Ionic transistors from organic and biological materials hold great promise for bioelectronics applications. Thus, much research effort has focused on optimizing the performance of these devices. Herein, we experimentally validate a straightforward strategy for enhancing the high to low current ratios of protein-based protonic transistors. Upon reducing the thickness of the transistors’ active layers, we increase their high to low current ratios 2-fold while leaving the other figures of merit unchanged. The measured ratio of 3.3 is comparable to the best values found for analogous devices. These findings underscore the importance of the active layer geometry for optimum protonic transistor functionality.

Ionic transistors from organic and biological materials represent an emerging class of devices for bioelectronics applications.1–9 Indeed, the processing and fabrication techniques required for the preparation of these transistors are simple, convenient, and inexpensive.1–5 The constituent organic or biological materials are also amenable to chemical modification and functionalization.1,5–7 In addition, the mechanical properties of organic materials are inherently compatible with those of biological systems.1,3,8 Moreover, organic and biological ionic conductors are well suited for the transduction of biochemical events into electronic signals.1,4–7,9 These key advantages have made ionic transistors from organic and biological materials exciting targets for further research and development.

Within the broader ionic transistor class of devices, there have been several reports of protonic transistors,10–16 including two notable examples from the Rolandi group.13,14 For these devices, the application of a voltage to the gate modulates the current flow between the source and the drain, in analogy to conventional unipolar field effect transistors.13,14 The magnitude of the current is determined by the proton charge carrier density in the device channel, as given by the equation nH+ = nH+0 − VGSCGS/et, where nH+ is the proton concentration at an arbitrary gate voltage, nH+0 is the proton concentration at a gate bias of 0 V, VGS is the gate voltage, CGS is the gate capacitance, e is the charge of the proton, and t is the thickness of the active layer.13,14 Thus, a negative gate voltage induces the injection of protons into the channel, leading to an increase in the source-drain current, and a positive gate voltage depletes the channel of protons, leading to a decrease in the source-drain current.13,14 This operating mechanism enforces limits on the ratio between protonic transistors’ high (on) and low (off) currents (IHIGH/ILOW), relative to standard field effect transistors. However, the IHIGH/ILOW ratio, in principle, can be improved by reducing the active layer thickness, thereby increasing the difference between the transistors’ high and low current states.

Recently, our group discovered that reflectin, a structural protein that plays a key role in the color-changing abilities of cephalopods,16–20 is an effective proton conducting material.21 This finding enabled the fabrication of protein-based protonic transistors with excellent figures of merit, including a proton mobility (μH+) of ∼7.3 × 10−3 cm2 V−1 s−1.21 However, due to active layers with thicknesses between ∼1 and ∼2 μm, the transistors possessed relatively poor IHIGH/ILOW ratios of ∼1.6.21 

Herein, we build upon our previous work and demonstrate improved IHIGH/ILOW current ratios for reflectin-based protonic transistors. We first fabricate two-terminal devices from thin reflectin films and characterize their conductivity when contacted with palladium (Pd) and palladium hydride (PdHx) electrodes. We next describe the electrical interrogation of reflectin films with an average thickness of ∼0.30 μm in a three-terminal transistor configuration. We find that the majority of the device metrics, including mobility and proton concentration, are comparable to those previously reported for protonic transistors from reflectin films with a thickness between ∼1 and ∼2 μm.21 However, we observe a 2-fold improvement in the thin protonic transistors’ IHIGH/ILOW current ratios. Overall, our findings highlight the importance of the active layer geometry for the performance of protein-based (and other) protonic transistors.

We began our studies by heterologously expressing a histidine-tagged Doryteuthis (Loligo) pealeii reflectin A1 isoform in E. coli according to previously reported protocols.20,21 We first extracted crude reflectin from E. coli inclusion bodies. The protein was then sequentially purified by immobilized metal affinity chromatography under denaturing conditions and high performance liquid chromatography (HPLC). The identity of the purified protein was definitively confirmed by in-gel tryptic digestion and tandem mass spectrometry. Notably, the optimized expression and purification procedure yielded >800 mg of reflectin per liter of E. coli cell culture with a purity of over 99%. This high yield and excellent purity enabled the high throughput fabrication of reflectin-based devices.

For our measurements, we fabricated two-terminal bottom-contact devices according to the scheme shown in Fig. 1.22 In brief, we deposited an array of palladium contacts onto the surface of clean silicon dioxide/silicon (SiO2/Si) substrates via electron-beam evaporation through a shadow mask. Next, we dropcast aqueous reflectin solutions onto the electrodes. The solvent was then allowed to evaporate, and the excess protein was removed from the substrate via mechanical scribing. The resulting completed devices were subjected to physical and electrical characterization.

FIG. 1.

Illustration of the fabrication of two-terminal reflectin-based devices. Palladium contacts are first deposited onto the surface of a silicon dioxide/silicon (SiO2/Si) substrate. Next, reflectin is dropcast directly onto the electrodes from aqueous solution. After drying, excess material is removed via mechanical scribing to furnish the completed device.

FIG. 1.

Illustration of the fabrication of two-terminal reflectin-based devices. Palladium contacts are first deposited onto the surface of a silicon dioxide/silicon (SiO2/Si) substrate. Next, reflectin is dropcast directly onto the electrodes from aqueous solution. After drying, excess material is removed via mechanical scribing to furnish the completed device.

Close modal

We first characterized our devices with both optical microscopy and atomic force microscopy (AFM).22 A typical optical image of a reflectin device is shown in Fig. 2(a). The optical image demonstrated that the film was uniform, with few apparent defects. A typical corresponding AFM image is shown in Fig. 2(b). The image indicated that the surface topography of the reflectin film was relatively smooth and featureless with a root mean square (RMS) roughness of ∼0.5 nm. The devices were now poised for electrical interrogation.

FIG. 2.

(a) A typical optical image of a device where a thin reflectin film bridges two palladium electrodes. (b) A representative AFM image of the reflectin film from (a). The film is relatively smooth and featureless with an RMS roughness of ∼0.5 nm.

FIG. 2.

(a) A typical optical image of a device where a thin reflectin film bridges two palladium electrodes. (b) A representative AFM image of the reflectin film from (a). The film is relatively smooth and featureless with an RMS roughness of ∼0.5 nm.

Close modal

We investigated the electrical properties of the reflectin films by recording current (I) as a function of voltage (V) at a relative humidity of 90%.24 The I-V characteristics of a typical reflectin-based device, when contacted with palladium electrodes, are shown in Fig 3. This 0.24 μm-thick device featured a low current density of 0.7 × 10−2 A/cm2 at 1.5 V, consistent with previous findings for ∼1 to ∼2 μm-thick reflectin films contacted with proton-blocking electrodes.21 Subsequently, we converted the device’s electron-conducting Pd contacts into proton-injecting PdHx contacts via exposure to hydrogen gas in situ (Fig. 3(a)). As illustrated in Fig. 3(b), the current density of the device increased by nearly an order of magnitude to 5.9 × 10−2 A/cm2 at 1.5 V, in agreement with literature precedent.13,21 These measurements confirmed the presence of protonic conductivity for thin reflectin films.

FIG. 3.

(a) An illustration of a two-terminal reflectin-based device, which undergoes in situ treatment with hydrogen (H2) gas. The palladium electrodes are converted to palladium hydride electrodes, enabling the injection of protons into the film. (b) The current versus voltage characteristics of a reflectin film contacted with palladium (red) and palladium hydride (blue) electrodes. The device current increases by an order of magnitude upon switching from palladium to palladium hydride electrodes in situ. Both the forward and reverse scans are shown for each measurement. The measurements were performed at a relative humidity of 90%. The device had a length of 50 μm, a width of 280 μm, and a thickness of 0.24 μm.

FIG. 3.

(a) An illustration of a two-terminal reflectin-based device, which undergoes in situ treatment with hydrogen (H2) gas. The palladium electrodes are converted to palladium hydride electrodes, enabling the injection of protons into the film. (b) The current versus voltage characteristics of a reflectin film contacted with palladium (red) and palladium hydride (blue) electrodes. The device current increases by an order of magnitude upon switching from palladium to palladium hydride electrodes in situ. Both the forward and reverse scans are shown for each measurement. The measurements were performed at a relative humidity of 90%. The device had a length of 50 μm, a width of 280 μm, and a thickness of 0.24 μm.

Close modal

We next studied the electrical properties of reflectin films featuring an average thickness of 0.30 (±0.06) μm in a three-terminal transistor configuration (Fig. 4(a)).22 The I-V characteristics of a typical protonic transistor are shown in Fig. 4(b). We found that the protonic source-drain current (ISD) in these devices, measured at different source-drain voltages (VSD), was dictated by the applied gate voltage (VGS). The protonic current decreased upon changing the VGS from +10 V to 0 V to −10 V at a relative humidity of 90% (Figs. 4(a) and 4(b)). This gating behavior was consistent with previous studies of maleic chitosan-based and reflectin-based protonic transistors.13,14,21

FIG. 4.

(a) Illustration of the protonic current for a protonic transistor under different applied gate voltages. The magnitude of the current decreases as the gate voltage (VGS) is changed from −10 V (left) to 0 V (middle) to +10 V (right). (b) The source–drain current (IDS) versus source–drain voltage (VDS) characteristics of a reflectin-based protonic transistor obtained at VGS = − 10 V (blue), VGS = 0 V (black), and VGS = + 10 V (red). IDS decreases as VGS is changed from a negative value to a positive value. The measurements were performed at a relative humidity of 90%. (c) A plot of the experimentally observed proton charge carrier density (nH+) as a function of VGS (blue squares). The data correspond to the IDS versus VGS characteristics in (b). The red line represents the calculated proton charge carrier density. There is good agreement between the experimental and calculated values of the proton charge carrier densities. The device had a length of 50 μm, a width of 280 μm, and a thickness of 0.35 μm.

FIG. 4.

(a) Illustration of the protonic current for a protonic transistor under different applied gate voltages. The magnitude of the current decreases as the gate voltage (VGS) is changed from −10 V (left) to 0 V (middle) to +10 V (right). (b) The source–drain current (IDS) versus source–drain voltage (VDS) characteristics of a reflectin-based protonic transistor obtained at VGS = − 10 V (blue), VGS = 0 V (black), and VGS = + 10 V (red). IDS decreases as VGS is changed from a negative value to a positive value. The measurements were performed at a relative humidity of 90%. (c) A plot of the experimentally observed proton charge carrier density (nH+) as a function of VGS (blue squares). The data correspond to the IDS versus VGS characteristics in (b). The red line represents the calculated proton charge carrier density. There is good agreement between the experimental and calculated values of the proton charge carrier densities. The device had a length of 50 μm, a width of 280 μm, and a thickness of 0.35 μm.

Close modal

We proceeded to calculate the proton mobility from the I-V characteristics of our transistors from thin reflectin films via established literature protocols.13,14,21 We therefore extracted the conductivity from the transistors’ ISD vs. VSD curves and used a linear fit of the dependence of this conductivity on VGS to calculate the corresponding proton mobilities.13,14,21 The average calculated value of μH+ = 7.7 (±2.1) × 10−3 cm2 V−1 s−1 for thin reflectin films was similar to the previously reported values of ∼7.3 × 10−3 cm2 V−1 s−1 for thick reflectin films,21 ∼3 × 10−3 cm2 V−1 s−1 for dilute acids,23 ∼3.9 × 10−3 cm2 V−1 s−1 for poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) films,24 and ∼4.9 × 10−3 cm2 V−1 s−1 for maleic chitosan nanofibers.14 

We used the calculated mobilities and conductivities to evaluate the proton charge carrier density of the thin reflectin films according to established procedures.13,14,21 For example, at a VGS = 0 V, we found an average charge carrier density of nH+ = 2.5 (±1.1) × 1017 cm−3. This value was similar to that previously reported for both thick reflectin films and other proton-conducting materials.13,14,21 We then evaluated the charge carrier density of the films at different values of VGS. Notably, the experimentally determined charge carrier density was in excellent agreement with the one theoretically predicted by the equation nH+ = nH+0 − VGSCGS/et (Fig. 4(c)).

Finally, we calculated the IHIGH/ILOW ratios of protonic transistors from thin reflectin films. We found a ratio of 3.3 (±0.3) between the high current state at VGS = − 10 V and the low current state at VGS = + 10 V. This IHIGH/ILOW ratio of 3.3 represented a 2-fold improvement over the value of ∼1.6 previously found for thick reflectin films.21 Although this ratio was certainly below values of >106 reported for organic transistors,25 it compared favorably to the best values found for protonic transistors under comparable conditions and in analogous configurations (Table I).12,13,21 These observations underscored the excellent performance of the thin reflectin film-based protonic transistors.

TABLE I.

A comparison of the estimated IHIGH/ILOW ratios for protonic transistors from various materials. The devices were tested and characterized under comparable conditions and in analogous configurations. The approximate corresponding thicknesses for each type of device are also indicated.

Device material Estimated IHIGH/ILOW ratio Approximate thickness (μm) Reference
Reflectin A1 (thick films)  1.6a  ∼1-2  21  
Mesoporous silica  2.9b  ∼0.1-0.2  12  
Maleic chitosan  3.3c  ∼0.1  13  
Reflectin A1 (thin films)  3.3a  ∼0.2-0.4  This work 
Device material Estimated IHIGH/ILOW ratio Approximate thickness (μm) Reference
Reflectin A1 (thick films)  1.6a  ∼1-2  21  
Mesoporous silica  2.9b  ∼0.1-0.2  12  
Maleic chitosan  3.3c  ∼0.1  13  
Reflectin A1 (thin films)  3.3a  ∼0.2-0.4  This work 
a

The value was estimated at VGS = − 10 V, VGS = + 10 V, and VDS = 1.5 V.

b

The value was estimated at VGS = − 1.0 V, VGS = + 1.0 V, and VDS = 2.0 V.

c

The value was estimated at VGS = − 15 V, VGS = + 15 V, and VDS = 1.5 V.

In conclusion, we have electrically characterized protein-based protonic transistors from reflectin films with an average thickness of ∼0.30 μm. The figures of merit measured for these devices, including a mobility of μH+ = 7.7 (±2.1) × 10−3 cm2 V−1 s−1 and a charge carrier density of nH+ = 2.5 (±1.1) × 1017 cm−3, are comparable to previously reported values for reflectin films with a ∼1 to ∼2 μm thickness.21 However, transistors from thin reflectin films feature a more than 2-fold improvement in their IHIGH/ILOW ratio relative to transistors from thick reflectin films. The measured ratio of 3.3 compares favorably to the best values reported for similar devices (Table I). In their totality, our findings highlight the importance of the active layer geometry for optimum protonic transistor functionality.

We acknowledge Professor Albert Yee’s Laboratory and Professor Matthew Law’s Laboratory for the use of their atomic force microscopes. We also thank Professor Albert Yee’s Laboratory for the use of their optical microscope. We are grateful to the Air Force Office of Scientific Research (FA9550-14-1-0144) for its financial support.

1.
R. M.
Owens
and
G. G.
Malliaras
,
MRS Bull.
35
,
449
456
(
2010
).
2.
M.
Irimia-Vladu
,
N. S.
Sariciftci
, and
S.
Bauer
,
J. Mater. Chem.
21
,
1350
1361
(
2011
).
3.
P.
Meredith
,
C. J.
Bettinger
,
M.
Irimia-Vladu
,
A. B.
Mostert
, and
P. E.
Schwenn
,
Rep. Prog. Phys.
76
,
034501
(
2013
).
4.
G.
Tarabella
,
F. M.
Mohammadi
,
N.
Coppedè
,
F.
Barbero
,
S.
Iannotta
,
C.
Santato
, and
F.
Cicoira
,
Chem. Sci.
4
,
1395
1409
(
2013
).
5.
J.
Rivnay
,
R. M.
Owens
, and
G. G.
Malliaras
,
Chem. Mater.
26
,
679
685
(
2014
).
6.
M.
Berggren
and
A.
Richter-Dahlfors
,
Adv. Mater.
19
,
3201
3213
(
2007
).
7.
G.
Lanzani
,
Nat. Mater.
13
,
775
776
(
2014
).
8.
G. G.
Malliaras
,
Biochim. Biophys. Acta
1830
,
4286
4287
(
2013
).
9.
K.
Svennersten
,
K. C.
Larsson
,
M.
Berggren
, and
A.
Richter-Dahlfors
,
Biochim. Biophys. Acta
1810
,
276
285
(
2011
).
10.
V. F.
Petrenko
and
N.
Maeno
,
J. Phys. Colloq.
48
, C1-115–C1-119 (
1987
).
11.
Z. G.
Chiragwandi
,
O.
Nur
,
M.
Willander
, and
N.
Calander
,
Appl. Phys. Lett.
83
,
5310
5312
(
2003
).
12.
R.
Fan
,
S.
Huh
,
R.
Yan
,
J.
Arnold
, and
P.
Yang
,
Nat. Mater.
7
,
303
307
(
2008
).
13.
C.
Zhong
,
Y.
Deng
,
A. F.
Roudsari
,
A.
Kapetanovic
,
M. P.
Anantram
, and
M.
Rolandi
,
Nat. Commun.
2
,
476
(
2011
).
14.
Y.
Deng
,
E.
Josberger
,
J.
Jin
,
A. F.
Rousdari
,
B. A.
Helms
,
C.
Zhong
,
M. P.
Anantram
, and
M.
Rolandi
,
Sci. Rep.
3
,
2481
(
2013
).
15.
A. M.
Deml
,
A. L.
Bunge
,
M. A.
Reznikov
,
A.
Kolessov
, and
R. P.
O’Hayre
,
J. Appl. Phys.
111
,
074511
(
2012
).
16.
W. J.
Crookes
,
L.-L.
Ding
,
Q. L.
Huang
,
J. R.
Kimbell
,
J.
Horwitz
, and
M. J.
McFall-Ngai
,
Science
303
,
235
238
(
2004
).
17.
R. M.
Kramer
,
W. J.
Crookes-Goodson
, and
R. R.
Naik
,
Nat. Mater.
6
,
533
538
(
2007
).
18.
A. R.
Tao
,
D. G.
DeMartini
,
M.
Izumi
,
A. M.
Sweeney
,
A. L.
Holt
, and
D. E.
Morse
,
Biomaterials
31
,
793
801
(
2010
).
19.
M.
Izumi
,
A. M.
Sweeney
,
D.
DeMartini
,
J. C.
Weaver
,
M. L.
Powers
,
A.
Tao
,
T. V.
Silvas
,
R. M.
Kramer
,
W. J.
Crookes-Goodson
,
L. M.
Mäthger
,
R. R.
Naik
,
R. T.
Hanlon
, and
D. E.
Morse
,
J. R. Soc. Interface
7
,
549
560
(
2010
).
20.
L.
Phan
,
W. G.
Walkup
 IV
,
D. D.
Ordinario
,
E.
Karshalev
,
J.-M.
Jocson
,
A. M.
Burke
, and
A. A.
Gorodetsky
,
Adv. Mater.
25
,
5621
5625
(
2013
).
21.
D. D.
Ordinario
,
L.
Phan
,
W. G.
Walkup
 IV
,
J. M.
Jocson
,
E.
Karshalev
,
N.
Hüsken
, and
A. A.
Gorodetsky
,
Nat. Chem.
6
,
596
602
(
2014
).
22.
See supplementary material at http://dx.doi.org/10.1063/1.4901296 for a detailed description of the device fabrication and characterization procedures.
23.
S.
Cukierman
,
Biochim. Biophys. Acta, Bioenerg.
1757
,
876
885
(
2006
).
24.
E.
Stavrinidou
,
P.
Leleux
,
H.
Rajaona
,
D.
Khodagholy
,
J.
Rivnay
,
M.
Lindau
,
S.
Sanaur
, and
G. G.
Malliaras
,
Adv. Mater.
25
,
4488
4493
(
2013
).
25.
I.
Kymissis
,
Organic Field Effect Transistors: Theory, Fabrication and Characterization, Integrated Circuits and Systems
(
Springer
,
New York
,
2009
).

Supplementary Material