This page describes how coffee leaf, depending on how it is processed, differs in aroma, taste, and mouthfeel — and summarises what published research has identified about the compounds responsible for those differences.
It is a sensory and chemical reference, written for the same reason Paper III's "Compass" exists: to give a vocabulary for what to taste for, and why. Where Paper III approaches this from the processor's side — what to do to move toward a sensory direction — this page approaches it from the drinker's side: what the finished leaf, prepared different ways, tends to taste and smell like, and what is currently known about why.
This page does not make claims about mood, stress, sleep, focus, or any other physiological or psychological outcome of drinking coffee leaf tea. Some of the research cited here measured such outcomes in specific small studies — that research is described in "A Note on the Research" below, in the terms the original studies used. It is not generalised here into claims about what a cup of coffee leaf tea will do for a reader. See the Foundation page's Evidence Discipline for why.
Companion to: The Foliage of Buna (Foundation) · Citane: Experimental Pathways (Paper III, the Compass)
Documented Research
A 2023 study presented to the International Coffee Convention analysed 24 coffee leaf samples produced under different processing methods, using descriptive sensory analysis and gas chromatography–mass spectrometry–olfactometry (GC-MS-O) — a method that identifies which specific aroma compounds are present and which of them a human panel can actually smell. The clearest finding was a consistent difference between fermented and non-fermented leaf.
| Attribute | Fermented Leaf | Non-Fermented Leaf |
|---|---|---|
| Primary aroma | Sweetish, fruity, honey-forward | Intense green, fresh, vegetal |
| Secondary notes | Floral undertones, dried-apricot and hay-like complexity | Citrus-like brightness, herbal clarity |
| Reported mouthfeel | Smoother, with lingering warmth | Crisper, lighter body |
| Aftertaste | Honey-like persistence | Clean, green finish |
Fermented coffee leaf samples showed more sweet and fruity aroma notes; non-fermented samples showed more intense green and vegetable aroma. This pattern is consistent with what Paper I describes mechanistically (Reservoir C, oxidation cascade) and what Paper III's Compass uses as the basis for Directions 1–2 (toward honey-like/floral) versus Direction 6 (toward grassy/green).
"Sweetish," "honey-forward," and "fruity" describe aroma character, not sweetness in the sense of sugar content or perceived intensity of taste — the study used descriptive sensory analysis (trained panellists describing what they perceive), not a sweetness-preference test. A leaf may smell "honey-like" without tasting noticeably sweet in the cup.
Source: Rigling, Steger, Lachenmeier, Schwarz & Zhang (2023), Sensorial and Aroma Profiles of Coffee By-Products — Coffee Leaves and Coffee Flowers, Proceedings 2023, 89, 5. CC BY 4.0.
Documented Research
The same GC-MS-O analysis identified five compounds as the key contributors to coffee leaf's aroma profile, present across samples in different ratios depending on processing. These are the same compounds Paper III's Compass uses to organise its six "sensory directions" — here they're described as what's actually been found in the leaf, rather than as a target to process toward.
Associated with fermented processing. The primary contributor to the "honey-forward" character of fermented leaf described above.
Found across processing types, more pronounced in non-fermented leaf — contributes to the citrus brightness noted in green leaf aroma.
Related to β-ionone (both are carotenoid breakdown products — see Paper I, Reservoir K) but with a more floral than honey character.
Contributes to fruity notes, distributed differently across processing types alongside the compounds above.
The aroma of fresh plant material — highest in non-fermented leaf, and the main contributor to the "intense green and vegetable" character noted above.
All five compounds were identified by GC-MS-O in coffee leaf samples and "distributed in different ratios" depending on processing — the source study does not give a single fixed concentration for each compound, but reports their relative presence across the fermented/non-fermented spectrum described in Section One.
A second study (Fibrianto et al., 2025 — see "A Note on the Research" below) additionally reported diphenyl sulfone and phenyl-methylene heptanoate as present at higher levels in oolong-style coffee leaf tea than in other styles tested, including Camellia sinensis teas. Both compounds are reported in that study as occurring in only trace amounts, and their general odour character is not established in the same way as the five compounds above — they are mentioned here for completeness, not as established contributors to a specific aroma note.
Source: Rigling et al. (2023), as above. See also Paper I, Reservoir K (Volatile Precursors) and Paper III, "The Citane Compass," for how these compounds relate to processing choices.
Documented Research
Beyond which leaf is used, how it is brewed also affects the cup — both through extraction generally, and specifically through the activity of polyphenol oxidase (PPO), the same enzyme central to Paper I's oxidative domain (Reservoir C).
One study found that brewing temperature and time can be adjusted to influence sensory acceptability: higher brewing temperatures were associated with greater phenolic extraction, which in turn affects astringency, bitterness, and overall flavour complexity. In that study, brewing at approximately 95°C for around 5 minutes was associated with favourable consumer perceptions of taste attributes including bitterness and astringency. A related finding was that higher phenolic content in processed leaf correlated with both antioxidant activity (a chemical measurement) and sensory acceptance (a panel's preference) — though the same research notes that excessive phenolic extraction can tip into undesirable bitterness, so "more phenolics" is not simply "better."
Brewing at ~95°C for ~5 minutes was associated with favourable consumer ratings for bitterness/astringency in the cited study. This aligns with Paper II, Section 22's brewing guidance (90°C, 5 minutes as a starting recommendation) — both point to a similar range, from different sources.
"Higher phenolic content correlates with sensory acceptance, up to a point" is a generalisation from a specific study's panel data. Leaf age, processing, and the brewer's own preferences (per Paper II, Section 22's "starting recommendation" framing) all interact with this — this page treats it as one data point among the brewing guidance Paper II already provides, not as a separate or competing recommendation.
Sources: Fibrianto et al. (2025), as below; Paper II, Section 22 (Brewing Parameters).
Exploratory Concept
One of the sources behind this page (Fibrianto, Bimo, Wulandari & Hendrawan, 2025) is a small consumer-panel study comparing oolong-style coffee leaf tea against other coffee leaf and Camellia sinensis tea styles. Because some of this study's findings have previously appeared, in this library and elsewhere, in language about stress relief, mood, and relaxation, this section describes what the study actually did and measured — in its own terms — so the difference between "what was measured" and "what it means for a cup of tea" stays visible.
Trained panellists drank samples of different coffee leaf and tea preparations under controlled conditions, after a standardised stress-inducing task (a timed arithmetic exercise). The study then measured: salivary and blood cortisol (a hormone used as a stress-physiology marker), self-reported relaxation on a numeric scale over a 45-minute window, GABA content of the brewed tea itself (measured chemically, not in the panellists), total phenolic content, and facial expression analysis via camera and software (Facial Emotion Recognition).
The oolong-style coffee leaf sample was associated with the largest reduction in measured salivary cortisol among the samples tested, the highest measured GABA and phenolic content among the coffee leaf samples, and self-reported relaxation scores comparable to a 200mg GABA supplement used as a reference sample in the same study.
This is a single study, with a trained panel of a size typical for sensory research, measuring specific physiological markers under specific controlled conditions (including a standardised stressor immediately beforehand). That is meaningful as a research finding — it is the kind of result that justifies further research. It is a different kind of statement from "drinking this tea will reduce your stress," which generalises a controlled measurement (cortisol in trained panellists, after a specific stressor, within a 45-minute window) into an everyday claim the study itself doesn't make.
GABA content of the brewed tea is a chemical measurement and is reported accurately in Section Two above. What GABA in a beverage does once consumed — whether and how much crosses into measurable physiological effect — is a separate question that this single study's design (measuring panellists' cortisol and self-report alongside the tea's GABA content) does not, on its own, fully resolve. Both are real findings; they answer different questions, and conflating them is the most common way this kind of research gets overstated.
Source: Fibrianto, Bimo, Wulandari & Hendrawan (2025), Functional and relaxing properties of coffee leaf tea: An integrative food cognition approach, Applied Food Research, 5, 101141.
Rigling, M., Steger, M. C., Lachenmeier, D. W., Schwarz, S., & Zhang, Y. (2023). Sensorial and Aroma Profiles of Coffee By-Products — Coffee Leaves and Coffee Flowers. Proceedings 2023, 89, 5. DOI: 10.3390/ICC2023-14837. CC BY 4.0. Used for the fermented/non-fermented sensory spectrum and the five-compound aroma profile (Sections One and Two).
Fibrianto, K., Bimo, I. A., Wulandari, E. S., & Hendrawan, Y. (2025). Functional and relaxing properties of coffee leaf tea: An integrative food cognition approach. Applied Food Research, 5, 101141. Used for brewing-parameter findings (Section Three) and described directly, in its own terms, in "A Note on the Research."
See also: Paper I (The Reactive Landscape) for the chemistry underlying these compounds; Paper II, Section 22, for brewing guidance; Paper III, "The Citane Compass," for these compounds as starting points for processing experiments.