A soft, wearabe microlfluidic device for the capture, storage, and colorimetric

sensing of sweat

Application
The goal of this study was to investigate the effectiveness of stretchable
microfluidic systems that can be worn on the skin and analyze perspiration using
colorimetric sensing methods to detect biologic markers such as chloride,
hydronium ions, glucose, and lactate. This report describes the results of human
trials and presents data regarding to these markers, as well as the rate at which the
people sweated, the total sweat loss, as well as the pH of the sweat.

Reaction
These devices utilize colorimetry to determine changes in chemical conditions in
sweat. Different reactions occur for the different markers being examined, which will
offer color changes that can be quantified and correlated with concentrations using
colorimetry. An example of this is using anhydrous cobalt (II) chloride to form
hexahydrate cobalt chloride, which will result in a change of color from deep blue to
purple. Individual detection reservoirs that produce different changes in color of the
materials used are utilized for each of the markers, with one being present for pH,
Lactate, Chloride, Glucose, and water.

Sensitivity
Because this device utilizes a cell phone camera in order to detect color changes
there are limitations on how accurate sensitive the readings can be. Though a
higher sensitivity could be achieved by using a more sophisticated sensor to make
colorimetric readings, the following values for sensitivity were given. For
concentration measurements of the biomarkers, the increments that were able to
be evaluated started at a baseline, and the next detectable increase corresponded
to double the initial detectable value. For lactate, the initial detectable
concentration was 1.6mM, with other detectable concentrations corresponding to
the doubles of the values, up until 100mM. All other markers followed this same
pattern with an initial detectable concentration, which are as follows: Glucose
1.6mM, Creatinine 15.6M, Chloride 39mM. The exception to this was pH, which
was only detectable in increments of 0.5.

Detection Limits
Because these rely on colorimetric techniques that are based on the absorbance of
light in the materials, there is an upper saturation limit for these materials when the
concentration of the marker being analyzed is too high to measure. This varies for
the different markers, but is given in the source material as being 100mM for
lactate, 25mM for glucose, 1mM for Creatinine, 625mM for Chloride, and a range of
pH from 5 to 8.5 (in 0.5 increments)
Accuracy
The sweat was tested against a reference method by which sweat was obtained by
the subjects. This was processed then analyzed using a Sigma-Aldrich colorimetric
analysis kit. The pH was determined using a Hanna Instruments micro pH meter.
Based on Figure F on page 9, the P value, which is related to the differences
between the two results, was 0.34 for chloride, <0.05 for glucose, 0.53 for lactate,
and 0.55 for pH.

Specificity
All of the chemical reagents that produce the color changes for colorimetric analysis
are believed to be highly selective towards the biomarkers that are being measured.
It is possible that there are other chemicals present in sweat that could impact the
Reproducibility
Though these devices are considered to be accurate and reproducible, there were
external conditions that are believed to impact them. Some of the external
conditions that were mentioned were humidity, temperature, and UV index.
However, all of these parameters were recorded, and it is believed that by
accounting for their impact, it would be possible to reproduce the results in a
reliable manner. The type of light also needs to be considered, as different results
were obtained when operating the device in different types of light. After
adjustments, the results were normalized, though some variation still existed.
Stability
The report discussed that these devices were used in high intensity conditions
without failure. After being removed, they were able to be stored for 125 hours after
removal without being compromised, however that is after having been sealed.
Without being sealed, they were found to last about 75 hours after removal. The
main concern is a degradation in the ability to conduct a colorimetric analysis, but it
was found that within this range, this was negligible.