Guide
Coffee Chemistry: Acids, Polyphenols and Roast Reactions
Quick Answer
Coffee contains a mixture of compounds, including caffeine, organic acids, polyphenols and aroma molecules. Roasting transforms the bean's chemistry; brewing transfers part of that mixture into water. Acidity, bitterness, aroma and strength describe different parts of the result, so one number cannot describe the whole cup.

On This Page8 Sections
The Useful Compound Groups
A green bean, a roasted bean and a cup of coffee are different samples. Green coffee supplies sugars and amino compounds that participate in roasting reactions. The brewed drink contains material extracted from the roasted grounds, at a concentration set by the preparation.
Swipe to see all columnsScroll to see all columns
| Group | Examples | Why it matters |
|---|---|---|
| Caffeine | Caffeine, an alkaloid | One contributor to bitterness; the amount consumed depends on the actual serving |
| Organic acids | Citric, malic, lactic and quinic acids | Part of the drink's acid chemistry, alongside its buffering and other constituents |
| Polyphenols | Chlorogenic acids, including caffeoylquinic acids | Named compounds that can be measured separately from overall antioxidant activity |
| Volatile aroma compounds | Pyrazines and many other families | Molecules reaching the nose help create roasted, fruity, floral and other aromas |
| Brown roasting products | Melanoidins | Products of browning chemistry that form as the beans roast |
The groups overlap: chlorogenic acids are both phenolic compounds and acids. For caffeine in defined servings, use the caffeine reference.
Trigonelline is another coffee alkaloid, distinct from caffeine. During roasting, part of it breaks down into compounds including nicotinic acid (niacin) and N-methylpyridinium. Those are separate substances, which researchers measure individually.
Is Coffee Acidic, and Which Acids Does It Contain?
Plain brewed coffee is normally acidic. It is an acidic mixture, not one chemical called “coffee acid.” Citric acid occurs in coffee itself, not only in products used to descale equipment. A 2024 experiment detected citric, malic, lactic and quinic acids in its brewed dark-roast Arabica samples using GC-MS and NMR analysis.
In Rao and Fuller's six-coffee experiment, hot and cold brews measured pH 4.85–5.13. Both used 35 g coffee per 350 ml (11.8 US fl oz) water. The hot coffee steeped in a French press carafe; both brews were paper-filtered before analysis. Their titratable acidity differed even where pH was comparable.
Those measurements answer different questions:
- pH describes the acid/base state of the liquid. Near room temperature, a value below 7 is acidic.
- Titratable acidity describes how much base a defined test needs to reach its chosen endpoint.
- Perceived sourness is a sensory experience shaped by the drink's composition and balance.
A lemon note on a bag is a flavor comparison, not a laboratory identification of citric acid. The flavor-notes guide explains that vocabulary. Brewing-water pH is another separate measurement; use the water guide for alkalinity and buffering.
Why Coffee Can Be Bitter and Acidic Together
An acidic drink can also taste bitter because bitterness is not the opposite end of the pH scale. Bitter compounds interact with bitter-taste receptors. Caffeine and cafestol are two coffee compounds that activated the human receptor TAS2R43 in a 2026 cell-assay study.
That helps explain why acidity does not exclude bitterness. A pH measurement cannot identify which bitter compounds are present or how strongly you will perceive them. Likewise, a bitter taste cannot tell you the coffee's pH or caffeine dose.
I would describe sourness and bitterness separately before changing a brew. If both are unpleasant, check the recipe and evenness rather than trying to make the coffee “less acidic” by one large grind change. The extraction guide connects those observations with practical checks.
What the Maillard Reaction Does During Roasting
Maillard reactions are a network of reactions involving reducing sugars and amino compounds, including amino acids. They help turn the chemistry of green coffee into the aromas and brown color of roasted coffee.
The useful sequence is:
- Sugars and amino compounds react. Their identities and the heating conditions affect the paths available.
- Reactive intermediates form and change. Carbonyl compounds can break down or react further with amino acids, including through Strecker reactions.
- Further reactions produce aroma and browning products. These include volatile compounds and larger brown products called melanoidins.
Pyrazines are one useful aroma family associated with this chemistry. 2,5-Dimethylpyrazine, detected in a coffee-roasting experiment, is associated with roasted and nutty odor. It is one contributor among many, rather than the single explanation for every hazelnut or almond note.
Caramelization is different sugar chemistry: it does not require an amino compound as a reactant. Several reaction pathways operate during roasting, so a caramel-like aroma is not proof that one reaction alone produced it.
A roaster's “Maillard phase” is also a timing label. Yellowing and first-crack markers do not switch chemical reactions on and off. The roast-curves guide explains how to read those events, temperature and rate of rise; the roast-level guide handles buying and taste choices.
Polyphenols, Chlorogenic Acids and Antioxidant Tests
Chlorogenic acids are a family of phenolic compounds found in coffee. Examples include 3-, 4- and 5-caffeoylquinic acid, often shortened to CQA. Measuring those compounds answers a different question from measuring a brew's overall radical-scavenging activity.
Caffeic acid is a phenolic acid; caffeoylquinic acids contain a caffeic-acid part linked to quinic acid. In Nardini and colleagues' 2002 tested brew, free caffeic acid was undetectable, while breaking down the bound forms released it. A result for free caffeic acid therefore answers a different question from total caffeic acid measured after that treatment.
A radical-scavenging assay measures how a sample reacts with a chosen test radical. Rao and Fuller used an ABTS assay, reporting activity in Trolox equivalents per litre. That result describes behavior in that chemical test; it is not a count of antioxidant molecules or a prediction of a health benefit from drinking the cup.
For a meaningful comparison, keep the preparation, serving basis and analytical method with the result. A value per litre of brewed coffee cannot be compared directly with a value per gram of green beans. Nor can an ABTS result be treated as interchangeable with another assay's score.
For choosing a daily coffee, I would give the flavor I enjoy and a repeatable preparation more weight than a highest-antioxidant claim.
Read the Measurement Before Drawing the Conclusion
Swipe to see all columnsScroll to see all columns
| If the result says… | Establish this first | Useful next step |
|---|---|---|
| “pH 5” | Is this the brewed drink or its water, and at what measurement temperature? | Keep the number separate from the taste description |
| “High polyphenols” | Which compounds, which sample and what unit? | Compare the same material and preparation basis |
| “Higher antioxidant activity” | Which assay, reporting unit and dilution? | Read it as that test's result |
| “1.35% TDS” | Is this coffee-specific dissolved-solids concentration? | Use the extraction guide, not the number as a chemical inventory |
| “Dense coffee” | Does the volume include gaps between beans or particles? | Use the density guide to match material and measurement method |
Chemistry is most useful when it sharpens the question. Ask which compound changed, what was measured and whether that difference matters to the cup you want. Keep a brewing adjustment with the relevant recipe, and keep a chemical comparison with its actual sample and method.
Acrylamide, Mycotoxins, Microplastics and Insect Damage
These describe different substances or defects, with different tests. A detection result identifies something in the tested sample; exposure also depends on its concentration and how much of that food or drink is consumed.
| Question | What the evidence describes |
|---|---|
| Acrylamide in coffee | A compound formed during roasting. The FDA explains that it forms when coffee beans are roasted, rather than when the coffee is brewed at home. |
| Mold and mycotoxins | Mold is a fungus; mycotoxins are chemicals produced by certain molds. Ochratoxin A is one such coffee contaminant assessed by EFSA. Visible mold and a measured toxin concentration are different findings. |
| Microplastics | Small plastic particles. The FDA highlights gaps in measurement methods and exposure research; current evidence does not demonstrate that the levels detected in foods pose a human-health risk. |
| Bugs and insect damage | Pests can infest or damage beans. The FDA's inspection method names the coffee berry borer and coffee bean weevil; it measures bean defects and contamination in a sampled lot. |
The FDA's green-coffee defect action level is an average of 10% or more insect-infested or insect-damaged beans by count. It does not mean coffee is 10% insects by weight, describe the average bag, or give a bug count for a brewed cup. The FDA also states that poor manufacturing can prompt action below a defect threshold.
For home storage, keep coffee dry and sealed, and discard visibly moldy, water-damaged or contaminated coffee. The storage guide separates those problems from ordinary flavor loss. For an analytical report, keep the sample, measured amount, units and test method together; a raw-bean measurement and a brewed-serving exposure are different bases.
Sources15
-
Rao and Fuller (2018), Acidity and Antioxidant Activity of Cold Brew Coffee: pH, titration, caffeoylquinic acids and ABTS measurements in the specified brews.
-
Bratthäll and colleagues (2024), Organic acids in coffee measured by GC-MS and NMR: detected acids and analytical context.
-
Kim and colleagues (2026), Structural insights into coffee bitter taste perception by TAS2R43: coffee-compound receptor assays.
-
Cha, Debnath and Lee (2019), Maillard reaction model systems: sugar/amino-acid reactions and their intermediate and volatile products.
-
Maillard reaction precursors and Arabica coffee beverage quality (2023): reducing sugars and soluble proteins in green and roasted samples.
-
Enhancing the Understanding of Aroma Formation during Coffee Roasting (2025): detected coffee volatiles, including pyrazines.
-
Kinetic modelling of Maillard reaction products during coffee roasting (2024): melanoidins measured during the tested roasts.
-
Lee, Oh and Lee (2024), Volatile and alpha-dicarbonyl compounds in Arabica coffee: roasting-reaction context.
-
Lang and Hofmann (2009), Trigonelline degradation during coffee roasting: named alkaloid and its roasting products.
-
Nardini and colleagues (2002), Phenolic acids in the tested coffee brew: free and bound caffeic acid and the hydrolysis distinction.
-
FDA: Acrylamide and Food Preparation: formation during bean roasting.
-
EFSA (2020), Risk assessment of ochratoxin A in food: toxin identity, food occurrence and exposure basis.
-
FDA: Microplastics and Nanoplastics in Foods: current evidence and analytical gaps.
-
FDA: Food Defect Levels Handbook: green-bean count basis and enforcement context.
-
FDA: Method for Coffee Beans: pests, bean defects and lot inspection.