Basics

Lexicon

• rH : relative Humidity (measured by humidity sensor)
• delta rH or ▲rH : sample’s delta rH. See dedicated paragraph for more details

Principle

To reduce the effect of water on VOC for a specific NeOse Advance, water’s odor is substracted to samples’ odor.

First we compute water’s affinity of this device:

Then, each VOC’s signatures is corrected as follow:

This requires:

• a water sample in the analyzed data ( at least 3 replicats);
  • the water sample must be from more or less the same date as the other samples analyzed data (Aryballe recommends +/- 24h);
  • the water sample must be from the same device and the same Core Sensor;
• For VOCs except water, rH is between 20% and 60%. Otherwise results are not garanteed.

Computation of a sample's▲rH

Here is a sensogram for a specific sample:

Captions

• colored lines: biosensors' response to the selected sample and water, a.k.a. sensogram
• grey zone: selected baseline timeframe
• light blue zone: selected analyte timeframe
• pink dashed line: rH measured by humidity sensor
• line with the water drop: the value of this sample's ▲rH which is used for the humidity compensation computation and when this value happens first.

The sample's ▲rH is computed following this rule: average humidity over the baseline timeframe, minus the highest value (or lowest value) within the analyte timeframe. Moreover, if the sample is dryer than the baseline, the delta rH <0, and then the water's odor will be added to the sample's odor.

Note that the humidity sensor (aka HIH) is usually late compared to the sensorgam because:

1. Slow convergence rate of HIH sensor;
2. Distance between the Core Sensor and HIH sensor (NOTE that the HIH sensor is placed after the Core sensor).

That’s why a long analyte is used in the example in the following paragraph.

Humidity compensation in DOH

Once a generated session is opened and analyte zones selected, click on Settings (double cogs icon):

The setting panel is then displayed. Select your water acquisition as the reference :

All results displayed in the session (expects sensogram and Intensity diagrams) are immediately affected.

Example

The following dataset contains 7 samples defined as such:

• Sample named “Gold X”  - 500 µL
• Sample named “Gold X”  - 500 µL + H2O 500 µL
• H2O 500 µL

In this example, here is the selected dataset:

And here is the average biosensors responses for each sample without any correction:

We can visualize water’s signature:

Once Humidity compensation is applied with “Water” as the calibrant:

“Water” signatures is no longer displayed and is no longer in the PCA either: