Clove

What is Clove?

A kilogram of clove buds must yield at least 150 mL of essential oil to satisfy the European Pharmacopoeia, and roughly four fifths of that oil is a single molecule, eugenol. Few culinary spices concentrate one compound that hard. Clove is a dried unopened flower bud, about 15 to 17 percent volatile oil by volume, and that oil is 75.0 to 88.0 percent eugenol by the Pharmacopoeia's own limits. The spice, the dental anesthetic and the poison are the same chemistry at three different concentrations.

The tree is an evergreen in the myrtle family, native to the Maluku Islands of Indonesia and now grown commercially in Indonesia, Madagascar, Tanzania, Sri Lanka and Brazil. Three separate essential oils come off it and they are not interchangeable. Bud oil is distilled from the flower buds, leaf oil from the leaves, stem oil from the flower stalks. The International Organization for Standardization publishes a separate standard for each, ISO 3141:1997 for leaf, ISO 3142:1997 for bud and ISO 3143:1997 for stem, and FDA lists all three plus eugenol itself as separate substances in 21 CFR 184.1257.

Common Names

• Clove, cloves, clove bud

• Syzygium aromaticum (L.) Merr. & L.M. Perry, the accepted binomial

• Eugenia caryophyllus (Sprengel) Bullock & S.G. Harrison, the synonym still used on labels and in the ISO standards

• Caryophylli flos (the bud) and Caryophylli floris aetheroleum (the oil), the European Pharmacopoeia names

• Lavang or laung in South Asian trade, ding xiang in traditional Chinese medicine, girofle, clavo de olor

Primary Active Compounds

• Eugenol, 4-allyl-2-methoxyphenol, CAS 97-53-0. Boiling point 254 degrees C, density 1.067 g/mL at 25 degrees C, water solubility about 2.46 g/L. The European Pharmacopoeia sets 75.0 to 88.0 percent for clove oil.

• Eugenyl acetate, also written acetyleugenol. European Pharmacopoeia range 4.0 to 15.0 percent in clove oil. Abundant in bud oil and close to absent in leaf oil, which makes it the single most useful marker for telling the two apart.

• Beta-caryophyllene, a sesquiterpene and a CB2 cannabinoid receptor ligand. European Pharmacopoeia range 5.0 to 14.0 percent.

• Methyleugenol, a trace constituent. FEMA's assessment of clove bud oil reports a mean of 0.04 percent, and EFSA's 2023 feed additive opinion cites up to 0.13 percent across clove oils. It matters out of proportion to its size, for reasons covered below.

• Non-volatile phenolics in the whole bud and in water or ethanol extracts, principally gallic acid, plus flavonoids including kaempferol, quercetin and rhamnetin. These do not appear in the essential oil at all.

Key Note

Clove is a medicinal herb with one well characterized action, local anesthesia in the mouth, and a long list of laboratory activities never tested in people. It is also one of very few kitchen spices whose concentrated form has a published record of putting infants into fulminant hepatic failure. Both facts belong in the same sentence, because the products sit on the same shelf.

Primary Functions & Benefits

Clove does four things the human evidence actually supports, and a much longer list that stops at the petri dish.

• Topical oral anesthesia. Eugenol blocks voltage-gated sodium currents in dental primary afferent neurons, both tetrodotoxin-resistant and tetrodotoxin-sensitive, independently of TRPV1 (Park and colleagues, Journal of Dental Research 2006, 85:900). This is the mechanism behind two thousand years of chewed cloves and a century of zinc oxide eugenol cement.

• Temporary toothache relief, the subject of the EMA traditional use monograph adopted 13 September 2011.

• Antimicrobial and antioxidant activity in food systems. A PRISMA review by Valarezo and colleagues (Plants 2025, 14:1958) screened 639 records and included 43 studies from 1999 to 2024 on clove essential oil in meat, at 2.65 mL/kg up to 5 percent.

• Glycemic and redox effects from water-soluble bud polyphenols, a completely different preparation from the oil, with two small industry-run trials behind it.

Claims with no human trial support at all, which appear on labels anyway, include liver detoxification, parasite elimination, cancer prevention and immune stimulation.

Forms & Standardization

This is where most clove purchasing goes wrong, because four products share one word.

Whole dried buds. The pharmacopoeial article. Ph. Eur. requires not less than 150 mL/kg of essential oil, 15 percent by volume per unit weight. A bud that fails that test is not Caryophylli flos. Whole buds hold their volatiles for years if kept sealed and cool.

Ground clove. The same material on a shorter clock. Grinding starts volatile loss immediately and a jar open for a year can run well below 15 percent, so treat eugenol arithmetic on ground spice as a range.

Clove bud oil. Steam distilled from the buds. Eugenol commonly 80 to 85 percent, with eugenyl acetate at 4 to 15 percent and beta-caryophyllene at 5 to 14 percent (Ph. Eur. limits). Specific gravity 1.036 to 1.060, refractive index 1.527 to 1.538 at 20 degrees C, optical rotation minus 2 degrees to 0 degrees, the figures 21 CFR 184.1257 states for clove oleoresin and other natural extractives, expressly distinguished there from the Food Chemicals Codex clove bud oil monograph. FEMA number 2323.

Clove leaf oil. Eugenol commonly 85 to 90 percent, per the composition figures compiled by Yao and colleagues in PLoS ONE in 2012. Higher eugenol, almost no eugenyl acetate, a fraction of the price. It is the standard adulterant of bud oil and the main industrial feedstock for isolated eugenol. FEMA number 2325.

Clove stem oil. Eugenol commonly 87 to 92 percent, the highest of the three. FEMA number 2328. FEMA's estimated per capita intakes are 2,350 micrograms per person per day for bud oil, 430 for leaf and 190 for stem.

Extraction method changes the answer. Liñán-Atero and colleagues (Antioxidants 2024) reported that supercritical CO2 extraction gave an oil that was 55.28 percent eugenol, 20.97 percent (E)-beta-caryophyllene and 7.08 percent alpha-humulene, while conventionally distilled oils in the same review ranged from 82.16 to 97.80 percent eugenol. A CO2 extract and a steam distillate labeled "clove oil" are not the same product.

Water-soluble polyphenol extract. A separate ingredient entirely. The published supplement trials used Clovinol at 250 mg per day. Thomas 2022 describes it as an ethanol and water extract of unopened buds containing 26.4 percent polyphenols, while Mohan 2019 specifies not less than 30 percent total polyphenols. It carries the gallic acid fraction, not the eugenol fraction.

Which form the human trials used. A homemade clove gel, a 1 percent clove oil cream, a 6.66 percent extract mouthwash, and 250 mg of the water-soluble polyphenol extract. Not one is a clove oil softgel, and no published human trial supports swallowing clove essential oil.

Food Sources

A level teaspoon of ground clove weighs about 2 grams by USDA reference weight (FoodData Central 171321). At 10 to 15 percent volatile oil and roughly 80 percent eugenol in that oil, one teaspoon carries 170 to 250 mg of eugenol. A quarter teaspoon, closer to what a single serving of a spiced dish contains, works out to 40 to 65 mg.

Eugenol is not unique to clove. It occurs in allspice, cinnamon leaf, bay, basil, nutmeg and lemon balm, and West Indian bay leaf and cinnamon leaf oils are assessed alongside clove in FEMA's natural flavor complex program precisely because they share the same phenol.

What cooking changes. Eugenol boils at 254 degrees C and is steam-distillable, which is the whole reason clove oil exists as a commodity, so long open simmering drives a meaningful fraction of it out of the pot. Whole buds fished out before serving deliver far less than the same mass ground and stirred in. Water solubility is only about 2.46 g/L, so eugenol partitions into fat, and a clove-spiced curry with coconut milk retains more of it than a clear broth.

Is food intake meaningful? For flavor and preservation, yes. For any therapeutic claim, no. The glycemic trials used a water-soluble polyphenol extract that is not present in meaningful quantity in the volatile-oil-dominated culinary spice, and the dental evidence is entirely topical.

One arithmetic point is worth sitting with. The JECFA acceptable daily intake for eugenol is 0 to 2.5 mg per kg of body weight, which is 175 mg per day for a 70 kg adult, and a full teaspoon of ground clove can exceed it. That does not make a teaspoon of clove dangerous. The ADI is a flavoring-exposure benchmark derived from a rat feeding study with a hundredfold safety factor already built in, not a toxicity threshold. Anyone quoting it at you as though crossing it were an event should be asked which number they think it represents.

Who Should Take Clove

The honest list is short.

• Someone with an open carious tooth waiting for a dental appointment. The one indication with a regulatory monograph behind it and a randomized trial finding no significant difference from a conventional agent. It buys days, not weeks, and does nothing about the cavity.

• Someone who wants a food preservative and antioxidant in a meat or fat-based dish. Well supported, in the kitchen sense.

• Someone specifically interested in the water-soluble clove polyphenol extract for glycemic support, with clear eyes about the evidence. Both trials are small, one is open label with 13 participants, and both were run by people employed by the ingredient manufacturer or a company selling it.

• Someone in an ICU setting where clove mouthwash is being trialed against chlorhexidine. A clinician's decision, not a consumer one.

Nobody needs clove. It is not a nutrient, and there is no deficiency state it corrects.

Who Should AVOID or Use Caution

Contraindications

• Children under 18, for the medicinal preparations. The EMA monograph of 13 September 2011 restricts both clove oil indications to adults because adequate data are lacking, and the US label for eugenol 85 percent toothache liquid says "children under 12 years: ask a dentist or doctor."

• Infants and toddlers, for clove oil in any amount, by any route that could end up swallowed. The case reports below describe severe outcomes in children aged 3 months to 2 years from volumes between about 5 and 10 mL.

• Known hypersensitivity to clove oil or to Peru balsam, the exact contraindication wording in the EMA monograph. Eugenol is one of the eight constituents of Fragrance Mix I.

• Anyone with existing liver disease considering oral clove oil, since the hepatotoxicity mechanism is glutathione depletion and a compromised liver has less reserve.

Use Caution

• Patients on warfarin or other anticoagulants. Details in the interactions section. The theoretical basis is strong and the clinical evidence is absent, which is an uncomfortable combination rather than a reassuring one.

• Anyone scheduled for surgery. Same reasoning.

• People with fragrance contact allergy. Diepgen and colleagues patch tested 3,119 people drawn from a random general-population sample of 12,377 across Germany, Italy, the Netherlands, Portugal and Sweden and found 2.6 percent contact allergy to Fragrance Mix I (95 percent CI 2.1 to 3.2), in the British Journal of Dermatology in 2015.

• Dental patients about to receive a bonded composite restoration. Carvalho and colleagues (Journal of Esthetic and Restorative Dentistry 2007) showed significantly reduced resin-dentin microshear bond strength after a zinc oxide eugenol temporary, for two self-etch adhesives (p less than 0.01).

• Pregnancy and lactation. The EMA monograph states that safety has not been established and use is not recommended, and there is no human reproductive toxicity data either way.

• People taking glucose-lowering medication, if using the polyphenol extract, given the reported reductions in fasting and postprandial glucose.

Critical Safety Point

Undiluted clove oil in a child's mouth or stomach is a medical emergency, not a home remedy gone slightly wrong. In Hartnoll's 1993 case in Archives of Disease in Childhood, a 2-year-old boy who swallowed 5 to 10 mL was drowsy at one hour, in deep coma with marked acidosis at three hours, and seizing with an unrecordable blood glucose at eight and a half hours. At 24 hours his INR was 6.51, his ALT was 1,549 IU/L and his platelet count was 56 times 10 to the ninth per liter. He recovered fully, but only after intravenous dextrose, diazepam, vitamin K, fresh frozen plasma, heparin, fibrinogen, antithrombin III, protein C and factor VII. If a child ingests clove oil, call poison control immediately and say the word eugenol. The clinical response, N-acetylcysteine on an acetaminophen protocol, is not obvious to a clinician who has not seen it before.

Recommended Dosages

Clove oil for temporary toothache relief (topical, adults). The EMA monograph permits undiluted essential oil, solutions of at least 50 percent, or gels at 20 percent, applied to the affected tooth on a cotton wool bud, repeated after 20 minutes then every two hours. US marketed products use eugenol at 85 percent, 850 mg per mL, on a cotton pellet placed in the cavity for one minute and removed, up to four times daily.

Clove oil mouthwash for minor oral or pharyngeal inflammation (adults). 1 to 5 percent essential oil in the rinse, several times daily, per the EMA monograph.

Clove extract mouthwash, investigational. 6.66 percent twice daily, from Jahanshir's 168-patient ICU trial.

Clove oil 1 percent cream, chronic anal fissure. The concentration used by Elwakeel in a 55-patient randomized trial, applied for six weeks.

Water-soluble clove polyphenol extract, oral. 250 mg once daily, standardized to 26.4 percent polyphenols in the 2022 trial and to not less than 30 percent in the 2019 pilot. This is the only oral clove dose with any published human trial behind it, and the glycemic and metabolic results come from two studies that randomized 83 participants and analyzed 77.

Whole or ground clove, culinary. No therapeutic dose exists because no therapeutic oral use has been established. Use it as a spice.

Clove bud capsules, typically 500 mg of powdered bud. A common retail format with no trial behind it at any dose. For scale, 500 mg of bud at an assumed 12 percent oil and 80 percent eugenol is roughly 50 mg of eugenol, about a quarter teaspoon of the spice.

Oral clove essential oil. No dose is recommended here. There is no human efficacy trial for swallowed clove oil at any dose, and there is a case report literature of severe toxicity from swallowing it.

Duration

The EMA monograph caps both topical indications at one week, and US eugenol toothache products carry the same seven day ceiling on their Drug Facts panel. The polyphenol extract trials ran 30 days and 12 weeks, the mouthwash trial 5 days. Nothing in the clove literature supports continuous long-term use of any preparation, and no study has followed oral clove past 12 weeks.

Timing & Administration

Topical dental use. Rinse the tooth clear of debris, then place the medicated pellet in the cavity rather than on the gum, and remove it after about a minute. Eugenol is caustic to soft tissue at 85 percent. The EMA's two hour redosing interval is not an invitation to leave a soaked cotton ball parked against the gingiva.

Mouthwash. Swish and spit. The EMA monograph is for oromucosal use, not swallowing, and the 1 to 5 percent range means one to five parts of oil in a hundred parts of vehicle, not drops of neat oil in a glass of water. Neat oil floats and arrives at the mucosa undiluted.

Polyphenol extract. Both trials dosed once daily. The 2019 pilot measured glucose two hours after the start of a meal, so the timing that produced the postprandial signal was daily supplementation rather than dosing with a specific meal.

With or without food. No pharmacokinetic study has compared fed and fasted eugenol absorption in humans. What is known, from Fischer and colleagues in Xenobiotica in 1990, is that oral eugenol in healthy volunteers is absorbed and metabolized rapidly and almost completely excreted in urine within 24 hours, with unmetabolized eugenol accounting for less than 0.1 percent of the dose.

Timeline of Effects

Minutes, for topical anesthesia. Alqareer, Alyahya and Andersson applied four agents to the maxillary canine buccal mucosa of 73 adult volunteers, waited five minutes, then delivered two needle sticks and measured pain on a 100 mm visual analogue scale (Journal of Dentistry 2006, 34:747). Both the clove gel and benzocaine 20 percent gel scored significantly lower than their matched placebos (p equals 0.005), with no significant difference between clove and benzocaine. Five minutes is the tested onset.

Five days, for the mouthwash outcome. Jahanshir and colleagues randomized 168 ICU patients at Kosar Hospital in Semnan, Iran, during 2021 and 2022 to clove extract mouthwash at 6.66 percent or chlorhexidine 0.2 percent, twice daily for five days. Ventilator-associated pneumonia occurred in 20.2 percent of the clove group and 41.7 percent of the chlorhexidine group, relative risk 2.06 for the control group (95 percent CI 1.26 to 3.37, p equals 0.005), though severity did not differ (p equals 0.557).

Twelve days, for the first glycemic signal. Mohan and colleagues saw postprandial glucose reductions at day 12 in both strata of their 13-person open-label pilot, 13.29 mg/dL in the group starting at or below 100 mg/dL and 16.67 mg/dL in the 101 to 125 mg/dL group (BMC Complementary and Alternative Medicine 2019). Preprandial glucose moved only in the higher group, and only at days 24 and 30.

Six weeks to three months, for anal fissure healing. Elwakeel's trial treated for six weeks and assessed healing at three months: 60 percent in the clove oil group versus 12 percent in the stool softener plus lignocaine 5 percent control (p less than 0.001).

Twelve weeks, for the metabolic outcomes. Thomas and colleagues randomized 70 participants and analyzed the 64 who completed, 33 on extract and 31 on control, over 12 weeks on 250 mg per day of the clove polyphenol extract against 250 mg per day of synthetic glutathione (Journal of Functional Foods 2022, volume 98), reporting a 23.47 percent HbA1c decrease against 2.09 percent, and a 51.42 percent HOMA-IR reduction against 3.27 percent.

Benefits of Taking Clove

Topical dental analgesia that matches a pharmacy standard. The Alqareer trial is small and uses a homemade gel, but it is a randomized subject-blinded comparison against benzocaine 20 percent with both placebos included, and clove held its own. For a household that wants something in the medicine cabinet for a cracked molar at midnight, this is a defensible choice.

A mechanism that is actually understood. Eugenol inhibits sodium currents in dental afferents in a TRPV1-independent way (Park 2006), activates TRPA1 with an EC50 of 261.5 micromolar (Chung and colleagues, Neuroscience 2014, 261:153), and inhibits the calcium-activated chloride channel TMEM16A (Yao and colleagues, PLoS ONE 2012). Very few botanicals have this level of target resolution.

Antifungal performance against a clinical comparator. A systematic review of denture stomatitis treatments (Carvalho-Silva and colleagues, Journal of Prosthetic Dentistry 2025) found clove and cinnamon extract gels performing at or above miconazole and clotrimazole.

Glycemic and redox effects from the polyphenol fraction. The 12-week trial reported a 55.23 percent rise in plasma glutathione against 9.15 percent for synthetic glutathione, and a 217.71 percent rise in Nrf2 against 17.77 percent. Large effect sizes, from an active-controlled trial run by the ingredient's developers with no placebo arm, never replicated independently.

Mosquito repellency. Trongtokit and colleagues screened 38 essential oils on human forearms (Phytotherapy Research 2005). Undiluted clove oil gave the longest complete repellency of any oil tested, 2 to 4 hours against three mosquito species.

Potential Negatives & Side Effects

• Mucosal irritation and chemical burn. The most common problem with dental use. The 85 percent product label warns specifically against contact with lips and mucous membranes because of burning, swelling and irritation.

• Allergic contact dermatitis and contact urticaria. Eugenol sits in Fragrance Mix I. Kieć-Świerczyńska and colleagues patch tested 1,937 patients at the Nofer Institute in Łódź between 2000 and 2005 and found 99 (5.1 percent) positive to fragrance mix, with eugenol among the most frequent individual sensitizers (Medycyna Pracy 2006). In the European general population sample, Fragrance Mix I positivity was 2.6 percent.

• Photoallergy overlap. Marmgren and colleagues found eugenol contact allergy in 23.3 percent of patients photoallergic to ketoprofen, against 0.4 percent in the general dermatitis group (Contact Dermatitis 2021, 85:660).

• Reduced bond strength of resin restorations after eugenol-containing temporary cements.

• Gastrointestinal upset from oral clove preparations, reported inconsistently and never quantified in a controlled trial.

• Respiratory injury from clove cigarettes. The AMA Council on Scientific Affairs (JAMA 1988, 260:3641) linked inhaled clove cigarette smoke to severe lung injury in susceptible individuals with a prodromal respiratory infection, and to aspiration pneumonitis caused by eugenol's local anesthetic action blunting the gag reflex.

• What has not been reported. No case of hepatotoxicity from culinary clove. No case of hepatotoxicity from clove bud capsules at label doses. No serious adverse event from the 250 mg polyphenol extract over 12 weeks. The harm signal is specific to the concentrated oil, and mostly to children.

Deficiency Symptoms

Clove is a spice, not a nutrient, and there is nothing in it your body is required to obtain. No intake level of clove is essential, there is no recommended intake, no biomarker of clove status, and no syndrome that arises from not eating it. Anyone marketing a "clove deficiency" is selling something.

What clove addresses

Localized dental pain from an exposed cavity, by anesthetizing the nerve endings around it. Oral microbial load, in the ICU mouthwash setting that has been tested. Lipid oxidation in stored meat. Possibly postprandial glucose and glutathione status, at the level of two small manufacturer-run trials. That is the complete list of things clove has been shown to do in humans.

Bottom line

Clove is optional and targeted. Its value is a specific preparation applied to a specific problem for a short period, and there is no baseline health case for taking it at all.

Toxicity Symptoms

At high intake

The clinical picture from concentrated clove oil ingestion, drawn from the published pediatric case reports, is consistent enough to recognize.

• Central nervous system depression, from drowsiness to deep coma. Hartnoll's 2-year-old was drowsy at one hour and comatose by three.

• Profound hypoglycemia, with seizure. Hartnoll's patient had an unrecordable blood glucose at eight and a half hours.

• High anion gap metabolic acidosis, the feature Lane and colleagues emphasized in their 7-month-old (Human and Experimental Toxicology 1991, 10:291), alongside CNS depression and urinary abnormalities.

• Acute hepatocellular injury. ALT 1,549 IU/L at 24 hours in Hartnoll's case, above 13,000 U/L at 24 hours in Janes's 15-month-old.

• Coagulopathy and disseminated intravascular coagulation. INR 6.51, platelets 56 times 10 to the ninth per liter, fibrin degradation products above 250 IU/L in Hartnoll's case. Brown, Biggerstaff and Savidge reported DIC with hepatocellular necrosis in a 2-year-old, possibly the same child, in Blood Coagulation and Fibrinolysis in 1992 (3:665).

• Renal involvement. Janes reported blood urea of 11.8 mmol/L and creatinine of 134 micromol/L.

The volumes are the part people get wrong. Eisen and colleagues described fulminant hepatic failure in a 3-month-old girl after less than 8 mL. Janes described it in a 15-month-old after 10 mL. Hartnoll's 2-year-old took 5 to 10 mL. The EMA assessment report describes Lane's 7-month-old as having received one teaspoon. Every one of those children survived, and Kim and colleagues reported a further case supporting early N-acetylcysteine in the Journal of Pediatric Gastroenterology and Nutrition in 2018.

JECFA's evaluation lists oral LD50 values for eugenol of 1,930 to 2,680 mg/kg in rats, 3,000 mg/kg in mice and 2,130 mg/kg in guinea pigs. Eight milliliters of clove bud oil is roughly 6.7 grams of eugenol, and in a 6 kg infant that is above 1,100 mg/kg. The cases are not mysterious. The dose was very large for the body receiving it.

Signs to reduce or stop

• Any burning, ulceration, whitening or swelling of the gum, cheek or lip where a clove preparation has been applied. Stop and rinse.

• Rash, itching, hives or eczema after topical use, which points to eugenol sensitization rather than irritation and means stopping permanently rather than reducing.

• Nausea, vomiting or abdominal pain after any oral clove product, or unusual bruising, prolonged bleeding from a dental site, or a rising INR in anyone on warfarin.

• Any symptom persisting past seven days on a topical preparation, the regulatory ceiling in both the EMA monograph and the US Drug Facts labeling.

General note

Clove has an unusual risk profile among culinary botanicals. The food is essentially harmless, the topical preparation has a mild and well defined adverse event list, and the concentrated oil is dangerous to small children in volumes that fit in a bottle cap. The risk is not spread evenly across the forms, and treating "clove" as one thing is the error that produces the case reports.

How Clove Works

Local anesthesia. Eugenol inhibits both tetrodotoxin-resistant and tetrodotoxin-sensitive sodium currents in rat dental primary afferent neurons, and pretreatment with the TRPV1 antagonist capsazepine fails to block the effect, so the sodium channel block is independent of the vanilloid receptor (Park 2006). This is the same broad class of action as lidocaine, reached by a different molecule.

Sensory receptor activity. Eugenol activates TRPA1 with an EC50 of 261.5 micromolar, confirmed in trigeminal ganglion neurons from TRPV1 knockout mice and in HEK293 cells expressing human TRPA1, and blocked by the antagonist HC-030031 (Chung 2014). It also acts at TRPV1. The warm prickle of clove on the tongue and the anesthesia that follows come from different parts of the same interaction.

Chloride channel inhibition. Fractionation of a herbal antidiarrheal preparation identified eugenol as an inhibitor of TMEM16A, the calcium-activated chloride channel that drives secretory fluid loss in the gut (Yao 2012).

Antioxidant and Nrf2 signaling. The phenolic hydroxyl on eugenol makes it a radical scavenger, which is both why it preserves meat and why it inhibits resin polymerization in a dentist's hands. The water-soluble polyphenol fraction, dominated by gallic acid, is where the reported Nrf2 induction and glutathione elevation come from.

The toxicity mechanism, which is the same chemistry pointed the wrong way. Eugenol's phenolic ring with an allyl group at the 4-position is oxidized by cytochrome P450 to a vinylogous quinone methide, a reactive electrophile. Mizutani, Satoh and Nomura showed that eugenol produces hepatic injury in mice depleted of glutathione with buthionine sulfoximine, and that the structural requirement is exactly that 4-allyl phenol arrangement (Research Communications in Chemical Pathology and Pharmacology 1991, 73:87). Bolton and colleagues characterized the quinone methide directly, with a 472-fold spread in reactivity across the related 4-alkyl series (Chemico-Biological Interactions 1995, 95:279). At ordinary intakes glutathione mops the intermediate up. At an overwhelming dose glutathione runs out and the electrophile binds hepatocyte proteins. That is why N-acetylcysteine works, and why the clinical course looks like acetaminophen poisoning.

Metabolism and clearance. Fischer and colleagues gave eugenol orally to healthy volunteers and recovered about 95 percent of the dose in urine, more than 99 percent as phenolic conjugates, with eugenol glucuronide and sulfate accounting for half of the conjugated metabolites. Nine metabolites were identified, including an epoxide-diol pathway. Less than 0.1 percent was excreted unchanged and clearance was complete within 24 hours, so there is no accumulation with daily use.

Methyleugenol, the trace problem. Methyleugenol is bioactivated to the proximate carcinogen 1'-hydroxymethyleugenol, chiefly by CYP1A2, and Jeurissen and colleagues measured a five-fold range in that activity across liver microsomes from 15 individuals, 0.89 to 4.30 nmol per minute per nmol P450 (Chemical Research in Toxicology 2006, 19:111). At a mean 0.04 percent of clove bud oil, a teaspoon of ground clove contributes micrograms.

Synergistic Supplements

This section is short because clove synergy claims mostly have not survived testing.

• Zinc oxide. Not a supplement, but the genuine article. Zinc oxide and eugenol form the chelate cement that has been dentistry's default temporary restorative for a century, and it is the reason eugenol is in a dental office at all.

• N-acetylcysteine. An antidote rather than a partner, but worth naming because the pairing is clinically load-bearing. NAC restores the glutathione that eugenol overdose depletes, and it is what Eisen, Janes and Kim treated their patients with. The three earliest reports did not use it, and Hartnoll proposed it rather than gave it.

• Other phenolic spices, with a caveat. Clove is routinely combined with cinnamon, oregano and thyme in antimicrobial blends on the theory that the phenols potentiate each other. Schlösser and Prange ran the actual checkerboard assay against Penicillium verrucosum and Aspergillus westerdijkiae, pairing eugenol with carvacrol, thymol, trans-cinnamaldehyde and 1,8-cineole, and reported fractional inhibitory concentration indices from 0.8 to 1.3 (FEMS Microbiology Letters 2018, 365). That range is the textbook definition of no interaction. Additive, not synergistic.

• What is not supported. Clove with black pepper for "bioavailability" has no human pharmacokinetic study behind it, and eugenol is already almost completely absorbed and conjugated within 24 hours.

Interactions & What NOT to Take

Warfarin. Heck, DeWitt and Lukes reviewed alternative therapy interactions with warfarin in the American Journal of Health-System Pharmacy in 2000 (57:1221) and placed clove on the list of herbs that may potentiate it. They also drew a line that matters: clove is on the theoretical list, not on their list of products with documented case reports. The basis is that eugenol is a potent cyclooxygenase inhibitor. Raghavendra and Naidu measured it in human platelets, reporting an IC50 of 0.5 micromolar against arachidonic acid-induced aggregation, 29-fold more potent than aspirin, with dose-dependent inhibition of thromboxane B2 formation (Prostaglandins, Leukotrienes and Essential Fatty Acids 2009, 81:73). That is an isolated platelet result at a concentration dietary intake does not reach in plasma. The prudent position is an INR check rather than a prohibition.

Aspirin, clopidogrel, ticagrelor, prasugrel, apixaban, rivaroxaban, dabigatran, heparin and enoxaparin. The same reasoning extends to each by mechanism, with the same absence of case reports. Oral clove oil alongside any of them is the scenario worth avoiding. Culinary clove is not.

Acetaminophen. Both are cleared largely by glucuronidation and sulfation, and both generate a reactive intermediate detoxified by glutathione. Nobody has run the interaction study in humans. The overlap is the reason clove oil poisoning is treated on an acetaminophen protocol in the first place, and a reason not to combine a large oral clove oil dose with acetaminophen.

Glucose-lowering drugs: metformin, glipizide, glyburide, glimepiride, insulin. The polyphenol extract trials reported fasting glucose reductions of 17.28 percent over 12 weeks and postprandial reductions of 13 to 17 mg/dL over 12 to 30 days. Additive hypoglycemia has not been reported, but nobody has looked for it in a co-treated population.

Dental resin adhesives. A measured, clinically relevant interaction rather than a theoretical one. Tell your dentist if you have been using clove oil on a tooth that is about to be restored with a bonded composite.

Glutathione S-transferase substrates. Van Bladeren noted that eugenol lowers glutathione S-transferase activity in humans (Biomedicine and Pharmacotherapy 1997, 51:324). No specific drug interaction follows from it yet, so treat it as an open question.

What the evidence does not show. There is no published human case report of a clove-warfarin bleeding event. There is no controlled interaction study of clove with any prescription medicine. There is no evidence that culinary clove affects any drug. Every interaction in this section is inferred from mechanism, and anyone who tells you otherwise has not read the source.

Quality, Testing & Adulteration

Clove is one of the more heavily adulterated items in the spice trade, and the specific frauds are well characterized.

Exhausted buds. Cloves already steam distilled for their oil, then dried and sold as whole spice. They look nearly identical and hold a fraction of the 150 mL/kg the Pharmacopoeia requires. The kitchen screen is that a sound clove sinks in water head-down while a spent one floats or lies flat, because the oil-heavy head is denser. A hint, not a test.

Clove stalks ground into clove powder. Cheap bulking material from the same plant, invisible to a botanical identity test.

Clove leaf oil sold as clove bud oil. The most consequential substitution, because leaf oil is cheaper and carries a higher eugenol percentage, so a total phenol assay passes it. The marker that catches it is eugenyl acetate, set at 4.0 to 15.0 percent by the European Pharmacopoeia and nearly absent from leaf oil. A certificate of analysis reporting only eugenol content is not evidence of bud origin.

Diluent oils. Selim, Darwish and Shawky built a near-infrared model for exactly this, testing clove powder against clove stalks and exhausted buds, and clove oil against clove leaf, olive, corn, rosemary, cinnamon and basil oils (Microchemical Journal 2023). Their SIMCA classification hit 100 percent sensitivity and specificity, with limits of quantitation from 1.8 to 4.9 percent. Vargas Jentzsch and colleagues published a handheld Raman method for the same purpose in 2018. The analytical tools exist. Most retail supply chains do not use them.

What a certificate of analysis should carry, and usually does not. GC-MS quantifying eugenol, eugenyl acetate and beta-caryophyllene individually rather than as total phenols. Specific gravity within 1.036 to 1.060, refractive index within 1.527 to 1.538 at 20 degrees C and optical rotation between minus 2 and 0 degrees, the figures 21 CFR 184.1257 sets for clove oleoresin and other natural extractives rather than the Food Chemicals Codex clove bud oil values. The plant part stated explicitly with the relevant ISO standard cited, 3141 for leaf, 3142 for bud, 3143 for stem. Methyleugenol quantified, given that it was first listed in the Report on Carcinogens in 2002 as reasonably anticipated to be a human carcinogen and added to California's Proposition 65 list on 16 November 2001. Heavy metals and pesticide residues, since clove oil (CAS 8000-34-8) and eugenol (CAS 97-53-0) are themselves EPA minimum risk pesticide active ingredients under FIFRA section 25(b).

A regulatory gap worth knowing. FDA's Over-the-Counter Monograph M022 for oral healthcare products, posted 14 October 2022, defines an "agent for the relief of toothache" as an ingredient used for the temporary relief of pain arising from an open tooth cavity, and then lists no active ingredient section for that category anywhere in its Part B. Eugenol does not appear in M022 at all.

Special Considerations

Children. The EMA monograph excludes everyone under 18 from both clove oil indications for lack of data, US toothache products direct anyone under 12 to a dentist, and the case literature is entirely pediatric. Clove oil in a home with small children deserves the storage discipline of a bottle of acetaminophen, without the packaging acetaminophen is required to have.

Teething. Clove oil on an infant's gums is a traditional remedy in several cultures and a genuinely bad idea. The gum is the exact tissue the 85 percent product warns against contacting, an infant swallows what is placed in the mouth, and the youngest documented case of fulminant hepatic failure was three months old.

Pregnancy and lactation. No data. EMA recommends against use. Culinary amounts are not the subject of that recommendation.

Liver disease. Because the mechanism is glutathione depletion, anyone with reduced hepatic reserve, including people with chronic alcohol use or cirrhosis, has less of the buffer that makes eugenol safe at ordinary doses.

Contact allergy. Once sensitized to eugenol, a person reacts to fragranced cosmetics, toothpaste, mouthwash and foods containing clove, cinnamon and allspice. Eugenol is one of the 26 fragrance allergens that Regulation (EC) No 1223/2009 requires to be individually declared on EU cosmetic labels above 0.001 percent in leave-on products and 0.01 percent in rinse-off products, a list expanded substantially by Commission Regulation (EU) 2023/1545 of 26 July 2023 with compliance deadlines of 31 July 2026 and 31 July 2028. US cosmetic labels are permitted to say "fragrance" and stop, which means a sensitized American consumer cannot read the ingredient off the package.

Research Status & Evidence Quality

Tier 1, randomized controlled trials with clinically meaningful endpoints. Five, and they are heterogeneous. Alqareer 2006 (n equals 73, topical anesthesia), Elwakeel 2007 (n equals 55, single-blind, anal fissure), Jahanshir 2023 (n equals 168, triple-blind, ventilator-associated pneumonia), Thomas 2022 (70 randomized and 64 analyzed, double-blind but active-controlled with no placebo, metabolic endpoints) and Mammen 2018 (n equals 16, double-blind placebo-controlled crossover, oxidative stress and hangover severity endpoints, Journal of Medicinal Food 21:1188). Those five randomized 382 people between them. Further randomized trials of clove preparations exist, including two of topical clove oil in children, so that figure is not a count of the whole literature.

Tier 2, open-label and uncontrolled human studies. Mohan 2019, n equals 13, no control, registered retrospectively (ISRCTN15680985), authored by employees of the ingredient manufacturer and of a retailer selling the finished product.

Tier 3, regulatory assessment without clinical trial support. The EMA monograph of 13 September 2011 is a traditional use monograph, the category the HMPC uses when long-standing use substitutes for clinical evidence. Its assessment report states plainly that no clinical data are available for clove or clove oil sufficient to support well-established use. The only non-targeted human study the HMPC could identify was Wagner and Sprinkmeyer's 1973 psychometric assessment of 72 subjects exposed to clove in room air, which found no differences in any parameter.

Tier 4, mechanistic and in vitro work. Extensive and genuinely good: sodium channel electrophysiology, TRPA1 patch clamp with a measured EC50, TMEM16A inhibition, quinone methide chemistry with quantified reactivity ratios, human metabolic mapping with nine identified urinary metabolites. Clove is better understood mechanistically than most botanicals are clinically.

Tier 5, animal toxicology. NTP Technical Report 223 fed eugenol to F344/N rats and B6C3F1 mice for 103 weeks, at 3,000 and 6,000 ppm for male rats, 6,000 and 12,500 ppm for female rats and 3,000 and 6,000 ppm for mice, 50 animals per dosed group. The conclusion was negative for carcinogenicity in both sexes of rats and equivocal in both sexes of mice, on increased liver neoplasms. JECFA used a rat NOAEL of 250 mg/kg per day with a hundredfold safety factor to set the ADI of 0 to 2.5 mg/kg at its twenty-sixth meeting in 1982, and has maintained it since.

Research Limitations

• The randomized evidence examined here is 382 participants across five trials in five unrelated indications, and no indication has been replicated independently.

• The trials supporting oral clove supplementation all used the same proprietary extract and were run by parties with a commercial interest in the result.

• The pediatric safety data is entirely case reports, which establishes that severe harm happens but not how often. There is no published population denominator for clove oil poisoning in any country. Janes and colleagues could only report a 14-fold increase in aromatherapy-related home accidents in a national database between 1994 and 1999, a proxy rather than a count.

• Essentially nothing is known about eugenol pharmacokinetics in infants, the population with the documented harm.

Summary & Key Takeaways

Bottom Line

Clove has one well-supported medical use, short-term topical relief of pain from an open dental cavity, and a concentrated form that is dangerous to small children. That dental use rests on a randomized trial finding no significant difference from benzocaine 20 percent at five minutes and an EMA traditional use monograph adopted 13 September 2011 capping it at one week in adults. Everything else being sold is unsupported or rests on a few small trials of a single proprietary polyphenol extract run by the people who make it. The correct mental model is three products: the spice, which is food and safe; the topical oil, a short-term analgesic with real irritant and allergenic potential; and the swallowed oil, which has no evidence base and a case report literature.

Key Safety Points

• Five published reports document severe toxicity in children aged 3 months to 2 years from clove oil ingestions of roughly 5 to 10 mL: Lane 1991, Brown 1992, Hartnoll 1993, Eisen 2004 and Janes 2005. Brown and Hartnoll both describe a 2-year-old with disseminated intravascular coagulation and hepatocellular necrosis managed at St Thomas' Hospital in London, eleven months apart, treated with the same unusual regimen of coagulation factor and inhibitor concentrates, so the two reports may describe one child rather than two. The published record does not settle it, and those five reports therefore cover four or five children. The smallest documented volume to cause fulminant hepatic failure was less than 8 mL in a 3-month-old.

• The syndrome is coma, seizure, hypoglycemia, high anion gap acidosis, hepatocellular necrosis and coagulopathy. It resembles severe acetaminophen overdose because the mechanism is similar, glutathione depletion by a reactive intermediate.

• N-acetylcysteine on an acetaminophen protocol is the reported treatment, and earlier administration appears better. Say "eugenol" when you call poison control.

• Clove oil is not on the 16 CFR 1700.14 list of substances requiring child-resistant packaging in the United States. Store it as if it were.

• Eugenol at 85 percent burns soft tissue. Keep it in the cavity, off the gum, and stop at seven days.

• Clove is on the theoretical warfarin interaction list on the strength of eugenol's platelet cyclooxygenase inhibition, an in vitro IC50 of 0.5 micromolar. No human case report of a clove-warfarin bleeding event has been published.

• Tell your dentist, because eugenol inhibits resin polymerization and measurably weakens composite bonding.

Special Note

The strangest fact about clove is regulatory rather than pharmacological. FDA's Over-the-Counter Monograph M022, posted on 14 October 2022, defines what a toothache relief agent is and then names no ingredient that qualifies as one. Eugenol, the compound in every clove-based toothache product sold in an American pharmacy, appears nowhere in it. Those products are labeled under M022 anyway, with "toothache relief agent" printed as the Drug Facts purpose. The European Medicines Agency adopted a monograph for the oil on 13 September 2011, then declined on 22 November 2011 to write one for the bud at all, because no single-ingredient clove bud products were on the EU market and stakeholders had given it low priority. Two regulators, one plant, decades of paperwork, and the practical result is a bottle on a shelf that anyone can buy, that no agency has finalized rules for, and that has put more than one infant in intensive care.