BGPT: Paper Review: Extraction of arsenic compounds from lichens

Explore by Goal

Quick Explanation Copied

Core takeaway (skeptical, evidence-first)

The paper shows that common single-step extractants recover ≤~23% of arsenic from lichens, while a soil-style sequential extraction can recover substantially more arsenic—about 45% in Hypogymnia physodes and 83% in Cladonia rei—suggesting a large “soil-particle / soil-like” arsenic pool and that lichens can be chemically complex rather than passive monitors.

Evidence:

Long Explanation

Paper Review (science-focused): Extraction of arsenic compounds from lichens

Talanta • 2005 • DOI: 10.1016/j.talanta.2005.10.011

1) What the paper set out to do (and what it did)

Problem: arsenic speciation in lichens is limited by low extraction efficiency and by lichens being complex matrices (including trapped particles).
Approach: test multiple extractants/conditions (water at 20/60/90°C, methanol/water mixtures, Tris buffer, acetone), then apply a sequential extraction scheme adapted from soils.
What was measured: total As by k0-INAA, total As in filtered extracts by ICP-MS after digestion, and arsenic speciation by HPLC-UV-HGAFS, with additional lipid-soluble assessment via diethyl ether + INAA.

2) Key results (what changed when extraction changed)

2.1 Extraction efficiencies for common extractants (single-step)

The paper reports extraction efficiencies (%) calculated by comparing total As in filtered extracts (ICP-MS) versus total As in whole material (k0-INAA).

Observed pattern: the best single-step efficiencies reported are ~19.2% (water 60°C) for H. physodes and ~22.4% (Tris) for C. rei, with some entries reported as “nd” (not detected) or “null” in the figure.

2.2 Sequential extraction recovers much more arsenic

The authors report sequential extraction (soil scheme based on Wenzel et al.) recovers about 45% of total As from H. physodes and 83% from C. rei.

Interpretation (as claimed by authors): sequential extraction results are consistent with a large portion of arsenic being associated with trapped soil particles inside lichens, and therefore “soil-like” arsenic dominates the measurable pool.

3) Speciation: what arsenic forms were reported

3.1 Species detected in lichen extracts (and one phosphate-ribose note)

The paper reports detection of: arsenous acid (As(III)), arsenic acid (As(V)), monomethylarsonic acid (MMAA), dimethylarsinic acid (DMAA), arsenobetaine (AsB), trimethylarsine oxide (TMAO), glycerol-ribose.
Additionally, phosphate-ribose was reported in Hypogymnia physodes.

3.2 A critical analytical caution: matrix effects & incomplete accounting

The paper states that chromatographic separations suffered from coelution/matrix effects (e.g., glycerol-ribose coeluting with multiple species in the anion exchange chromatogram; cation exchange also had partial coelution).
It also notes that even within “best” extract conditions, some arsenic measured as total As in extracts could remain unassigned or invisible to the detection system used—potentially ~half of total arsenic in extracts.

Skeptical implication: “missing arsenic” and coelution mean the chemical speciation picture is directionally informative but may not be a complete census of all arsenic species present in the true biological matrix.

4) Mechanistic claim vs. measurement reality (where uncertainty lives)

4.1 Main mechanistic claim

The authors propose that a large fraction of arsenic in lichens is “soil-like” and associated with trapped particles, because sequential extraction adapted from soils recovers far more arsenic than common biological extractants.

What could disprove/complicate it? the same data also shows that even after sequential extraction, the system may still not recover all arsenic species; therefore, “soil-like” association could be partly conflated with extraction detectability limits and chromatographic recovery biases.

4.2 Sample-size and locality limits

Only two lichen species were analyzed (one epiphytic and one terricolous), and only one representative soil sample was used.
The soil was from a mine-impacted area near a former lead–zinc mine/smelter; these results may not generalize to other arsenic sources (e.g., different mineralogy or organic arsenic profiles).

5) Reproducibility & methodological transparency (what seems solid vs. fragile)

5.1 What looks reproducible from the provided text

Detailed analytical workflow is described: k0-INAA irradiation conditions, extraction procedure steps (freeze-dry, liquid nitrogen, shaking bath), and ICP-MS calibration strategy (external calibration, checks after sample blocks).
They performed duplicate extractions and report RSDs for ICP-MS and speciation where possible.

5.2 Fragility / bias risks inside the measurement chain

Matrix effects and coelution can bias species assignment and the inferred concentrations of specific arsenosugars.
Spiking artifacts are acknowledged: the paper reports ~30% losses for AsB during spiking and interprets the lack of TMAO increase as not confirming AsB→TMAO degradation under those conditions.

6) Bottom line

Best-supported claim

The central empirical message is robust: single-step extractants underperform (max ~19–22% extraction efficiency), while sequential extraction increases recovery substantially (to ~45–83%), and the measured arsenic species inventory includes both inorganic and organic arsenicals plus arsenosugars.

Most important uncertainty

The mechanistic inference that much of the arsenic is “soil-like” is plausible and supported by the sequential extraction success, but it is not uniquely determined because (i) the species accounting is incomplete due to detectability/binding and (ii) matrix effects/coelution alter quantification and potentially affect interpretation of specific arsenosugars.

7) Author review links (bespoke BGPT queries)

Feedback:

Updated: April 01, 2026