Ethanol Interference in Formaldehyde Sensors: What the Evidence Supports

Scope of the review

The targeted search examined accessible primary studies, supporting information and a laboratory evaluation report available through 10 September 2026. It also considered model-based estimates and numerical examples in public patent documents. We checked what was measured, how the reference concentration was obtained, which gases were present and whether percentages described concentration errors or other metrics.

Three selected sources illustrate the findings below. This was a targeted, non-systematic search: some full texts and underlying paired measurements were unavailable, and the result does not establish that no matching observation exists. Their different instruments and test conditions do not support a single error percentage for formaldehyde sensors as a class. This note reports no new CoastalQuant sensor measurements and has not undergone journal peer review.

What “above 200%” means

For a positive reference concentration, the signed relative concentration error is:

relative error = (indicated HCHO − reference HCHO) / reference HCHO × 100%, with reference HCHO > 0.

An error of +200% means that the indicated concentration is three times the reference. An error strictly above +200% requires a reading more than three times the reference. A reading described as “200% of the true value” instead corresponds to a +100% error.

A formaldehyde-equivalent response to ethanol is a different quantity. It expresses the signal produced by ethanol on an instrument’s formaldehyde scale. If no nonzero formaldehyde reference is present, that response alone cannot establish a percentage error in formaldehyde concentration. A model prediction also needs separate identification: a large prediction error is not automatically a large observed overreading by the instrument.

Three sources, three different interpretations

A measured ethanol cross-response

Won and colleagues introduced interfering gases with formaldehyde and verified formaldehyde by DNPH sampling and HPLC. Figure 6(a) includes an independent reference near 0.045 ppm. For the electrochemical sensor COTS_1, the fitted slope of 0.3308, approximately 33%, describes the increase in formaldehyde-equivalent response per unit ethanol concentration. It is not relative concentration error against HPLC. [1]

At the highest plotted ethanol challenge, approximately 0.1 ppm as inferred from the plotted response and fitted line, the mean reading is about 0.066–0.067 ppm, roughly 50% above the reference. Without ethanol, the plotted mean is about 0.030 ppm, approximately one-third below the reference. These are our approximate graph readings, not author-reported percentages or uncertainty intervals. The total error against HPLC differs from the increase relative to the ethanol-free baseline. Within this approximate ethanol range, none of the plotted means reaches +200% error; the figure provides no measurements at higher ethanol concentrations. Panel (b) concerns limonene and cannot establish the magnitude of ethanol interference. [1]

Interference in the reference instrument

Van den Broek and colleagues developed a portable sensor with a separation column. Their supporting information also documents ethanol-related interference in the proton-transfer-reaction time-of-flight mass spectrometer (PTR-ToF-MS) used as a reference instrument. Ethanol fragmentation contributes to the apparent formaldehyde signal, and the authors describe correcting this contribution. [2]

Supporting Figure S3 therefore needs to be read as evidence about the reference instrument. It does not establish an ethanol-induced error above 200% in the portable sensor. Converting a response measured during an ethanol exposure into such a percentage would require a nonzero formaldehyde reference for that exposure, which this test does not provide. [2]

A large discrepancy in a mixed-gas evaluation

Sidheswaran and colleagues describe the Interscan mixture readings as roughly twice the DNPH values. Reading the centres of Figure 7’s five unscrubbed points, we find only the lowest above +200%; the other four suggest approximately +40% to +160%. The lowest is about 5 ppb by DNPH and 36 ppb by the monitor. These are our graphical estimates, not author-reported percentages or assessments of error bars. [3]

The lowest point lies below the lowest nonzero calibration level of 10 ppb. The report discusses drift; raw paired data and full uncertainty were unavailable to this review. Tests at 45–50% relative humidity included 20–30 ppb ethanol among other volatile organic compounds. The scrubber removed both these compounds and water, leaving ethanol’s contribution unresolved. [3]

Table 1 gives a manufacturer ratio of 127 ppb ethanol per 1 ppb formaldehyde-equivalent response. If that coefficient and linear scaling transfer to these conditions, our calculation predicts about 0.2 ppb for 20–30 ppb ethanol. This conditional estimate is not a measured contribution. The authors describe interference exceeding manufacturer-based predictions for the mixture, without attributing it to ethanol. [3]

Implications for measurement validation

The evidence supports including ethanol in the CoastalQuant HCHO Research validation program. Planned testing should compare formaldehyde-only conditions with matched formaldehyde-plus-ethanol conditions using an independent reference method. Concentrations, temperature, humidity, calibration range and measurement uncertainty need to be stated. Checks on the reference method itself are also necessary.

Sources

1. Won et al. (2011). Characterizing Commercially Available Formaldehyde Sensors. (opens in a new tab) Indoor Air 2011 conference proceedings; full text in the NRC Publications Archive. Figure 6(a–b). NRCC 54466; NRC-IRC-21726. NRC record. (opens in a new tab)
2. van den Broek et al. (2020). Selective formaldehyde detection at ppb in indoor air with a portable sensor. (opens in a new tab) Journal of Hazardous Materials, 399, 123052. DOI: 10.1016/j.jhazmat.2020.123052. (opens in a new tab) Supporting information, Figure S3 (DOCX). (opens in a new tab) Open published version: ETH Zurich repository, 10.3929/ethz-b-000418226. (opens in a new tab)
3. Sidheswaran et al. (2013). Performance Evaluation of Real Time Formaldehyde Monitors: PTR-MS and Interscan 4160-500B Portable Monitor. (opens in a new tab) Berkeley Lab technical report, July 2013. LBNL-6357E; OSTI 1165367. Figure 7 and Tables 1–2. OSTI record. (opens in a new tab) DOI: 10.2172/1165367. (opens in a new tab)

Change history

1. v1.3 · 10 September 2026

   First publication.