Understanding the effect of carbon in carbon/salt/adhesive electrodes for surface electromyography measurements

• In general, salt/adhesive (SA) electrodes are suitable to collect sEMG signals in a similar manner to carbon/salt/adhesive (CSA) electrodes.
• Correlation of sEMG signals between CSA and SA electrodes was high in time (≥ 0.87) and frequency (≥ 0.94) domain.
• CSA electrodes, which contain carbon in the adhesive layer, provide a better capability to manage motion, but are more susceptible to noise corruption and are less sensitive to myoelectric activity.
• SA electrodes, made without adding carbon to the adhesive layer, specifically a mixture with 15% salt, provided a better response to myoelectric activity and seem to be the most suitable alternative for sEMG data collection.

AbstarctDry electrodes that do not require silver and hydrogel might provide the advantages of having very long shelf life and lower cost, compared to the gold standard Ag/AgCl hydrogel electrodes. Recently, we compared novel carbon/salt/adhesive (CSA) electrodes with Ag/AgCl electrodes for surface electromyography (sEMG) signal collection. We found no significant differences in amplitude, but CSA electrodes outperformed Ag/AgCl in response to noise and motion artifacts. However, the carbon component may be redundant, and the salt/adhesive (SA) mixture might be as effective as CSA for such a task. In the SA electrodes, the salt concentration is the only tunable factor. To determine if carbon contribution is necessary for effective sEMG measuring capabilities, we varied the salt concentration in the SA electrodes to 10%, 15%, and 25% and their performance was compared to the functional capabilities of CSA electrodes. Twenty subjects were recruited to collect simultaneous recordings of sEMG signals using CSA and SA electrodes, side-by-side on triceps brachii, tibial anterior muscles, biceps brachii and quadriceps femoris. SA 15% and SA 25% electrodes detected higher amplitude values during contraction in biceps, tibials and quadriceps, compared to CSA. All SA electrodes exhibited high mean correlation with CSA electrodes, on the linear envelopes (≥ 0.887), RMS envelope (≥ 0.87) and power spectrum density (≥ 0.94). SA 15% and SA 25% electrodes performed better in response to noise and were more sensitive to myoelectric activity than CSA electrodes, but CSA electrodes exhibited better response to motion artifacts than SA electrodes. SA 10% electrodes presented high electrode-skin impedance, producing some lower values in sEMG signals during contraction, worse motion corruption and spectral deformation compared to CSA. Results suggest that carbon improves capability to manage motion, but at the expense of more susceptibility to noise corruption. Higher salt concentration reduced motion artifacts and spectral deformation, but reduced the sensitivity to myoelectric signals. In conclusion, SA electrodes, specifically the mixture with 15% salt, provided a better response to myoelectric activity and seem to be the most suitable alternative for sEMG data collection.