JP2015095557A - Field effect transistor using carbon nanotube aggregate - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a field effect transistor which forms a channel by using a carbon nanotube (CNT) aggregate and achieves high on-off ratio.SOLUTION: A carbon nanotube aggregate in which an amount of a metal type CNT having a diameter of 0.8nm of more and 1.3nm or less obtained by dividing a peak area derived from the metal type CNT observed in a region of 200 cmor more and 280 cmor less in Raman spectroscopic analysis with a wavelength of 532 nm by a peak area derived from a semiconductor type CNT observed in a region of 130 cmor more and 200 cmor less, is 0.1 times or less of an amount of a semiconductor type CNT having a diameter of 0.8nm of more and 1.3nm or less obtained by dividing a peak area derived from the semiconductor type CNT observed in a region of 190 cmor more and 280 cmor less in Raman spectroscopic analysis with a wavelength of 785 nm by a peak area derived from a metal type CNT observed in a region of 140 cmor more and 190 cmor less, is used, and the carbon nanotube aggregate is divided into strip regions to form a channel layer 106 of a field effect transistor 100.

Description

  The present invention relates to an electronic device using a carbon nanotube aggregate. In particular, the present invention relates to a field effect transistor in which a channel is formed by an aggregate of carbon nanotubes mainly containing semiconducting carbon nanotubes.

A carbon nanotube (hereinafter also referred to as “CNT”) has a mobility of 10 times or more that of silicon and can pass a current at a high current density of 10 ⁹ A / cm ² , so that it is a next-generation electronic material. It is attracting attention as. The leading role of electronic materials so far has been silicon semiconductors, and the performance of silicon MOSFETs (metal-oxide-semiconductor field-effect transistors) produced thereby has been improved along with the development of miniaturization technology. However, as the miniaturization of integrated circuits progresses, the gate insulating layer of the MOSFET needs to be thinned, so that the influence of quantum effects such as carrier tunneling cannot be ignored, and it is said that the limit of miniaturization is approaching. Yes.

  Accordingly, high-performance devices that cannot be realized with silicon MOSFETs are being developed by using a carbon nanotube aggregate in the active layer (channel) of a field effect transistor. For example, an example in which a logic circuit is formed by a field effect transistor using a carbon nanotube aggregate as a channel and its operation has been demonstrated has been reported (see Non-Patent Document 1).

  In a logic circuit, a field effect transistor is used as a switch, so that excellent switching characteristics are required. One index representing the switching characteristics is an on / off ratio, and if this value is small, the operation margin of the logic circuit is narrowed. Therefore, the field effect transistor is required to have a high on / off ratio.

However, as disclosed in Non-Patent Document 1, in order to increase the on / off ratio of a field effect transistor using a carbon nanotube aggregate, it is necessary to increase the channel length. FIG. 11 shows the channel length dependence of the on / off ratio in a field effect transistor using a conventional aggregate of carbon nanotubes. Referring to FIG. 11, when the channel length is about 10 μm, the on / off ratio can be obtained only about one digit, and in order to increase the on / off ratio to 10 ⁶ or more, the channel length must be about 100 μm or more. Has been.

  In a field effect transistor using a carbon nanotube aggregate, it is considered that the on / off ratio decreases when the channel length is short due to the influence of a component having high electrical conductivity contained in the carbon nanotube aggregate. A carbon nanotube aggregate is synthesized by an arc discharge method, a laser ablation method, a vapor phase growth method, or the like. It is known that both are included.

  When a channel of a field effect transistor is formed by a carbon nanotube aggregate in which both semiconductor-type carbon nanotubes and metal-type carbon nanotubes are mixed, the influence of conductivity due to the metal-type carbon nanotubes cannot be ignored. For example, when a metal-type carbon nanotube forms a conductive path that bridges between the source electrode and the drain electrode, the conductive path does not depend on the gate voltage, and a current flows due to a potential difference between the source electrode and the drain electrode. . For this reason, the off-state current of the field effect transistor inevitably increases.

  In order to solve such a problem, a method of selectively removing metal-type carbon nanotubes from a synthesized carbon nanotube aggregate has been studied. Metallic carbon nanotubes and semiconducting carbon nanotubes have very little structural difference, so it is extremely difficult to separate them. In view of recovery rate, purity, cost, etc., practically and economically The reality is that the solution has not been achieved to a satisfactory level. For example, a separation process for separating and refining semiconductor carbon nanotubes from a carbon nanotube aggregate in which metal-type and semiconductor-type carbon nanotubes are mixed has been developed (see Patent Documents 1 and 2). However, there are problems of damage to carbon nanotubes during the separation process, separation costs, and residual impurities such as a dispersant.

JP 2008-266112 A JP 2011-166070 A

Dong-ming Sun, Marina Y. Timmermans, Ying Tian, Albert G. Nasibulin, Esko I. Kauppinen, Shigeru Kishimoto, Takashi Mizutani and Yutaka Ohno, “Flexible high-performance carbon nanotube integrated circuits”, NATURE NANOTECHNOLOGY, PUBLISHED ONLINE: 6 FEBRUARY 2011, pp.1-6.

  According to the prior art, a field effect transistor using a carbon nanotube aggregate needs to have a long channel length in order to increase the on / off ratio, and specifically, the channel length must be 100 μm or more. On the other hand, since the on-state current of the field effect transistor is inversely proportional to the channel length, there is a problem that if the channel length is increased in order to obtain a high on / off ratio, the on-current is reduced.

  Accordingly, an object of the present invention is to achieve both a high on-current and a high on-off ratio in a field effect transistor in which a channel is formed by an aggregate of carbon nanotubes.

A field effect transistor according to an embodiment of the present invention includes a source layer and a drain electrode, and a channel layer including a carbon nanotube aggregate provided so as to be in contact with the source electrode and the drain electrode, and the channel layer at least partially overlaps the channel layer. This aggregate of carbon nanotubes has a gate insulating layer and a gate electrode, and this carbon nanotube aggregate has a peak area derived from metal-type carbon nanotubes observed in a region of 200 cm ⁻¹ or more and 280 cm ⁻¹ or less by Raman spectroscopy at a wavelength of 532 nm of 130 cm ^−. The amount of metallic carbon nanotubes having a diameter of 0.8 nm or more and 1.3 nm or less obtained by dividing by the peak area derived from a semiconductor type carbon nanpotube observed in a region of ^{1 to} 200 cm ⁻¹ is Raman having a wavelength of 785 nm. 190cm ^-1 more than 280cm ^-1 with a spectral analysis Obtained by dividing the peak area derived from metallic carbon nanotubes observed peak areas derived from the semiconducting carbon nanotube is observed in the regions beneath the 140cm ^-1 or 190 cm ^-1 or less in the region, or more in diameter 0.8nm The carbon nanotube aggregate has a property of being 0.1 times or less of the amount of semiconductor-type carbon nanotubes of 1.3 nm or less, and a short side in contact with the source electrode and the drain electrode, and a short side with respect to the short side It exists in a belt-like region surrounded by long sides extending in the vertical direction, the short side has a length of 5 micrometers or less, and a plurality of the belt-like regions are arranged apart from each other.

  In the aggregate of carbon nanotubes forming the channel layer of a field effect transistor, the amount of metallic carbon nanotubes having a diameter of 0.8 nm or more and 1.3 nm or less is defined as semiconductor type carbon nanotubes having a diameter of 0.8 nm or more and 1.3 nm or less. The carbon nanotube aggregate is divided into a plurality of band-shaped regions while reducing the off-current by eliminating the conductive path composed of metal-type carbon nanotubes, and the semiconductor-type carbon. A high on-current derived from a conductive path made of nanotubes can be maintained.

  According to an embodiment of the present invention, the carbon nanotube aggregate forming the channel layer is divided into a plurality of band-like regions, thereby obtaining a high on-current ratio while obtaining a high on / off ratio even when the channel length is short. Can be obtained.

It is a figure explaining the structure of the field effect transistor which concerns on one Embodiment of this invention. It is a figure explaining the synthesis | combination process of the carbon nanotube aggregate which concerns on one Example of this invention. It is a figure explaining the synthesis | combination process of the carbon nanotube aggregate which concerns on one Example of this invention. It is a graph which shows the Raman spectrum obtained when a plurality of carbon nanotube aggregates concerning one example of the present invention are analyzed by Raman spectroscopy. It is a graph which shows the Id-Vg characteristic of the field effect transistor which used the some carbon nanotube aggregate | assembly which concerns on one Example of this invention as a channel layer. It is a figure explaining the synthesis process of the carbon nanotube aggregate in a comparative example. It is a figure explaining the structure of the field effect transistor which concerns on one Example of this invention. It is a graph which shows the Id-Vg characteristic of the field effect transistor which used the some carbon nanotube aggregate | assembly which concerns on one Example of this invention as a channel layer. It is a graph which shows the channel width (The width | variety of the short side of a strip | belt-shaped area | region, and an on-off ratio) of the field effect transistor which concerns on one Example of this invention. It is a graph which shows the channel width (The width | variety of the short side of a strip | belt-shaped area | region, and the on-current density) of the field effect transistor which concerns on one Example of this invention. It is a graph which shows the channel length dependence of the on-off ratio in the transistor using the conventional carbon nanotube aggregate.

  Embodiments of the present invention will be described below with reference to the drawings. However, the present invention can be implemented in many different modes and should not be construed as being limited to the description of the embodiments exemplified below. Note that as to the contents of the invention described below, the same portions or portions having similar functions are denoted by the same reference numerals in different drawings, and repeated description is omitted unless there are special circumstances in that case.

[About field effect transistors]
FIG. 1 shows a configuration of a field effect transistor 100 according to an embodiment of the present invention. The field effect transistor 100 includes a gate electrode 103, a gate insulating layer 104, and a channel layer 106 made of a carbon nanotube aggregate. The source electrode 108 and the drain electrode 110 are provided in contact with the channel layer 106. A passivation layer 112 made of an insulating material may be provided on the channel layer 106.

  The aggregate of carbon nanotubes constituting the channel layer 106 is divided into a plurality of regions. Each region of the aggregate of carbon nanotubes is provided so that one end is in contact with the source electrode 108 and the other end is in contact with the drain electrode 110, and the individual regions are arranged at intervals. That is, the aggregate of carbon nanotubes is a strip (or strip) having a short side (d) in contact with the source electrode 108 and the drain electrode 110 and a long side extending in a direction substantially perpendicular to the short side (d). Separated into areas. In the channel layer 106, a plurality of adjacent band-like regions are arranged with a gap without contacting each other.

  The band-shaped region composed of the aggregate of carbon nanotubes has a short side (d) length of 10 micrometers, preferably 5 micrometers or less, and a mutual distance of 0.1 micrometers or more, preferably 1 macrometer or more. For example, it arrange | positions so that it may become 2 micrometers.

  The channel length of the field effect transistor 100 according to an embodiment of the present invention may be 100 micrometers or less, preferably 20 micrometers or less, more preferably 10 micrometers or less, and even more preferably 5 micrometers or less. . In this case, the channel width is a total value of the lengths of the respective short sides (d) of the band-shaped region composed of the carbon nanotube aggregate. Note that there is no limitation on the channel width, and the on-current changes in proportion to the channel width. Therefore, the value of the channel width can be set as appropriate depending on the number of band-like regions made of carbon nanotube aggregates.

[About carbon nanotube aggregates]
The aggregate of carbon nanotubes used for the channel layer 106 is ideally composed only of semiconductor-type carbon nanotubes, but it is difficult to purify only the semiconductor-type carbon nanotubes from the aggregate of carbon nanotubes to be synthesized. Therefore, in one embodiment of the present invention, the aggregate of carbon nanotubes used for the channel layer 106 is allowed to include not only semiconductor carbon nanotubes but also metal carbon nanotubes, while reducing the proportion of specific metal carbon nanotubes. I am letting. Accordingly, an increase in off-current can be prevented even if the channel length is shortened, and a value of 1000 or more can be obtained as the on-off ratio.

  By the way, the existence of a thick carbon nanotube and a thin carbon nanotube is confirmed for each of the semiconductor-type carbon nanotube and the metal-type carbon nanotube included in the carbon nanotube aggregate. In one embodiment of the present invention, it is preferable that the ratio of the thin semiconductor-type carbon nanotubes to the thin metal-type carbon nanotubes is high. Since the thin semiconductor-type carbon nanotube can switch the current value by the gate voltage, the use of a carbon nanotube aggregate having a large ratio as a channel of the field effect transistor can increase the on / off ratio.

Information on semiconductor-type carbon nanotubes and metal-type carbon nanotubes contained in the aggregate of carbon nanotubes and their thickness can be confirmed by performing Raman spectroscopic analysis of the aggregate of carbon nanotubes. The thin metal-type carbon nanotube is a carbon nanotube having a diameter of 0.8 to 1.3 nm and showing a signal at a Raman shift of 200 to 280 cm ⁻¹ in the Raman analysis with an excitation wavelength of 532 nm, which is thicker than this. The metal-type carbon nanotube is a carbon nanotube having a diameter of 1.3 to 1.7 nm that shows a signal at a Raman shift of 140 to 190 cm ⁻¹ during Raman analysis with an excitation wavelength of 785 nm. The thin semiconductor-type carbon nanotube is a carbon nanotube having a diameter of 0.8 to 1.3 nm which shows a signal at a Raman shift of 190 to 280 cm ⁻¹ in the Raman analysis with an excitation wavelength of 785 nm, and is thick. The semiconducting carbon nanotube is a carbon nanotube having a diameter of 1.3 to 1.7 nm which shows a signal at a Raman shift of 130 to 200 cm ⁻¹ in Raman analysis with an excitation wavelength of 532 nm.

In one embodiment of the present invention, the aggregate of carbon nanotubes applied to the channel layer 106 is observed in a region of at least 200 to 280 cm ⁻¹ of the Raman shift obtained when observed by Raman spectroscopy at a wavelength of 532 nm. It is preferable to use one having a reduced proportion of thin metallic carbon nanotubes.

When a carbon nanotube aggregate according to an embodiment of the present invention is characterized from a spectrum obtained by Raman spectroscopic analysis, a metal-type carbon nanotube observed in a region of 200 cm ⁻¹ or more and 280 cm ⁻¹ or less by Raman spectroscopic analysis at a wavelength of 532 nm The amount of metal-type carbon nanotubes having a diameter of 0.8 nm or more and 1.3 nm or less obtained by dividing the origin peak area by the peak area derived from a semiconductor-type carbon nanpotube observed in a region of 130 cm ⁻¹ or more and 200 cm ⁻¹ or less. Are observed in the region of 190 cm ⁻¹ or more and 280 cm ⁻¹ or less in the Raman spectroscopic analysis at a wavelength of 785 nm, and the metal type carbon nanotubes observed in the region of 140 cm ⁻¹ or more and 190 cm ⁻¹ or less of the peak area derived from the semiconductor type carbon nanotubes Divide by the peak area of origin And, 1 times or less of the following amounts of semiconducting carbon nanotubes than a diameter of 0.8 nm 1.3 nm, preferably of 0.1 times or less.

Further, as another index, among the Raman shifts obtained when the aggregate of carbon nanotubes is observed by Raman spectroscopic analysis at a wavelength of 532 nm, (A) a thin metal-type carbon nanotube observed in the region of 200 to 280 cm ^−1. And (B) when comparing the peak area intensity with the thick semiconductor-type carbon nanotubes observed at 130 to 200 cm ⁻¹ , it is preferable that the number of thin metal-type carbon nanotubes is small. A value obtained by dividing by (B) ((A) / (B)) is 1 or less, preferably 0.5 or less.

  Of course, it is ideal if thick metal-type carbon nanotubes can be reduced at the same time, but may be included in the channel layer 106 as long as it does not affect the characteristics of the field effect transistor.

  As described above, one embodiment of the present invention includes particularly thin metal-type carbon nanotubes while allowing a mixture of semiconductor-type carbon nanotubes and metal-type carbon nanotubes in the carbon nanotube aggregate used in the channel layer 106. By reducing the ratio, the on / off ratio of the field effect transistor is increased.

  In one embodiment of the present invention, the channel layer 106 uses an aggregate of carbon nanotubes divided into a plurality of band-like regions, but there is no limitation on other structures of the field effect transistor, and the channel layer is formed by a gate insulating layer. It is preferable to have an insulated gate structure insulated from the gate electrode. FIG. 1 shows an example in which a silicon substrate is used as the gate electrode 103. The gate insulating layer 104 is a silicon oxide film formed on the surface of a silicon substrate, and a carbon nanotube aggregate is synthesized on this surface to form a channel layer 106. A passivation layer 112 is provided on the channel layer 106. Instead of this passivation layer, an ion gel film may be provided and used as a gate insulating film.

  According to one embodiment of the present invention, the ratio of the semiconductor-type carbon nanotubes to the metal-type carbon nanotubes in the carbon nanotube aggregate forming the channel layer is set to a specific ratio, and the carbon nanotube aggregate is divided into a plurality of band-shaped regions. By dividing, a high on / off ratio can be realized in a field effect transistor having a channel length of 10 μm or less. Hereinafter, details of the field effect transistor will be described with reference to examples.

[Production of CNT]
With reference to FIGS. 2 and 3, an example of synthesizing a carbon nanotube aggregate having a high content ratio of semiconductor-type carbon nanotubes will be described. In this embodiment, a method for synthesizing an aggregate of carbon nanotubes applicable to a channel layer of a field effect transistor will be described.

(1) Formation of Catalyst Particles As shown in FIG. 2A, a substrate 102 is prepared for synthesizing a carbon nanotube aggregate. Catalyst particles 105 are formed on the surface of the substrate 102. In this embodiment, a silicon wafer is used as the substrate 102. A silicon oxide film 104 is formed on the surface of the substrate 102. The thickness of the silicon oxide film 104 is not limited, but may be 100 nm, for example.

  The surface of the silicon oxide film 104 is cleaned by a cleaning process. This is to prevent impurities from being mixed when the carbon nanotube aggregate is synthesized. Moreover, it is for preventing a carbon nanotube aggregate from growing abnormally. Various methods can be applied to the cleaning process, but the surface of the silicon oxide film 104 is cleaned by, for example, a plasma process using oxygen. Then, catalyst particles 105 are formed on the surface of the silicon oxide film 104.

In this embodiment, iron (Fe) is used as a catalyst for growing a carbon nanotube aggregate. To lay the particles by the catalyst dispersed in the surface of the silicon oxide film 104, including 0.01 mol / L of _NH 3 H _· H _{2 O} 2 in the ^{_{FeCl 3 2 × 10 -5 mol /}} l containing solution 50ml After immersing the substrate 102 in a solution to which 1000 μl of the solution has been added, the substrate 102 is rinsed with ultrapure water, and then washed with isopropyl alcohol (IPA) and dried. In this way, the catalyst particles 105 are laid on the surface of the silicon oxide film 104. The catalyst particles 105 are dispersed in the form of fine particles without aggregating on the silicon oxide film 104.

  Next, a carbon nanotube aggregate is synthesized using a substrate in which catalyst particles are dispersed. This process is performed by a stage of pretreatment of catalyst particles and a stage of synthesizing a carbon nanotube aggregate that is performed thereafter.

(2) Pretreatment of catalyst fine particles The pretreatment (the period indicated as “Formation” in FIG. 3) is intended to oxidize the catalyst particles appropriately, thereby synthesizing the next aggregate of carbon nanotubes. In this step, the content of the semiconductor-type carbon nanotube aggregate can be increased. In the pretreatment, the substrate 102 on which the catalyst particles 105 are formed on the surface of the silicon oxide film 104 is placed in a synthesis furnace and heated. After setting the substrate in the synthesis furnace, the substrate is heated while flowing atmospheric gas and reducing gas. The substrate is heated by heating the temperature in the furnace to 600 to 900 ° C., for example, 780 ° C. The atmosphere gas is preferably an inert gas. For example, helium (He) can be used, and hydrogen (H ₂ ) can be used as the reducing gas. The ratio of the reducing gas to the atmospheric gas may be 1/10 or less in terms of the flow rate ratio. In this embodiment, the flow rate of He is 1900 sccm, and H ₂ is flowing at 100 sccm.

  While the temperature in the synthesis furnace is heated to a predetermined temperature (for example, 780 ° C.), impurities remaining in the synthesis furnace are removed by maintaining this state for a certain time while flowing the atmospheric gas and the reducing gas. Impurities to be removed from the synthesis furnace are adsorbed carbon and oxygen, and also residual moisture is removed (a period indicated as “Transit” to “Wait” in FIG. 3). In particular, it is preferable to sufficiently remove moisture remaining in the synthesis furnace, and the water vapor amount in the synthesis furnace is set to 50 ppm or less by a moisture meter installed on the gas exhaust side.

In the synthesis furnace, after 99.9% or more of the gas components are occupied by the atmospheric gas and the reducing gas, water vapor (H ₂ O) is introduced while reducing the flow rate of the reducing gas, and the catalyst particles are converted into water vapor. Exposure is performed (period shown as “X” in FIG. 2B and FIG. 3). For example, when the gas flow rate of He is 2000 sccm, H ₂ is 4 sccm, and He gas containing 2000 ppm of H ₂ O is 50 sccm.

  In this way, after removing carbon, oxygen, and water vapor remaining in the synthesis furnace as impurities, the catalyst particles are uniformly exposed to water vapor by supplying a controlled amount of water vapor into the synthesis furnace. The catalyst particles are oxidized appropriately. As a result, a carbon nanotube aggregate in which the ratio between the metal type and the semiconductor type is controlled can be grown.

(3) Synthesis of aggregate of carbon nanotubes After the catalyst particles are exposed to water vapor (H ₂ O) for a certain period of time, the supply of water vapor (H ₂ O) is completely stopped. Then, the raw material gas is introduced into the synthesis furnace in a state in which the atmospheric gas and the reducing gas are flowed (period indicated as “Growth” in FIG. 3). The source gas is preferably a hydrocarbon gas, for example, ethylene (C ₂ H ₄ ). By supplying the source gas for a certain time, an aggregate of carbon nanotubes grows on the substrate (FIG. 2C).

  In this case, the growth efficiency of the thin metal-type carbon nanotubes is selectively suppressed by appropriately oxidizing the catalyst particles by applying the catalyst particles as described above and the steam treatment in the synthesis furnace. Thereby, it is possible to synthesize a carbon nanotube aggregate in which the proportion of metal carbon nanotubes in the whole is reduced while including semiconductor carbon nanotubes and metal carbon nanotubes.

  When the supply of the source gas is shut off, the growth of the carbon nanotube aggregate is stopped. The temperature is kept constant with the atmospheric gas flowing, and after a predetermined time has elapsed, the temperature in the synthesis furnace is lowered and the substrate is taken out to obtain an aggregate of carbon nanotubes (“Keep” in FIG. 3). ”And the subsequent periods).

[Characteristics of carbon nanotubes]
(1) Raman spectroscopic analysis From Raman spectroscopic analysis at a wavelength of 532 nm, the peak observed in the region of 200 to 280 cm ⁻¹ is a thin metal-type carbon nanotube, and the peak observed in the region of 130 to 200 cm ⁻¹ is a thick semiconductor type carbon. The presence of nanotubes can be confirmed. In addition, the presence of thick metal-type carbon nanotubes was confirmed from the peak observed in the region of 190 to 280 cm ^{−1 and} the peak observed in the region of 140 to 190 cm ^{−1 by} the Raman spectroscopic analysis at a wavelength of 785 nm. can do.

FIG. 4 shows Raman shift data obtained by Raman spectroscopic analysis at an excitation wavelength of 532 nm and Raman shift data obtained by Raman spectroscopic analysis at an excitation wavelength of 785 nm. From the graph shown in FIG. 4, in the Raman shift at an excitation wavelength of 532 nm, it is observed that the peak intensity is extremely weak in the region of 200 to 280 cm ⁻¹ and the proportion of thin metal-type carbon nanotubes is reduced. . On the other hand, a peak is observed in the region of 140 to 190 cm ⁻¹ from Raman shift data with an excitation wavelength of 785 nm, and it is confirmed that thick metal-type carbon nanotubes still exist.

Table 1 shows the peak area of the Raman spectrum for each excitation wavelength and the area intensity ratio. As shown in Table 1, the aggregate of carbon nanotubes obtained in this example has a peak area 67 derived from metal-type carbon nanotubes observed in a region of 200 cm ⁻¹ or more and 280 cm ⁻¹ or less by Raman spectroscopy at a wavelength of 532 nm. (count) was obtained by dividing the semiconducting carbon Nanpo tube from the peak area 1582 (count) observed at 130 cm ^-1 or more 200 cm ^-1 or less in the region, or more in diameter 0.8 nm 1.3 nm or less of the metal The amount of type carbon nanotubes is 0.042. On the other hand, the peak area 883 (count) derived from semiconductor-type carbon nanotubes observed in a region of 190 cm ⁻¹ or more and 280 cm ⁻¹ or less in a Raman spectroscopic analysis with a wavelength of 785 nm is a metal observed in a region of 140 cm ⁻¹ or more and 190 cm ⁻¹ or less. The amount of semiconducting carbon nanotubes having a diameter of 0.8 nm or more and 1.3 nm or less obtained by dividing by the peak area 605 (count) derived from the carbon nanotubes is 1.45. The value obtained by dividing the former by the latter is 0.03. From the results, it can be seen that although the proportion of thin metal-type carbon nanotubes is reduced, thick metal-type carbon nanotubes are present as they are.

  As described above, the amount of impurities in the reaction furnace before the synthesis of the carbon nanotube aggregate is reduced once, in particular, the amount of moisture, and then acts on the catalyst particles with a controlled amount of moisture, so that the thin metal-type carbon nanotubes An aggregate of carbon nanotubes with a small proportion can be synthesized.

[Characteristics of field effect transistor]
FIG. 5 shows the characteristics of the same field effect transistor having the structure shown in FIG.

  In the field effect transistor manufactured in this example, a silicon substrate is used as the substrate 102 and this is used as the gate electrode 103 with reference to FIG. The gate insulating layer 104 is a silicon oxide film and has a thickness of 100 nm. In addition, a laminate of titanium (Ti) and gold (Au) is used as the source electrode 108 and the drain electrode 110, and titanium (Ti) is brought into contact with the channel layer 106 formed of the carbon nanotube aggregate.

  Note that the field-effect transistor manufactured in this example has a channel length of 10 micrometers. In the channel layer 106, 30 carbon nanotube aggregates each having a short side (d: length in the channel width direction) of 1 micrometer are provided at intervals of 2 micrometers. Thus, the substantial channel width is 30 micrometers.

  The data shown in FIG. 5 is the drain current vs. gate voltage characteristics (Id-Vg characteristics), the source-drain voltage (Vds) is kept constant at 0.1 V, the source-gate voltage is swept, and the drain current is read. Is measured.

From the Id-Vg characteristics shown in FIG. 5, the on-current is 1.4 × 10 ⁻⁶ A, and the on-current density obtained by dividing the on-current by the channel width is 4.06 × 10 ⁻⁸ A / μm. The field effect mobility is 15 cm ² / V · sec. As described above, according to this embodiment, although the channel length is 10 μm, the on / off ratio is 3470 and a very high value is realized while maintaining a high on-current density.

  As a result, the conductive path formed by the metal-type carbon nanotube is divided by dividing the channel layer 106 made of the carbon nanotube aggregate into a plurality of band-like regions, so that the conductive path independent of the gate voltage is obtained. Suggests a substantial decrease. At this time, since the ratio of the semiconductor-type carbon nanotube is higher than that of the metal-type carbon nanotube, the characteristics as a field effect transistor are not deteriorated.

As described above, according to the present example, the carbon nanotube aggregate forming the channel layer is divided into a plurality of band-like regions, so that a high on / off ratio is 1000 or more even when the channel length is short, A field effect transistor having a value obtained by dividing the density by the channel width of 1 × 10 ⁻⁸ A / μm or more can be obtained.

[Comparative example]
FIG. 6 shows the characteristics of a field effect transistor produced using the carbon nanotube aggregate produced in Example 1 without dividing the channel layer into a plurality of band-like regions. The field effect transistor of the comparative example has a channel length of 10 μm and a channel width of 100 μm. Since the field effect mobility is 35 cm ² / V · sec and the on-current is as high as 25 × 10 ⁻⁶ A, the on-current density is as high as 2.5 × 10 ⁻⁷ A / μm. Is only 8.5.

  The result of the comparative example shows that the on / off ratio and the on-current density differ depending on the form of the channel layer even when the composition of the carbon nanotube aggregate, that is, the mixing ratio of the semiconductor-type carbon nanotube and the metal-type carbon nanotube is the same. In other words, a carbon nanotube aggregate in which the proportion of the thin metal-type carbon nanotubes is reduced is used, and the width (length in the channel width direction) of the carbon nanotube aggregate is set to 10 micrometers or less, and a plurality of carbon nanotube aggregates with a predetermined interval are used. By providing this and using it as a channel layer, both a high on-off ratio and an on-current density can be achieved.

  FIG. 7 shows the configuration of the field effect transistor according to this example. The difference from the field effect transistor shown in FIG. 1 is that a gate insulating film 116 made of an ion gel film is provided on the channel layer 106.

  As illustrated in FIG. 7, the field effect transistor 101 includes a source electrode 108, a drain electrode 110, and a gate electrode 114 provided on a gate insulating film 104. The gate electrode 114 can be provided in the same layer as the source electrode 108 and the drain electrode 110. A gate insulating film 116 made of an ion gel film is provided so as to cover at least the channel layer 106. An electric field is applied to the gate insulating film 116 made of an ion gel film by the gate electrode 114.

  The ion gel film is provided so as to cover the upper surface and side surfaces of the band-shaped region made of the carbon nanotube aggregate. As the ion gel film, for example, a film containing a π-conjugated polymer is used. By using this as the gate insulating film 116 and controlling the gate voltage with respect to the voltage between the source and the drain, the π-conjugated polymer is partially oxidized and drained. The current can be controlled (switched).

From the Id-Vg characteristics shown in FIG. 8, the on-current is 2.8 × 10 ⁻⁶ A, the on-current density obtained by dividing the on-current by the channel width is 9.3 × 10 ⁻⁸ A / μm, and the on-off The ratio is 6220.

  FIG. 9 shows the relationship between the width of the short side (d) of the strip region in the channel layer 106 and the on / off ratio. The on / off ratio is significantly increased by setting the width of the short side (d) to 10 micrometers or less, preferably 5 micrometers or less.

  FIG. 10 shows the relationship between the width of the short side (d) of the band-like region in the channel layer 106 and the on-current density. In this case, even if the width of the short side (d) of the belt-like region is changed from 100 micrometers to 1 micrometer, the decrease in the on-current density is within one digit.

  According to this example, an extremely high on-off ratio and sufficient on-current density can be realized by using a channel layer obtained by dividing a carbon nanotube aggregate into strip-like regions and an ion gel film as a gate insulating film. . In this case, the width of the short side (d) of the band-like region is preferably smaller than 10 micrometers, while the on-current density does not greatly depend on the width of the band-like region, and a high on-current density is obtained. Can be maintained.

100: field effect transistor 102: substrate 103: gate electrode 104: gate insulating layer 105: catalyst particle 106: channel layer 108: source electrode 110: drain electrode 112: passivation film 114: gate electrode 116: gate insulating film

Claims (4)

A source electrode, a drain electrode, a channel layer made of a carbon nanotube aggregate provided so as to be in contact with the source electrode and the drain electrode, and a gate insulating layer and a gate electrode at least partially overlapping the channel layer ,
The aggregate of carbon nanotubes is observed peak areas derived from the metallic carbon nanotubes observed in 200 cm ^-1 or more 280 cm ^-1 or less in the region at the Raman spectroscopic analysis of wavelength 532nm to 130 cm ^-1 or more 200 cm ^-1 or less in the region The amount of metal-type carbon nanotubes having a diameter of 0.8 nm or more and 1.3 nm or less obtained by dividing by the peak area derived from the semiconducting carbon nanpotube is 190 cm ⁻¹ or more and 280 cm ⁻¹ or less in Raman spectroscopic analysis with a wavelength of 785 nm. The peak area derived from semiconductor-type carbon nanotubes observed in the region of ¹ is divided by the peak area derived from metal-type carbon nanotubes observed in the region of 140 cm ⁻¹ or more and 190 cm ⁻¹ or less. .The amount of semiconducting carbon nanotubes of 3 nm or less Have the property of a .1 times or less,
The aggregate of carbon nanotubes is present in a belt-like region surrounded by a short side in contact with the source electrode and the drain electrode and a long side extending in a direction substantially perpendicular to the short side, A field-effect transistor having a length of 5 micrometers or less and a plurality of the band-like regions arranged apart from each other.   2. The field effect transistor according to claim 1, wherein a distance between the source electrode and the drain electrode is 10 micrometers or less. 3. The field effect transistor according to claim 1, wherein the on / off ratio is 1000 or more and the field effect mobility is 10 cm ² / V · sec or more. 4. A value obtained by dividing an on-current density by a channel width when the potential difference between the source electrode and the drain electrode is 0.1 V is 10 ⁻⁸ A / μm or more. The field effect transistor as described in any one.