Salts as Additives: A Route to Improve Performance and Stability of n-Type Organic Electrochemical Transistors

Organic electrochemical transistors (OECTs) are becoming increasingly ubiquitous in various applications at the interface with biological systems. However, their widespread use is hampered by the scarcity of electron-conducting (n-type) backbones and the poor performance and stability of the existing n-OECTs. Here, we introduce organic salts as a solution additive to improve the transduction capability, shelf life, and operational stability of n-OECTs. We demonstrate that the salt-cast devices present a 10-fold increase in transconductance and achieve at least one year-long stability, while the pristine devices degrade within four months of storage. The salt-added films show improved backbone planarity and greater charge delocalization, leading to higher electronic charge carrier mobility. These films show a distinctly porous morphology where the interconnectivity is affected by the salt type, responsible for OECT speed. The salt-based films display limited changes in morphology and show lower water uptake upon electrochemical doping, a possible reason for the improved device cycling stability. Our work provides a new and easy route to improve n-type OECT performance and stability, which can be adapted for other electrochemical devices with n-type films operating at the aqueous electrolyte interface.

of an overall smoother surface.Additionally, the oscillatory behavior of the Minkowski connectivity curves has previously been used to describe randomly and poorly connected valleys/peaks. [1]The lower strength of the oscillatory behavior for P75-TBAClO 4 and P75-LiClO 4 indicates a more interconnected pore network than for the other films.This is in line with the generally improved ion transport observed for these films (e-QCMD).

Figure S1 .
Figure S1.Output and transfer curves, and the transconductance vs. gate voltage plots.Arrows indicate the scan direction.Transfer curves were obtained at V D = 0.6 V.The scan rate was 100 mV/s.Error bars represent the standard deviation measured from at least six different devices.Average channel thicknesses were 60 nm, 130 nm, 175 nm, and 180 nm for P75, P75-TBAClO 4 , P75-TBAPF 6 , and P75-LiClO 4 , respectively.

Figure S2 .
Figure S2.UV-VIS-NIR absorption spectroscopy of as cast thin films on ITO.Thin films have an average thickness of 50 nm.

Figure S3 .
Figure S3.The dependence of film capacitance on its geometry.The capacitance was determined from the electrochemical impedance spectra collected at 0.1 Hz at a DC voltage of -0.5 V vs. Ag/AgCl.Error bars represent the standard deviation over at least 3 different devices per geometry.

Figure S5 .
Figure S5.Operational stability of P75, P75-TBAClO 4 , P75-TBAPF 6 , and P75-LiClO 4 .The channel current stability against pulsed gate voltages (V G = 0.4V) applied for 2 hours.V G varied between 0 and 0.4 V, with 10-second intervals.V D was fixed at 0.4V.Note that the arrow indicates the addition of fresh 0.1M NaCl electrolyte.Inset presents the variation of the current in the first 500 seconds of pulsing.I 0 is the drain current corresponding to the first pulse.

Figure S8 .
Figure S8.Influence of oxygen on OECT gate current (I G ).The currents were measured at V G = 0.6 V and V D = 0.6 V. Error bars represent the standard deviation over at least six different devices.

Figure S9 .
Figure S9.Influence of oxygen on cyclic voltammograms.Arrow indicates the scan direction.Scan rate is 50 mV/s.The average film thickness is 50 nm.

Figure S10 .
Figure S10.High-resolution Li 1s XPS spectra of P75-LiClO 4 as cast and exposed films.The peak observed here overlaps with (or corresponds to) the Au 5p 3/2 peak which has a typical binding energy of 57.2 eV.

Figure S13 .
Figure S13.Secondary ion mass spectrometry data showing F -, P -, and Cl -signals in as-cast and exposed thin films.

Figure S17 .
Figure S17.In-plane (q r ) line cut of P75-TBAClO 4 as cast and exposed thin films.The disappearance of the salt peak at 0.61 A -1 shows that the salt species leave the film in the testing conditions.

Figure S18 .
Figure S18.The height distribution curve for each AFM image.The height distribution size (Xaxis) represents the relative size feature of each individual scan due to the lack of a common zero baseline.The pristine polymer displays a reversible surface distribution, while the average height and width distribution decreased for the TBA-salt polymers and increased for P75-LiClO 4 after exposure and electrochemical doping.Note that the height distribution size (X-axis) represents the relative size feature to each scan due to the lack of a common zero baseline.

Figure
S21 represents the character of the Minkowski connectivity function distribution.The Minkowski connectivity (χ) curves for all the polymers present an oscillatory form.Valleys are indicated by negative values where the minimum corresponds to the highest density of valleys.Conversely, positive values represent peaks, and the maximum is the highest density of peaks.P75-TBAClO 4 and P75-LiClO 4 Minkowski connectivity curves are both dominated by negative values, indicating the dominance of valleys (pores) over peaks on the surface.Alternatively, the pristine P75 shows a very sharp transition around 20 nm, characteristic