of products.
Finding
In addition to pre-existing contamination, oligomeric polyolefin hydrocarbons from primary plastic packaging and MOSH/MOAH from recycled cardboard secondary packaging were found in many samples. Total MOSH/POH ranged from 1.2 to 69.2 mg/kg (on average 20.9 mg/kg), exceeding the German benchmark level of 9 mg/kg in half of the samples from Italy. Most of the samples had MOAH >0.5 mg/kg.
Does not establish: It does not apportion the contamination between jute, processing and packaging for any one product, and it does not assess risk.
Limits: A market survey in two countries with an unstated number of samples. The total figure combines mineral oil with polyolefin oligomers from plastic, which are different substances.
How does a biobased biodegradable food tray compare with a conventional plastic one when production, food preservation, end of life and marine litter are all counted?
How
Life cycle assessment of falafel packaged in PHBV and in conventional polypropylene trays, extended with a new indicator for lifetime costs on marine ecosystem services.
Finding
Per kilogram of packaged falafel consumed, food ingredients made the largest contribution. The results indicated a clear preference for polypropylene trays, mainly because of the higher mass and volume of the alternative tray. Because PHBV is less persistent, its lifetime costs for marine ecosystem services were about seven times lower.
Does not establish: It does not show that biodegradable packaging is worse or better in general, and it says nothing about barrier performance for chocolate.
Limits: One food, one pair of trays, and an indicator the authors say needs further refinement. The food is not chocolate.
Across published life cycle assessments, which stages of the chocolate life cycle carry the environmental burden?
How
Systematic review that collected 25 articles on life cycle assessment of the chocolate industry.
Finding
The raw material production and chocolate manufacturing stages accounted for 77-97% of total impacts, with milk powder, sugar and cocoa derivatives the important contributors. Dark chocolate had the lowest reported carbon emissions (1.67 kg CO2 eq/kg product) and chocolate confectionery products the highest (6.76 kg CO2 eq/kg product).
Does not establish: The abstract does not isolate packaging's share, so it cannot be read as a measured packaging percentage; packaging falls within the remainder along with distribution and other stages.
Limits: The underlying studies use different system boundaries, functional units and data, so the figures are not strictly comparable.
How much mineral oil do cocoa beans take up from jute bags during storage and transport, and where does it go in processing?
How
On-line HPLC-GC analysis of 28 whole cocoa bean samples before and after transport in jute bags contaminated with mineral oil hydrocarbons, analysis of the bags, and follow-up through the supply chain into cocoa mass, butter and powder.
Finding
Average MOSH increased from 0.9 to 9.6 mg/kg, with MOAH reaching 2.3 mg/kg. On average 40 % of the hydrocarbons that migrated into whole beans reached the shelled beans. The authors estimated that MOAH in the n-C10-25 range exceeding 220 mg/kg in new jute bags expose whole beans to contamination above 2 mg/kg. Contamination carried into cocoa mass and partitioned into butter and powder according to fat content.
Does not establish: It does not show that all cocoa shipped in jute is contaminated to this degree, and it makes no statement about health risk.
Limits: 28 samples; the abstract does not give origins or routes, and the bags studied were ones known to be contaminated.
We would like to set two Google Analytics cookies to count page views, and to count them in our own cookie-free measurement as well. They do not identify you and nothing is shared with advertisers. Decline and the site works exactly the same — see cookies for what each one does.