3D graphene transistor
11121334 · 2021-09-14
Assignee
Inventors
- Sameer Sonkusale (Arlington, MA, US)
- Shideh Kabiri Ameri Abootorabi (Medford, MA, US)
- Pramod Kumar Singh (Ambedkar Nagar, IN)
Cpc classification
H01L29/778
ELECTRICITY
G01N27/4145
PHYSICS
H01G9/2013
ELECTRICITY
International classification
G01N27/414
PHYSICS
A61B5/1473
HUMAN NECESSITIES
H01L29/06
ELECTRICITY
H01L29/16
ELECTRICITY
H01L29/778
ELECTRICITY
Abstract
A field effect transistor having a channel that comprises three-dimensional graphene foam. The subject matter of the invention concerns a three dimensional field-effect transistor having a channel based on graphene foam and the use of ionic liquid as a gate. The graphene foam is made of a three-dimensional network of single and double layer graphene that extends in all the three dimensions. Metal contacts on either end of the graphene foam form the drain and source contacts of the transistor.
Claims
1. An apparatus comprising a field-effect transistor having a channel that comprises a three-dimensional graphene foam that comprises an interconnected network of graphene strips, wherein said graphene strips are made of a form of graphene selected from the group consisting of monolayer graphene and bilayer graphene.
2. The apparatus of claim 1, further comprising an ionic liquid, wherein said ionic liquid bathes said foam.
3. The apparatus of claim 2, wherein said ionic liquid comprises 1-Butyl-3-methylimidazolium hexafluorophosphate.
4. The apparatus of claim 1, wherein said strips are made of monolayer graphene.
5. The apparatus of claim 1, further comprising a hydrophilic agent on said graphene.
6. The apparatus of claim 1, wherein said agent comprises HfO.sub.2.
7. The apparatus of claim 1, further comprising a sensor, wherein variation in a property of said transistor forms a basis for a measurement by said sensor.
8. The apparatus of claim 7, wherein said sensor comprises a biological sensor.
9. The apparatus of claim 8, wherein said biological sensor is configured for in vivo measurements.
10. The apparatus of claim 7, wherein said sensor comprises a chemical sensor.
11. The apparatus of claim 7, wherein said sensor comprises strain sensor.
12. The apparatus of claim 7, wherein said sensor comprises a pH sensor.
13. The apparatus of claim 1, wherein said transistor incorporates a scaffold.
14. The apparatus of claim 1, wherein said strips are made of bilayer graphene.
15. The apparatus of claim 1, wherein variation in a property of said transistor forms a basis for a measurement of both pH and strain.
16. The apparatus of claim 1, wherein said transistor is a multimodal device that measures multiple inputs at the same time.
17. The apparatus of claim 1, wherein said field-effect transistor comprises a source terminal and a drain terminal, wherein said channel defines a plane having a top side a bottom side, and a perimeter, and wherein said source and drain terminals connect to said perimeter.
18. The apparatus of claim 1, wherein said graphene is suspended above a substrate.
Description
BRIEF DESCRIPTION OF THE FIGURES
(1)
(2)
(3)
(4)
(5)
(6)
(7)
DETAILED DESCRIPTION
(8)
(9) The left side of
(10) The next step is to form drain and source contacts (step 16). This is carried out by deposition through a shadow mask using electron beam evaporation. In one embodiment, the titanium is deposited to a thickness of 10 nm followed by deposition of a 50 nm gold layer. In another embodiment, the titanium layer is 20 nm and the gold layer is 220 nm.
(11) In another deposition step (step 18), 10 nm of titanium is deposited through another shadow mask to cover the gold on the source and drain where contact is made with the channel. This forms a protective layer against ion leakage.
(12) Optionally, the contacts are covered with epoxy to prevent contact between the solution and the electrodes.
(13) Finally, the shadow mask is removed and ionic liquid is added to the channel (step 20). A metal electrode is suspended in the ionic liquid to apply a gate voltage. In one embodiment, the ionic liquid is 1-Butyl-3-methylimidazolium hexafluorophosphate (BmimPF.sub.6, 98%). In one embodiment, the device is about 600 microns long and 5 mm wide.
(14) In one embodiment that is suited for use as a pH meter, the graphene is coated with 20 nm of HfO.sub.2 using ALD. This is because the sensing mechanism for a graphene FET is based on absorbing chemical and biological molecules on the surface. Since graphene is inherently hydrophobic, adding a layer of HfO.sub.2 creates a hydrophylic surface and increases the pH sensitivity of the resulting sensor. In this embodiment, a solution with a pH between 3 and 10 is prepared by adding NaOH and HCL to a phosphate buffered saline (PBS). This serves as the gate with an Ag/AgCl reference electrode used to apply a gate voltage bias.
(15) The resulting device is ideal as a scaffold with inherent pH responsivity for tissue engineering applications. Such a device is more sensitive than known pH meters. This superior sensitivity is believed to arise from the unique three-dimensional structure of the device, which provides a higher surface area and therefore a greater pH-dependent field effect of transistor behavior. This greater sensitivity is improved further by the use of the hafnium oxide layer, which provides a hydrophilic surface that improves interaction with an aqueous based ionic liquid in the channel.
(16)
(17) The graphene layers in the foam are naturally suspended above the substrate 24. This suppresses undesirable substrate effects that plague conventional two-dimensional graphene transistors, thus achieving high carrier mobility. For example, in a planar two-dimensional graphene transistors realized on a substrate, overall mobility of the charge carriers is diminished because of extrinsic scattering from phonons due to trapped charges at the graphene-dielectric interface. Mobility of high quality graphene ranges between 10.sup.3 cm.sup.2V.sup.−1s.sup.−1 for graphene grown on silicon. But for suspended graphene, this mobility rises to 10.sup.6 cm.sup.2V.sup.−1s.sup.−1. Moreover, using ionic liquid as a gate results in a large capacitance at the liquid-channel interface allowing for low voltage operation of the device.
(18) In one embodiment, the transistor has a gate length of 600 μm and width of 5 mm. The effective length and width of the transistor will be different than the drawn length due to the three-dimensional path between source and drain. The device sizes were chosen for convenience of measurement and for proof of concept demonstration. These sizes can easily be scaled down using standard photolithography.
(19)
(20) Graphene layers do not have any structural support once the copper has been etched away. As a result, micro and nano-sized shrinks tend to develop. These shrinks degrade the overall mobility of the carriers. Estimated electron and hole mobility of graphene foam after etching away the copper are only 2497 cm.sup.2V.sup.−1s.sup.−1 and 4023 cm.sup.2V.sup.−1s−.sup.1 respectively.
(21) The mobility of the carriers can be improved by decreasing the ripples and defects created in graphene and by preventing the collapse and shrinkage in the active layer after etching copper. This can be done by using a supportive layer while etching the copper away and by etching in a more controlled manner, for example by adding etchant at a constant flow rate and by taking steps to reduce overall mechanical vibrations.
(22) While the observed mobility of graphene has been lower than what has been demonstrated for suspended graphene, it is adequate for most analog and sensing applications. Moreover, proposed liquid gated FETs are expected to somewhat inherently slow anyway, simply because of the low mobility of the heavy ions in the ionic liquid. The mobility degradation caused by graphene shrinkage is therefore not a limiting factor and not noticeably affect the overall speed performance of the transistor.
(23)
(24) The behavior shown in the left-hand plot in
(25) The type of charge carrier in the channel depends on the gate voltage. A negative gate voltage creates a p-type graphene channel and positive gate voltage creates an n-type channel.
(26) The extracted minimum contact resistance at the source/drain terminal is 6.87 Ωmm and the extracted equivalent sheet resistance of the active layer is 620 Ω/□. Contact resistance and sheet resistance can be improved by decreasing the defects and the ripples in the graphene foam after etching away the copper. The contact resistance decreases as the interface between metal and graphene is improved.
(27) Another way to improve contact resistance is to control base pressure during metal deposition. In the embodiments described herein, the base pressure during the deposition of titanium is 10.sup.−6 Torr. However, further reduction will improve contact resistance. For example, at a base pressure of 8×10.sup.−9 Torr, contact resistance decreases to less than 250 Ω.μm.
(28) The ionic solution used as a liquid gate is 1-Butyl-3-methylimidazolium hexafluorophosphate (BMM PF.sub.6, 98%), which creates a double layer capacitance of approximately 15-nF at the interface with the graphene. This serves as the gate capacitance of the transistor. The on/off current ratio in the device described herein is about 5. In a long channel graphene transistor, the on/off current ratio can be between 2 and 20.
(29) The plot on the right-hand side of FIG. 7 shows the drain current versus drain voltage characteristic of the transistor. As the carrier density increases with increasing drain voltage, the drift velocity does not saturate for small values of V.sub.DS. Therefore saturation behavior is not observed in graphene transistors.
(30) The electronic performance indicates that liquid gating using high conductivity ionic liquid as a gate provides a high level of electrostatic control over the graphene transistor. This results in high carrier concentration and high mobility in the transistor with very low operating voltages.
(31) A transistor as described herein need not be restricted to a long channel transistor. Such a transistor could be scaled down considerably with standard lithography to achieve even better electronic properties.
(32) A transistor as described herein, with its three dimensional random network, high surface area, and liquid gating is ideally suited as chemical and biological sensor for variety of applications. Due to graphene's highly stable carbonaceous form, it is also biocompatible. The proposed three-dimensional liquid gated transistor is therefore ideal as an electronic interface with a biological system through which it is possible to record electrical signals, to apply electrical stimuli, and to sense chemical and biological parameters in vivo.