Overview
Noor and coauthors compared Merapi Robusta coffee roasted by traditional, semi-modern, and modern methods at two roasting durations (20 and 25 min). The primary study focus is quality and sensory performance; the three best-scoring hedonic samples (sample 1: traditional 20 min; sample 3: semi-modern 20 min; sample 6: modern 25 min) were then tested by ICP-MS for Pb and Cd and by LC-MS/MS for ochratoxin A. All three samples were below the limits of detection for all three analytes.
Key numbers
Table 2 reports SNI 01-3542-2004 ground-coffee compliance values across the six roasting conditions:
| Criterion | Reported range | Meets SNI |
|---|---|---|
| Water | 0.13 +/- 0.02% to 0.29 +/- 0.01% w/w | yes |
| Coffee extract | 30.70 +/- 0.15% to 36.59 +/- 0.02% w/w | yes |
| Caffeine (anhydrous) | 1.32 +/- 0.00% to 1.52 +/- 0.01% w/w | yes |
Table 4 reports the metal contamination and ochratoxin screening on the three best hedonic samples (samples 1, 3, 6) with the analytical limits of detection:
| Analyte | Unit | Sample 1 | Sample 3 | Sample 6 | LOD |
|---|---|---|---|---|---|
| Lead (Pb) | mg/kg | Not detected | Not detected | Not detected | 0.00025 |
| Cadmium (Cd) | mg/kg | Not detected | Not detected | Not detected | 0.0005 |
| Ochratoxin A | mg/kg | Not detected | Not detected | Not detected | 1.01 |
The methods section states that Pb and Cd were tested by ICP-MS and ochratoxin A by LC-MS/MS.
Methods (brief)
The study used Grade A Merapi Robusta green coffee from the Tunggak Semi Farmers Group in Cangkringan, Sleman. Six roasting conditions were tested: traditional, semi-modern, and modern roasting for 20 and 25 minutes. Pb and Cd were measured by ICP-MS in selected top-performing samples; ochratoxin was measured by LC-MS/MS.
Implications
Certification: Three-sample left-censored occurrence evidence for ground Robusta coffee from one Indonesian origin (Sleman, Yogyakarta). Pb and Cd LODs (0.25 ppb and 0.5 ppb in mg/kg terms) are sufficiently sensitive that left-censored values would not bias a pooled occurrence percentile if treated as LOD/2 or zero in synthesis.
Courses: Useful for teaching that quality and sensory studies frequently contain contaminant screens, and that when LODs are reported a non-detect carries quantitative meaning (an upper bound).
App: Supports a low-confidence non-detect note for single-origin Indonesian Robusta only; insufficient on its own to anchor a contamination profile.
Related evidence
Update history
No substantive edit history is available in this build. The full commit record is available in git.