Industries / Consumer goods / Candles and home fragrance
Fragrance
Fragrance work for the home is choosing and blending scent oils so they mix into a wax or liquid, fill a room when heated or left to evaporate, and stay within industry safety limits and labelling law.
6
Milestones achieved, from 1973 to 2019.
2
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1973
IFRA sets shared limits
The International Fragrance Association (IFRA) was founded in 1973 to represent the fragrance industry. Its Standards ban, limit or set purity rules for single fragrance ingredients, based on safety assessments by the Research Institute for Fragrance Materials (RIFM), and they are mandatory for its members; with the 51st Amendment in 2023 there were 263 of them, by IFRA's count. Candle and diffuser makers get an IFRA certificate from their oil supplier that states the most of each oil they may use in each type of product.
LessonThe main safety limits for home fragrance are set by the industry itself, not by a government.
Credit
International Fragrance Association, Geneva; Research Institute for Fragrance Materials
2006
Scent chemistry continues in rooms
A 2004 review by William Nazaroff and Charles Weschler set out how terpenes, the citrus and pine molecules common in scent oils, react with ozone indoors to form new pollutants such as formaldehyde. In 2006 a Lawrence Berkeley National Laboratory team ran a plug-in scented-oil air freshener for several days in a 50 m³ chamber set up as a room: with ozone present, product use raised formaldehyde by about 10 ppb (parts per billion), and cleaning products used with ozone sent fine particles above 100 µg/m³ in some runs.
LessonWhat a fragrance adds to a room depends on the room's air, not only on the oil.
Credit
William Nazaroff (UC Berkeley) and Charles Weschler (review); Brett Singer, Hugo Destaillats, Alfred Hodgson and colleagues, Lawrence Berkeley National Laboratory (chamber tests)
2009
Throw becomes a lab number
Cold throw is how strongly an unlit candle smells, and hot throw is how strongly it smells while burning. A Givaudan patent, filed in 2005 and granted in 2009, gave each scent molecule a numerical cold throw and hot throw value, measured by sampling the air above test candles and analysing it by gas chromatography-mass spectrometry (GC-MS), which separates and identifies the molecules, and chose molecules by those values.
LessonLarge fragrance houses measure throw with lab instruments, but no shared public method followed.
Credit
Givaudan Fragrances Corp. (inventors Addi Fadel, Richard Turk, Grant Mudge, Jill Mattila and Robert Burke)
2017
California air care discloses allergens
California's Cleaning Product Right to Know Act (SB 258) covers air care products as well as cleaners. It requires makers to disclose listed fragrance allergens present at or above 100 ppm (0.01%), using a list compiled under the EU Cosmetics Regulation, online from 1 January 2020 and on labels from 1 January 2021.
LessonDisclosure stops at a threshold, so allergens below it can still be listed only as "fragrance".
Credit
California Legislature
2019
Swiss lab checks candle labels
Under European chemical labelling rules (CLP), which Switzerland applies through its own law, a candle counts as a mixture carried on a wick, so a scented candle needs an allergen warning such as "Contains (name). May produce an allergic reaction" when a skin-sensitising fragrance passes set levels. In spring 2018 the Canton of Thurgau collected 33 scented candles, wax melts and incense products. Lab analysis found total allergen contents of 0.0 to 5.2% (median 0.6%), and the Swiss Federal Office of Public Health reported in February 2019 that about one fifth failed the labelling rules, in every case by missing the allergen warning. Only 17 of the 33 sellers supplied a safety data sheet.
LessonLabels lag what the lab finds in the wax.
Credit
Chemicals office of the Canton of Thurgau; Swiss Federal Office of Public Health
2019
Candles get their own category
IFRA notified its 49th Amendment on 12 December 2019, moving to a method that adds up a person's skin exposure across products. Candles and wax melts moved from Category 11 to Category 12, where, because people barely touch them, the Standards do not limit fragrance concentration in the finished product. Reed diffusers went to the stricter Category 10A, because people flip the soaked reeds by hand. The 51st Amendment of June 2023, whose limits applied to existing fragrances from 30 October 2025, kept this layout. In practice the wax sets the fragrance load, about 6 to 10% in candles by the US supplier CandleScience's figure.
LessonThe category depends on how people handle a product, not on how it smells.
Credit
IFRA and RIFM
Watch for
The 52nd IFRA amendment
IFRA says its 52nd Amendment will add 48 new restriction Standards for skin sensitisation and revise 15 more, with formal notice expected in January 2027, and that the new limits will apply 11 months after notice for new fragrances and 30 months after for existing ones. Its 2023 guidance said the amendment would include a full review of Categories 10, 11 and 12, where diffusers, sprays and candles sit; its 2026 note already calls reed diffusers "Category 10B products".
Achieved whenthe Notification is published and suppliers' IFRA certificates for candle and diffuser oils show the new limits and categories.
Claimed by
IFRA, September 2026
AI-formulated focus scents
Aera, the home diffuser brand of Prolitec, launched three scents built with FocusMax, which it calls "an AI-powered formulation platform" drawing on 40,000 fragrances and over 1 million consumer interviews, and says users did better on attention tasks with them. The report names no test lab and no sample size.
Achieved whenan independent, blinded test with a named lab, a stated sample size and a placebo scent repeats the attention effect.
Claimed by
Aera (Prolitec), March 2026
Open
01
Instrument score for scent throw
A low-cost sensor box and a published protocol that turn a candle's hot and cold throw into a number, calibrated against a trained human sniff panel, so that small makers can compare waxes, wicks and fragrance loads without a panel each time.
The briefCredit · unclaimed02
Allergen label-versus-lab audit
A public, product-by-product table for air fresheners and scented candles that sets the fragrance allergens each maker discloses (on California SB 258 web pages, or in the EU allergen warning on the label) against what an accredited lab measures in the same product.
The briefCredit · unclaimed
Open 01 · the brief
Instrument score for scent throw
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A low-cost sensor box and a published protocol that turn a candle's hot and cold throw into a number, calibrated against a trained human sniff panel, so that small makers can compare waxes, wicks and fragrance loads without a panel each time.
Why it is still open
Suppliers still judge throw by smell: CandleScience evaluates hot throw "after burning the candle for at least two hours and forming a full melt pool". Research groups measure terpenes from candles and wax melts with proton-transfer-reaction mass spectrometers (PTR-TOF-MS), far beyond a small maker's budget, and do not link the readings to how strong people find the smell. The one numerical throw method found is in Givaudan's patent.
Built with what exists
An Ion Science MiniPID 2 HS photoionisation sensor and a Sensirion SGP41 gas sensor on an ESP32 or Raspberry Pi logger; a Sensirion SCD41 carbon dioxide sensor to measure the room's air exchange from CO2 decay; a sniff panel of 10 or more people scoring against the ASTM E544 odour intensity scale, which compares a smell with set strengths of n-butanol; Python with pandas and scikit-learn. Candles are burned attended in a ventilated room.
First step
Build one box and burn one unscented and one scented candle from the same batch three times each in a room with measured air exchange, then post the raw logs.
Done looks like
A public protocol and dataset for at least 30 candles (for example 3 waxes and 10 oils at the same load), each burned 3 times, with panel scores and sensor traces; the rank correlation between sensor score and panel score reported on a held-out set of at least 10 candles, along with the burn-to-burn spread of the sensor score.
Open 02 · the brief
Allergen label-versus-lab audit
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public, product-by-product table for air fresheners and scented candles that sets the fragrance allergens each maker discloses (on California SB 258 web pages, or in the EU allergen warning on the label) against what an accredited lab measures in the same product.
Why it is still open
Checks so far are one-off and do not publish product-level matches. The Swiss check of 2018-2019 covered 33 products in one canton and reported totals. Steinemann and colleagues found in 2011 that of 133 volatile organic compounds given off by 25 fragranced products, "only 1 was listed on any product label", but that came before SB 258's allergen disclosures. Searches of GitHub, Zenodo, Hugging Face and Kaggle found cosmetic ingredient datasets but no air-care lab-versus-label data.
Built with what exists
A contract lab accredited for EN 16274:2021, the European GC-MS method for 24 suspected fragrance allergens, with wax dissolved in hexane first as in the Swiss method; Python with requests and BeautifulSoup to collect disclosures; a small low-cost AI model such as Meta's Muse Ember to pull allergen lists out of disclosure pages and safety data sheets into one table (see the Note "The jobs agents repeat"); Zenodo for the data and lab reports.
First step
Record the SB 258 web disclosures for 40 air-care products sold in California in a spreadsheet, and get quotes from two accredited labs.
Done looks like
At least 40 products with disclosed allergen lists captured and at least 20 of them tested by the lab; the number and share with any undisclosed listed allergen at or above 100 ppm (0.01%), with every lab report linked.