JPH0992816A - Field-effect transistor and its manufacture - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the breakdown voltage of a gate and also reduce the dispersion in element characteristics and the resistance in lateral direction of a high electron mobility transistor(HEMT). SOLUTION: Sequentially laminated on an InP substrate 1 are a buffer layer 21 (film thickness of 100nm), a p type InGaAs layer 22 (film thickness of 50nm), a barrier layer 23 (film thickness of 10nm) comprising updoped InAlAs (In 52%), a channel layer 3 (film thickness 15nm) comprising undoped InGaAs (In 53%), a spacer layer 4 (film thickness of 5nm) comprising undoped InAlAs (In 52%), a doped layer 5 (film thickness of 10nm) comprising n type InAlAs (In 52%), and a cap layer 11 (film thickness 5nm) comprising an n type InGaAs (In 80%). Top of a cap layer 11 comprises SiN and an insulating film 12 having an opening 12a is formed, and n type impurities are introduced to a region A through the opening 12a. A gate electrode 8 is formed on the insulating film 12, a source electrode 9 and a drain electrode 10 is formed on the cap layer 11 through the opening 12a, thereby forming HEMT 100.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a field effect transistor (hereinafter, referred to as FET) used for a high frequency amplification element, a high speed switching element and the like.

[0002]

2. Description of the Related Art With the recent increase in demand for radio waves, compound semiconductor FEs operating at higher frequencies than silicon transistors
T, for example, GaAs MESFET (metal semicondu
ctor FET) and high electron mobility transistor (High Electro
n Mobility Transistor: hereafter called HEMT), the demand for FET is increasing. In order to support the higher frequency millimeter wave band, InAlAs / In
HEMTs using a new structure called GaAs are being developed. This InAlAs / InGaAs HE
The structure of MT200 is shown in FIG. Substrate 1 made of InP
The buffer layer 2 made of InAlAs (having a film thickness of 100
nm), a channel layer 3 made of InGaAs (film thickness 15
nm), a spacer layer 4 made of InAlAs (film thickness 5n
m) A doped layer 5 (thickness: 10 n) made of n-type InAlAs doped to a carrier concentration of 1 × 10 ¹⁹ cm ^−3.
m), a gate contact layer 6 (film thickness 10 nm) made of InAlAs, and a cap layer 7 (film thickness 20 nm) made of n-type InGaAs doped with a carrier concentration of 1 × 10 ¹⁹ cm ⁻³ . . Here, the non-n-type layer is an undoped layer. The composition ratio of In is 0.52 in the InAlAs layer and 0.53 in the InGaAs layer. The cap layer 7 is provided only at the portion where the gate electrode 8 is provided.
Recess etching is performed to form the recess portion 7a, and the gate electrode 8 having a T-shaped cross section is formed on the exposed gate contact layer 6. A source electrode 9 and a drain electrode 10 are formed on the cap layer 7.

The purpose of forming the cap layer 7 on the gate contact layer 6 is to prevent oxidation of the gate contact layer 6 and to reduce lateral resistance from the source electrode 9 and the drain electrode 10 to the vicinity of the gate electrode 8 to improve device performance. To improve. Therefore, since the doping is performed at a considerably high concentration, it is necessary to form the gate electrode 8 after removing the cap layer 7 in order to form the gate electrode 8 that requires the Schottky characteristic. InAlA in FIG.
The structure of s / InGaAs HEMT200 is shown.
GaAs MESFET and conventional AlGaAs / GaAs
In the system HEMT as well, a heavily doped n-type cap layer is provided for the same purpose, and it is necessary to form the gate electrode after removing this by recess etching.

[0004]

However, in the above disclosed technique, in order to remove the cap layer 7, it is necessary to stop the etching with the gate contact layer 6 exposed. Since the film thickness of the gate contact layer 6 is about 10 nm, this etching requires fairly precise control. When the cap layer 7 remains or the doped layer 5 is exposed, the gate electrode 8 does not exhibit the Schottky characteristic and the FET does not operate. In addition, since the characteristics change when the distance from the gate electrode 8 to the channel layer 3, that is, the remaining etching thickness changes, it is necessary to perform etching control with an accuracy of about 1 nm in order to obtain a device with stable characteristics. There is. Therefore, an attempt has been made to improve the etching accuracy by utilizing the difference in etching rate due to the difference in material (InGaAs and InAlAs in FIG. 9). The etching conditions are described in detail, for example, in J. Electrochem. Soc., Vol 139, pp831-835 (1992). As a result of the experiment, InGaAs and In
The etching rate difference from AlAs is about 20: 1 depending on the conditions. However, even in this case, there are variations in the characteristics, and since it depends on the surface state (dirt or the like) in the initial stage of etching, the etching amount may largely change.

Therefore, in view of the above problems, an object of the present invention is to omit an etching step for forming a gate electrode,
It is an object of the present invention to provide a FET having a reduced gate resistance and a reduced lateral resistance and improved gate breakdown voltage.

[0006]

Means for Solving the Problems To solve the above problems, the means described in claim 1 can be adopted. According to this means, the p-type first semiconductor layer, the undoped n-type second semiconductor layer, and the undoped or n-type third semiconductor layer are sequentially formed on the substrate. A gate electrode and source and drain electrodes on both sides thereof are formed on the third semiconductor layer. N-type impurities are introduced into the first, second, and third semiconductor layers below the source and drain electrodes. With such a configuration, it is not necessary to carry out etching to form a recess portion as in the conventional case for forming a gate electrode. Therefore, a HEMT in which variation in element characteristics caused by etching is reduced is obtained. realizable. Further, conventionally, the thickness of the third semiconductor layer was limited because it was necessary to etch the third semiconductor layer, but in the present invention, the n-type impurity in the p-type first semiconductor layer is introduced. The formed portion plays a resistance reducing effect, and since there is no limitation on the film thickness of this layer, the lateral resistance can be reduced.

Further, by adopting the means described in claim 2, since the gate electrode is formed on the third semiconductor layer via the insulating film, MIS (Metal Insulator Semicondu
The gate breakdown voltage can be improved. By adopting the means described in claim 3, an InAlAs / InGaAs HEMT having excellent device characteristics can be realized. By adopting the means described in claim 4, an AlGaAs / GaAs HEMT having excellent device characteristics can be realized. By adopting the means according to claim 5, AlGaAs / InGaA excellent in device characteristics is obtained.
An s-based HEMT can be realized. By adopting the means according to claim 6, a GaAs-based MESFE excellent in device characteristics
T can be realized. By adopting the means described in claim 7, since InGaAs has high electron mobility and low resistance, the lateral resistance can be reduced by increasing the film thickness of this layer. By adopting the means described in claim 8, the contact resistance of the ohmic electrode can be reduced by increasing the In composition ratio of the second semiconductor layer. By adopting the means described in claim 9, the gate electrode can be easily brought into Schottky contact with the second semiconductor layer.

By adopting the means according to claim 10, the gate electrode can be formed on the third semiconductor layer without etching, and a HEMT having an MIS structure with excellent device characteristics can be manufactured. In addition, it is not necessary to remove the insulating film used as the mask when introducing the n-type impurity, and the manufacturing process can be simplified. Further, by adopting the means described in claim 11, it is possible to manufacture a HEMT having a simpler structure.

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION

(First Embodiment) Hereinafter, the present invention will be described based on specific embodiments. FIG. 1 is a schematic structural diagram showing a configuration of a HEMT 100 (corresponding to a field effect transistor). I
On the substrate 1 made of nP, a buffer layer 21 made of InAlAs, a p-type InGaAs layer 22 (corresponding to the first semiconductor layer) lightly doped with a carrier concentration of about 1 × 10 ¹⁶ cm ⁻³ , made of InAlAs Barrier layer 23, I
A channel layer 3 made of nGaAs (corresponding to a second semiconductor layer), a spacer layer 4 made of InAlAs, 1 × 10 ¹⁹
n-type InA doped with a carrier concentration of about cm ^-3
Doped layer 5 made of 1As (corresponding to a semiconductor layer made of n-type InAlAs), 1 × 10 ¹⁹ c with an In composition ratio of 0.8
n-type InGa doped with a carrier concentration of about m ⁻³
A cap layer 11 made of As (corresponding to a third semiconductor layer)
Are sequentially laminated. An insulating film 12 made of SiN is formed on the cap layer 11 and has an opening 12a. A source electrode 9 and a drain electrode 10 are formed on the cap layer 11 through the opening 12a, and a gate electrode 8 is formed on the insulating film 12. A region A (corresponding to an n-type impurity introduction region) in the drawing is a region into which an n-type impurity is introduced. In this way, InAlAs / In
A GaAs HEMT 100 is constructed.

A region A in the figure is a region into which n-type impurities are introduced using the insulating film 12 as a mask. As for the amount of n-type impurities introduced into the region A, it is necessary that at least the lightly-doped p-type InGaAs layer 22 is inverted to n-type. In fact, the higher the n-type concentration of the InGaAs layer 22 is, the more advantageous it is to reduce the lateral resistance.
The n-type impurity concentration of the GaAs layer 22 is 10 ¹⁸ cm ⁻³ to 10
It is desirable to introduce impurities so that the ^{concentration is} about ¹⁹ cm ^-3 . In the above structure, the layer that is not the n-type is an undoped layer. In addition, the In composition not particularly described is
The InAlAs layer is 0.52, and the InGaAs layer is 0.53.

Next, a method of manufacturing the HEMT 100 will be described with reference to FIG. FIG. 2 is a schematic structural diagram showing the manufacturing method of the HEMT 100. First, molecular beam crystal growth (Mo
lecular Beam Epitaxy: hereinafter referred to as MBE)
A buffer layer 21 having a film thickness of 100 nm and InG is formed on the substrate 1.
The aAs layer 22 has a thickness of 50 nm, and the barrier layer 23 has a thickness of 10
nm, the channel layer 3 has a thickness of 15 nm, the spacer layer 4 has a thickness of 5 nm, the doped layer 5 has a thickness of 10 nm, and the cap layer 11 has a thickness of 10 nm.
Are sequentially grown to a film thickness of 5 nm (see FIG. 2A).
At this time, each layer may be formed by another crystal growth method, and the film thickness of each layer is not limited to the above.
Subsequently, the substrate shown in FIG. 2A is taken out from the MBE apparatus, and the insulating film 12 is formed on the entire surface of the cap layer 11 to a thickness of 1
The insulating film 12 is formed to a thickness of 0 nm and only the portion where the gate electrode 8 is to be formed is left, and the opening 12a is formed. At the time of introducing impurities in the next step, if the insulating film 12 having a film thickness of 10 nm is insufficient as a mask, the SiO ₂ layer 13 is formed on the insulating film 12 (see FIG. 2B).

Then, the insulating film 12 and the SiO ₂ layer 1
3 is used as a mask to introduce an n-type impurity by ion implantation or introduction to form an n-type impurity introduction region A (FIG. 2).
(C)). If ion implantation is performed at this time, it is necessary to perform activation treatment by rapid thermal annealing or the like. After that, the source electrode 9 is formed by the lift-off method.
Then, the drain electrode 10 is formed on the region A of the cap layer 11, and an alloying process is performed (see FIG. 2D). Since the insulating film 12 made of semiconductor and SiN is hardly dissolved in hydrofluoric acid, only the SiO ₂ layer 13 is removed using hydrofluoric acid. Then, the gate electrode 8 having a T-shaped cross section is formed on the insulating film 12 by using a normal gate electrode forming method. In this way, the HEMT 100 is manufactured (FIG. 2).
(E)).

The operation of the part of the HEMT 100 having the above structure, which is different from the conventional structure, will be described below. The layer newly provided in this embodiment is a p-type InGaAs layer 2
2, the barrier layer 23, the cap layer 11, and the insulating film 12. When the n-type impurity is introduced, the p-type InGaAs layer 22 is inverted to the n-type except the lower part of the insulating film 12, that is, the lower part of the gate electrode 8. This layer 22 is InGaA
Since it is composed of s, it has high electron mobility and low resistance.
However, since the polarity of this InGaAs layer 22 between the source electrode 9 and the drain electrode 10 is n / p / n,
InGaAs layer 2 between the source electrode 9 and the drain electrode 10
No current flows through 2. This is important in obtaining the pinch-off characteristic of HEMT100.

The current path in this case is shown in FIG. In the region A, the current flows through the n-type InGaAs layer 22 and the channel layer 3 having a low resistance, and under the gate electrode 8, only the channel layer 3 flows. Conventionally, when an n-type cap layer is provided on the surface, the film thickness is limited due to the necessity of etching, and the limit is 20 to 30 nm. In the present embodiment, the lower InGaAs layer 22 is responsible for reducing the lateral resistance, and since there is no limitation on the film thickness of this layer, it is possible to reduce the lateral resistance by increasing the film thickness. .

The barrier layer 23 is an energy barrier layer for improving the confinement property of the two-dimensional electron gas formed in the channel layer 3, and reduces the short channel effect generated when the gate length is shortened. Therefore, the element characteristics can be improved. The cap layer 11 has an action of preventing oxidation of the underlying doped layer 5 and a function of reducing contact resistance between the source electrode 9 and the drain electrode 10. The effect of reducing the contact resistance is that I of the cap layer 11
A larger n composition is more effective, but it is not particularly limited to this value (In composition 0.8 here). Also,
Although the carrier concentration of the p-type lightly doped InGaAs layer 22 is set to about 1 × 10 ¹⁶ cm ⁻³ , the carrier concentration is not limited to this value.

The insulating film 12 has a role as a mask for introducing impurities and a role as a MIS structure for increasing the gate breakdown voltage. However, if the insulating film 12 is too thick, the distance between the gate electrode 8 and the channel layer 3 becomes large and the performance deteriorates.

The structure of this embodiment eliminates the semiconductor etching process under the gate electrode 8 and eliminates the need for precise etching control. In addition, the gate electrode 8 and the channel layer 3
The distance between the gate electrode 8 and the channel layer 3 depends on only the film thicknesses of the spacer layer 4, the doped layer 5, and the cap layer 11 at the time of stacking, but these layers can be formed with high accuracy. It has excellent controllability and can reduce variations in device characteristics. Furthermore, since it is a laminated structure that prevents an increase in lateral resistance that occurs when only the etching process is eliminated from the conventional structure, it is possible to make the lateral resistance the same or to reduce it, and to make the element characteristics the same or It can also be improved.

(Second Embodiment) A second embodiment according to the present invention will be described below with reference to FIG. Figure 4 shows HEMT10
1 is a schematic structural diagram showing the constitution of No. 1, and the layers below the doped layer 5 have the same constitution as in the first embodiment. A cap layer 1 made of n-type GaAs undoped or doped with a carrier concentration of about 1 × 10 ¹⁷ cm ^-3 is formed on the doped layer 5.
10 (thickness 3 nm) is formed. GaAs and In
The lattice mismatch with AlAs is about 3.7%, but good crystal growth without dislocation is possible within the critical film thickness. According to the energy balance model, the critical film thickness is 4.5n
The thickness is about m, and GaAs with no defects can be obtained when the film thickness is 3 nm as in the present embodiment. Further, unlike the first embodiment, in this embodiment, the insulating film below the gate electrode 8 is not provided, and the source electrode 9, the drain electrode 10 and the gate electrode 8 are formed on the cap layer 110. H with such a configuration
In the manufacturing process of the EMT 101, an insulating film is formed on a portion of the cap layer 110 where the gate electrode 8 is to be formed, and impurities are introduced using this insulating film as a mask to form an n-type impurity introduction region A. Then, after removing the insulating film, the source electrode 9 and the drain electrode 10 are formed on the n-type impurity introduction region A, and the gate electrode 8 is formed on the region in which the impurities are not introduced. In this embodiment, GaAs
Since the cap layer 110 made of is provided, the Schottky characteristic can be obtained even if the gate electrode 8 is directly formed on a region of the cap layer 110 where impurities are not introduced.
The same effect as the first embodiment can be obtained.

(Third Embodiment) Next, a third embodiment of the present invention will be described with reference to FIG. Figure 5 shows HEMT10
2 is a schematic structural diagram showing the configuration of FIG. 2, and in this embodiment,
Undoped or 1 × 10 ¹⁷ cm ⁻³ in the second embodiment
A cap layer 110 (thickness: 3 nm) made of n-type GaAs doped to a carrier concentration of about 5 nm, and a cap layer 210 (thickness: 5 nm) made of undoped InGaAs.
The configuration is the same as that of the second embodiment except that the above is changed. With such a structure, the gate electrode 8 can be formed in Schottky contact on the region of the cap layer 210 where impurities are not introduced, and the same effect as that of the first embodiment can be obtained.

(Fourth Embodiment) A fourth embodiment according to the present invention will be described with reference to FIG. Figure 6 shows AlGaA
It is a typical structure figure showing composition of s / GaAs HEMT103. On a substrate 70 made of semi-insulating GaAs, a p-type GaAs layer 71 (film thickness 50 nm) light-doped to a carrier concentration of about 1 × 10 ¹⁶ cm ⁻³ and a channel layer 72 (film made of undoped GaAs) are formed. Thickness 50n
m), a spacer layer 7 made of undoped AlGaAs
3 (film thickness 5 nm), a doped layer 74 made of n-type AlGaAs doped to a carrier concentration of about 2 × 10 ¹⁸ cm ⁻³ (corresponding to a semiconductor layer made of n-type AlGaAs: film thickness 30 nm), undoped or A cap layer 75 (film thickness 5 nm) made of n-type GaAs doped with a carrier concentration of about 1 × 10 ¹⁷ cm ⁻³ is sequentially laminated.
Here, Al in the spacer layer 73 and the doped layer 74
The composition ratio of both is 0.3. With such a configuration, the gate electrode 8 can be formed in Schottky contact on a region of the cap layer 75 where impurities are not introduced,
The same effect as the first embodiment can be obtained.

(Fifth Embodiment) A fifth embodiment according to the present invention will be described with reference to FIG. Figure 7 shows AlGaAs / In
FIG. 3 is a schematic structural diagram showing a configuration of a GaAs HEMT 104. 1 on a substrate 80 made of semi-insulating GaAs
Lightly doped p-type GaAs layer 81 (thickness 50 nm), undoped GaAs layer 82 (thickness 50 nm), undoped In with a carrier concentration of about 10 ¹⁶ cm ^-3
A channel layer 83 (thickness 20 nm) made of GaAs, a spacer layer 84 (thickness 5 nm) made of undoped AlGaAs, and a doped layer 85 made of n-type AlGaAs doped to a carrier concentration of about 2 × 10 ¹⁸ cm ^−3. (N
Type AlGaAs semiconductor layer: 30n
m), an undoped layer or a cap layer 86 (thickness: 5 nm) made of n-type GaAs doped with a carrier concentration of about 1 × 10 ¹⁷ cm ⁻³ is sequentially laminated. here,
The Al composition ratio in the spacer layer 84 and the doped layer 85 is 0.15, and the In composition ratio in the channel layer 83 is also 0.15. With such a configuration, the cap layer 8
The gate electrode 8 can be formed in Schottky contact on the region where the impurity 6 is not introduced, and the same effect as that of the first embodiment can be obtained.

(Sixth Embodiment) A sixth embodiment according to the present invention will be described with reference to FIG. Figure 8 shows GaAs MESF
It is a schematic structural diagram showing a configuration of ET105. 1 × 10 ¹⁶ cm ⁻³ on a substrate 90 made of semi-insulating GaAs
P-type GaA light-doped to a carrier concentration
An s layer 91 (thickness 100 nm), an undoped GaAs layer 92 (thickness 50 nm), and a channel layer 93 (thickness 200 nm) made of n-type GaAs doped with a carrier concentration of about 4 × 10 ¹⁷ cm ^−3. It is sequentially laminated. With such a configuration, the gate electrode 8 can be formed in Schottky contact on the region of the channel layer 93 where impurities are not introduced, and the same effect as that of the first embodiment can be obtained.

In the above-mentioned first to third embodiments,
Although the buffer layer is provided on the substrate, the buffer layer is a layer provided to improve the crystallinity of the channel layer, and the buffer layer may not be provided if necessary. In the first to fifth examples, the spacer layer is provided on the channel layer. However, the spacer layer is a layer provided to improve the traveling speed of electrons in the channel layer, and is required. Therefore, the spacer layer may not be provided.

As indicated above, according to the present invention,
A p-type semiconductor layer is arranged on the substrate side of the channel layer, an insulating film having an opening including a region where source and drain electrodes are formed is formed on the cap layer, and n-type impurities are introduced from the opening. Forming an n-type impurity introduction region from the p-type semiconductor layer to the cap layer, forming source and drain electrodes on the cap layer through the opening, and forming a gate electrode on the insulating film. Thus, since the gate electrode can be formed without etching, a field effect transistor having excellent device characteristics can be realized.

[Brief description of drawings]

FIG. 1 is a schematic structural diagram showing a configuration of a first embodiment according to the present invention.

FIG. 2 is a schematic diagram showing a manufacturing method of a first embodiment according to the present invention.

FIG. 3 is a schematic structural diagram showing a current path in the first embodiment according to the present invention.

FIG. 4 is a schematic structural diagram showing a configuration of a second embodiment according to the present invention.

FIG. 5 is a schematic structural diagram showing a configuration of a third embodiment according to the present invention.

FIG. 6 is a schematic structural diagram showing the configuration of a fourth embodiment according to the present invention.

FIG. 7 is a schematic structural diagram showing the configuration of a fifth embodiment according to the present invention.

FIG. 8 is a schematic structural diagram showing the configuration of a sixth embodiment according to the present invention.

FIG. 9 Conventional InAlAs / InGaAs HEMT
Schematic structural diagram showing the configuration of.

[Explanation of symbols]

 1 InP Substrate 3 Undoped InGaAs Channel Layer 4 Undoped InAlAs Spacer Layer 5 n-type InAlAs Doped Layer 8 Gate Electrode 9 Source Electrode 10 Drain Electrode 11 n-type InGaAs Cap Layer 12 Insulating Film 21 Undoped InAlAs Buffer Layer 22 p-type InGaAs Layer 23 Undoped InAlAs Barrier layer 100 HEMT A n-type impurity introduction region

Claims (11)

[Claims] 1. A substrate, a p-type first semiconductor layer formed on the substrate, a second semiconductor layer formed on the first semiconductor layer and having no impurities added thereto, An impurity-added or n-type third semiconductor layer formed on the second semiconductor layer; a source electrode and a drain electrode formed on the third semiconductor layer; A gate electrode that is formed between the source electrode and the drain electrode on the semiconductor layer and controls a current flowing through the source electrode and the drain electrode by applying a voltage; and the source electrode and the drain electrode. Field effect transistor formed in the first, second, and third semiconductor layers at the bottom of the field effect transistor. 2. The gate electrode is formed on the third semiconductor layer via an insulating film.
3. The field-effect transistor according to claim 1. 3. The second semiconductor layer is made of InGaAs, and a semiconductor layer made of n-type InAlAs is arranged between the second semiconductor layer and the third semiconductor layer. The field effect transistor according to claim 1. 4. The second semiconductor layer is made of GaAs, and a semiconductor layer made of n-type AlGaAs is arranged between the second semiconductor layer and the third semiconductor layer. The field effect transistor according to claim 2. 5. The second semiconductor layer is made of InGaAs, and a semiconductor layer made of n-type AlGaAs is arranged between the second semiconductor layer and the third semiconductor layer. The field effect transistor according to claim 2. 6. The field effect transistor according to claim 2, wherein the second semiconductor layer is made of GaAs, and the third semiconductor layer is made of n-type GaAs. 7. The field effect transistor according to claim 1, wherein the first semiconductor layer is made of InGaAs. 8. The field effect transistor according to claim 1, wherein the third semiconductor layer is made of InGaAs. 9. The field effect transistor according to claim 2, wherein the third semiconductor layer is made of GaAs. 10. A step of forming a p-type first semiconductor layer on a substrate, a step of forming a second semiconductor layer to which impurities are not added on the first semiconductor layer, No impurities are added on the second semiconductor layer,
Alternatively, a step of forming an n-type third semiconductor layer, and a step of forming, on the third semiconductor layer, an insulating film having an opening at a predetermined portion including a portion where a source electrode and a drain electrode are formed. A step of introducing an n-type impurity into the first, second, and third semiconductor layers through the opening, a step of forming a gate electrode on the insulating film, the source electrode and the drain electrode Is formed on the third semiconductor layer through the opening, and a method for manufacturing a field effect transistor. 11. A step of forming a p-type first semiconductor layer on a substrate, a step of forming a second semiconductor layer to which impurities are not added on the first semiconductor layer, No impurities are added on the second semiconductor layer,
Alternatively, a step of forming an n-type third semiconductor layer, and a step of forming, on the third semiconductor layer, an insulating film having an opening at a predetermined portion including a portion where a source electrode and a drain electrode are formed. And a step of introducing an n-type impurity into the first, second, and third semiconductor layers through the opening, and after removing the insulating film, the gate electrode is insulated from the third semiconductor layer. And a step of forming the source electrode and the drain electrode on a region of the third semiconductor layer where the opening of the insulating film was formed, the step of forming the source electrode and the drain electrode on the region where the film was formed. A method of manufacturing a field effect transistor, characterized in that