A) Field of the Invention
The present invention relates to a semiconductor device and its manufacture method, and more particularly to a semiconductor device including minute transistors and its manufacture method.
B) Description of the Related Art
The integration degree of semiconductor integrated circuit devices is improved more and more. For high integration degree, transistors as constituent elements are made finer. Under the present developments, the gate length of a CMOS transistor formed by 90 nm rules is 40 nm or shorter. As a transistor is miniaturized, the short channel effects appear such as leak current due to punch-through.
In order to prevent the short channel effects, the source/drain regions are formed by extension regions having a shallow junction and outer source/drain regions having a deep junction. Even if shallow extension regions are formed by short range ion implantation, subsequent heat treatment at a high temperature diffuses doped impurities and deepens the junction depth.
It is therefore desired to perform heat treatment such as activation after the ion implantation process at a low temperature. As impurities are activated by a low temperature process, insufficient activation occurs and a transistor drive current may lower.
In order to prevent the punch-through between source/drain regions, the shallow extension regions are covered in some cases with pocket (halo) regions having a conductivity type opposite to that of the extension regions. For example, the pocket region is formed by ion implantation oblique to a substrate normal direction.
In order to realize a high performance semiconductor integrated circuit device, it is desired to improve the integration degree and retain or increase a transistor drive current.
As shown in
After a clean surface of the active region 104 is exposed, the silicon surface is thermally oxidized to form a gate insulating film 105. Thereafter, on the gate insulating film 105, a gate electrode layer 106 of polysilicon is deposited by chemical vapor deposition (CVD).
As shown in
As shown in
By using the gate electrode Gp and side wall spacers SW as a mask, p-type impurity ions are implanted deeply into the active region 104 to form deep high concentration source/drain regions 114. In this manner, a p-channel MOS (PMOS) transistor is formed. In manufacturing a CMOS device, each ion implantation process is performed independently by separating an n-channel MOS (nMOS) region and pMOS region with resist masks.
As a transistor is miniaturized, the gate length becomes short. If the conventional gate height is to be used, the gate height is too high so that it becomes unstable. As the scaling of transistors advances, it is desired to lower the gate height.
Boron (B) is mainly used as the p-type impurity of a pMOS transistor. As the gate height is lowered, in the process of implanting p-type impurity ions B for forming deep source/drain regions, the phenomenon occurs in which B ions implanted into the gate electrode pierce through the gate insulating film and reach the channel region. New countermeasures are desired to prevent B ions from piercing through the gate insulating film.
As shown in
By using the gate electrode 106 as a mask, p-type impurity ions B are implanted at a low acceleration energy to form shallow p-type extensions 111. Since ion implantation is performed at a low acceleration energy, the phenomenon is hard to occur in which B ions implanted into the gate electrode 106 pierce through the gate oxide film 105.
As shown in
As shown in
Since the upper portion of the gate electrode Gp is the amorphous layer 109, an ion implantation depth is constrained so that B ions are prevented from piercing through the gate oxide film. Since the amorphous layers are formed also in the active region 104, the ion implantation depth is constrained so that high concentration source/drain regions 114s having a constrained junction depth are formed.
Thereafter, implanted impurity ions are activated to complete a PMOS transistor. With this manufacture method, since the implantation depth of p-type impurity ions B is constrained, the phenomenon of piercing of B through the gate insulating film can be prevented.
However, the implantation depth of the high concentration source/drain regions is also constrained. An impurity concentration gradient of the high concentration source/drain regions becomes sharp. It is difficult for a depletion layer to widen when a negative voltage is applied to the drain region, so that parasitic capacitances of the source/drain regions increase. An increase in parasitic capacitance results in a lowered operation speed.
For example, Japanese Patent Laid-open Publication No. HEI-9-23003 discloses a pMOS transistor manufacture method in which after a gate electrode is formed, In ions are implanted to form p-type extension regions, side wall spacers are formed, Si ions are implanted for channeling prevention, and thereafter B ions are implanted to form high concentration source/drain regions.
An object of the present invention is to provide a semiconductor device manufacture method capable of forming a micro pMOS transistor which can operate at high speed and has a large drive current.
Another object of the present invention is to provide a semiconductor device manufacture method capable of lowering a gate electrode height, preventing piercing of B through a gate insulating film and suppressing an increase in parasitic capacitances of the source/drain regions.
Still another object of the present invention is to provide a semiconductor device having a pMOS transistor which has good stability, can operate at high speed, has a large drive current and can suppress the short channel effects.
Another object of the present invention is to provide a semiconductor device having a pMOS transistor which can constrain a gate electrode height, suppress B impurities from piercing through the gate insulating film and entering the channel region, and reduce parasitic capacitances of the source/drain regions.
According to one aspect of the present invention, there is provided a method for manufacturing a semiconductor device comprising steps of: (a) forming a gate insulating film on a semiconductor substrate including a first conductivity type active region defined by an element isolation region; (b) depositing a gate electrode layer of polycrystalline semiconductor on the gate insulating film; (c) implanting impurity ions to transform an upper portion of the gate electrode layer into an amorphous layer; (d) patterning the gate electrode layer to form a gate electrode; (e) forming side wall spacers on side walls of the gate electrode at a temperature not crystallizing the amorphous layer; and (f) implanting impurity ions of a second conductivity type into the first conductivity type active region by using as a mask the gate electrode and the side wall spacers, to form high concentration source/drain regions.
According to another aspect of the present invention, there is provided a semiconductor device comprising: a semiconductor substrate including a first conductivity type active region defined by an element isolation region; a gate insulating film formed on the first conductivity type active region; a gate electrode of polycrystalline semiconductor formed on the gate insulating film, the gate electrode containing impurities and second conductivity type impurities; side wall spacers formed on side walls of the gate electrode; high concentration source/drain regions formed by implanting ions of the second conductivity type impurities into the first conductivity type active region outside of the side wall spacers, the high concentration source/drain regions not containing the impurities; and a channel region defined in the first conductivity type active region under the gate electrode, the channel region not substantially containing the second conductivity type impurities for doping into the gate electrode.
According to still another aspect of the present invention, there is provided a semiconductor device comprising: a single crystal semiconductor substrate including a first conductivity type active region defined by an element isolation region; a gate insulating film formed on the first conductivity type active region; a gate electrode formed on the gate insulating film, the gate electrode including a polycrystalline lower layer and an amorphous upper layer and containing impurities and second conductivity type impurities; side wall spacers formed on side walls of the gate electrode; single crystal source/drain regions formed by implanting ions of the second conductivity type impurities into the first conductivity type active region outside of the side wall spacers and not by implanting ions of the impurities; and a single crystal channel region defined in the first conductivity type active region under the gate electrode, the single crystal channel region not substantially containing the second conductivity type impurities for doping into the gate electrode.
The present inventors have analyzed current technologies and studied possible methods for solving the conventional problems.
According to the technologies illustrated in
The abscissa represents temperature, low, middle and high temperatures, and the ordinate represents a degradation factor of a drain current in the unit of % where a drain current Id of a transistor having a gate electrode height of 70 nm and annealed at a high temperature is set to 100%. The higher the percentage, the degradation is larger.
The measurement results of nMOS transistors are shown in the left area of
The degradation of the drain current is large, particularly for PMOS. The drain current Id of a pMOS transistor at a gate electrode height of 100 nm and at low temperature annealing degrades by 30% or more than that of a pMOS transistor at a gate electrode height of 70 nm and at high temperature annealing. If the gate electrode height is set to 70 nm, the degradation of the drain current Id is smaller than 15% even at low temperature annealing.
In order to suppress the degradation of the drain current, it is therefore desired to set the gate electrode height to 100 nm or lower. As the gate electrode height is lowered, there arises the problem of piercing of B ions through the gate insulating film when deep and high concentration source/drain regions of a pMOS transistor are formed.
Samples were formed by depositing a polysilicon layer having a thickness of 200 nm and by vertically implanting B+ ions at an acceleration energy of 3 to 5 keV and a dose of 5×1015 cm−2. A distribution of a B concentration was measured by secondary ion mass spectroscopy (SIMS).
A curve s3 indicates the distribution of B in a depth direction when ion implantation is performed at an acceleration energy of 3 keV. Similarly, curves s4 and s5 indicate the distributions of B in the depth direction when ion implantation is performed at acceleration energies of 4 keV and 5 keV, respectively. As the acceleration energy is increased, the peak position of a B concentration moves to a deeper position. After the peak, the B concentration lowers. The curve s3 has a gentle reduction near at a depth of 40 nm. As compared to the curve s3, the curves s4 and s5 have B concentration lifted shapes from the peak to the depth of about 75 nm.
The distributions in the area at a depth of about 75 nm or deeper are generally the same, irrespective of the acceleration energy. No B concentration difference is recognized in the area at a depth of 80 nm or deeper, irrespective of the acceleration energy. At the depth of 75 nm, the B concentration is in the order of about 1019 cm−2. At a depth of 105 nm, the B concentration eventually becomes higher than 2×1018 cm−2. It can be anticipated from these results that as the gate electrode height is set low at 70 nm, a fair amount of B ions pierces through the gate insulating film and reaches the underlying channel region.
If B ions of a non-negligible amount pierce through the gate insulating film and enter the channel region, the threshold value of a pMOS transistor becomes unstable and the pMOS transistor cannot operate stably.
The B concentration distribution shown in
It is known that amorphousizing is effective for preventing channeling. It is also known that ion implantation of an element having a relatively large mass is effective for amorphousizing silicon single crystal. Conductivity imparting impurities such as As, Sb and In may be used. In order to avoid electric influences, neutral ions of the same group as that of silicon, Ge, Si and the like may be used. Ge among others has a large mass and is effective for amorphousizing.
As the acceleration energy increases, the peak value of the Ge concentration distribution moves to a deeper position and the whole concentration distribution moves to the deeper position. At the Ge concentration of 1×1019 atoms cm−3, as the acceleration energy is increased from 5 keV, to 10 keV, to 15 keV and to 20 keV, the depth becomes deeper from about 33 nm, to about 41 nm, to about 50 nm and to about 56 nm.
A curve b (g5) indicates a B concentration distribution when B+ ions are implanted after Ge ions are implanted at an acceleration energy of 5 keV. Similarly, curves b (g10) and b (g20) indicate B concentration distributions when B+ ions are implanted after Ge ions are implanted at acceleration energies of 10 keV and 20 keV, respectively. A curve b (g0) indicates a B concentration distribution when Ge ions are not implanted. A curve b (a-Si) indicates a B concentration distribution when B+ ions are implanted into an amorphous silicon layer instead of a polysilicon layer.
Although the curve b (g0) has a large skirt portion, the curve b (a-Si) has almost no skirt portion, indicating that the amorphous layer is effective for suppressing the abnormal distribution. The curve b (g20) has generally the same distribution as that of the curve b (a-Si), indicating that as Ge+ ions are implanted by about 1×1015 cm−2 at an acceleration energy of 20 keV, generally the same results as those of the amorphous silicon layer can be obtained.
Although the curve b (g5) shows the suppression of the abnormal distribution as compared to the curve b (g0) without Ge ion implantation, the suppression effects are limited. It can be considered that the acceleration energy of Ge+ ions of 5 keV is insufficient.
The curve b (g10) has a distribution like that of the curve b (g20), particularly in the shallow region, and suppresses the abnormal distribution considerably. Although it has a skirt in the deep region, its width is limited.
The B concentrations at a depth of 75 nm of the curves b (g0), b (g5), b (g10) and b (g20) are higher than 1×1019 cm−3, 6×1018 cm−3, 3×1018 cm−3, and about 5×1017 cm−3, respectively.
In order to suppress the B abnormal distribution, it can be considered that Ge ion implantation is executed in an acceleration energy range of 10 keV to 20 keV. The suppression effects are small at an acceleration energy lower than 10 keV. At an acceleration energy higher than 20 keV, it is hard to expect the suppression effects to be improved more. Conversely, there is a possibility that Ge pierces through the gate insulating film and is doped in the channel region, adversely affecting the electric characteristics of the channel region.
It is confirmed that an amorphous layer formed by implanting Ge ions into the gate electrode prior to B ion implantation into the source/drain regions and gate electrode, is effective for constraining the depth of the subsequent B ion implantation. However, if Ge ions are implanted into the silicon substrate, the source/drain regions become shallow. It is preferable not to perform Ge+ ion implantation into the silicon substrate in order to widen the B concentration distribution in the source/drain regions, to form a junction at a sufficiently deep position, and to reduce parasitic capacitances.
In the following, description will be made on main processes of a semiconductor device manufacture method according to an embodiment of the invention.
As shown in
After the wells are formed, a gate oxide film 5 having a thickness of, e.g., about 1 nm, is formed on the clean surface of the active region, by thermal oxidation. On the gate oxide film 5, a polysilicon layer 6 thinner than 100 nm, e.g., about 75 nm, is formed by thermal CVD.
As shown in
Ge ion implantation is preferably executed in an acceleration energy range of 10 keV to 20 keV. At an acceleration energy lower than 10 keV, the amorphousizing effects are small and the abnormal distribution suppression effects of the subsequent B ion implantation are small. At the acceleration energy of 20 keV, B ion implantation presents the sufficient abnormal distribution suppression effects approximately equal to those of a-Si.
As shown in
If the subsequent B ion implantation provides a sufficiently high concentration, the above-described B ion implantation may be omitted. In this case, the mask 8 may be omitted for Ge ion implantation shown in
The execution order of the processes shown in
As shown in
As shown in
Since the acceleration energy is low and the upper portion of the gate electrode layer is the amorphous layer 9, implanted B ions will not pierce through the gate insulating film. P+ ions are implanted at an acceleration energy of 10 keV and a dose of 1×1013 cm−2 to form pocket regions Pn. The pocket regions are effective for suppressing the short channel effects.
After the resist mask 10 is removed, a new mask is formed covering the PMOS region and ion implantation processes for the nMOS region are performed to form shallow n-type extension regions and p-type pocket regions. For example, As as n-type impurities is implanted at an acceleration energy of 1 keV and a dose of 1×1015 cm−2, and B as p-type impurities is implanted at an acceleration energy of 7 keV and a dose of 1×1013 cm−2.
As shown in
A silicon oxide film having a thickness of, e.g., 80 nm, is deposited on the whole surface of the silicon substrate by low temperature CVD at a temperature of, e.g., 600° C. The silicon oxide film is subjected to reactive ion etching (RIE) to remove the silicon oxide on the flat surface. Side wall spacers SW of the silicon oxide film are therefore formed only on the side walls of the gate electrodes Gp and Gn.
As shown in
Therefore, p-type impurity ions B are implanted into the gate electrode Gp made of a lamination of the amorphous silicon layer and polysilicon layer and into the single crystal silicon regions outside the side wall spacers SW. A B abnormal distribution in the gate electrode Gp is suppressed by the amorphous silicon layer 9p. The channel region (n-well) 4 under the gate electrode does not substantially undergo B ion implantation.
If the whole thickness of the gate electrode layer is transformed into an amorphous layer, impurities under the gate electrode are not sufficiently activated by subsequent activation, and activation insufficiency occurs. As the polysilicon layer 6p itself is used as the lower portion of the gate electrode, subsequent impurity activation can be performed properly.
Since an amorphous layer does not exist in the single crystal region, B ions are distributed deeply having a skirt portion, and it becomes possible to form the source/drain regions 14 deep enough to form small junction capacitances.
After the ion implantation for the source/drain regions in the pMOS region, the resist mask 13 is removed and a new resist mask is formed covering the pMOS region. In the nMOS region, for example, P+ ions are implanted at an acceleration energy of 6 keV and a dose of 5×1015 cm−2 to form deep high concentration n-type source/drain regions. Even if an amorphous layer does not exist in an nMOS transistor, there is no problem because piercing of n-type impurity P through the gate insulating film is not still recognized.
However, if the gate electrode becomes further low, there is a possibility that n-type impurity P pierces through the gate insulating film. In this case, the Ge ion implantation shown in
As shown in
In the above manner, a pMOS transistor and an nMOS transistor are formed. Thereafter, by using well-known processes, an interlayer insulating film, lead wirings, multilayer wirings and the like are formed to complete a semiconductor integrated circuit device. For general semiconductor integrated circuit manufacture processes, for example, refer to U.S. Pat. Nos. 6,465,829, 6,492,734, and 6,707,156, and US publication U.S. 2003/0227086 A1, the whole contents of which are incorporated herein by reference.
If the concentration of the channel region is N (ch), the junction depth formed by the concentration distribution b2 becomes much shallower than the junction depth formed by the concentration distribution b1, and the B concentration lowers sharply near the junction.
In the case of the junction formed by the concentration distribution b1, the p-type impurity concentration gently lowers near the junction, and a broad depletion can be formed easily. It is therefore possible to maintain small the parasitic capacitances of the source/drain regions. In the case of the junction formed by the concentration distribution b2, p-type impurity concentration lowers steeply near the junction. Formation of a broad depletion is suppressed and the parasitic capacitances of the source/drain regions become large.
Since the gate electrode has the amorphous layer, the concentration distribution with the skirt portion shown by the curve b1 is not formed, but the junction depth is constrained as indicated by the curve b2. It is therefore possible to efficiently prevent B ions from piercing through the gate insulating film.
B impurities are not substantially doped into the channel region under the gate electrode. The channel region under the gate electrode does not substantially contain B impurities used for doping into the gate electrode and has the B concentration distribution substantially the same as that of the regions under the side wall spacers SW. The term “substantially” has a meaning to be used when the electric characteristics are taken into consideration.
If the active region surface is amorphousized, the B concentration distribution is constrained when the source/drain regions are formed, and shallow source/drain regions 14x are formed. The impurity concentration distribution changes steeply, and as described above, the depletion of the p-type source/drain regions 14x is constrained and the parasitic capacitances of the source/drain regions increase.
The impurity concentration of the channel region changes in the depth direction with the threshold value adjustment ion implantation and the like. As the junction depth moves into the threshold value adjustment region 7, the impurity concentration of the channel region increases and the high concentration p-type region contacts the high concentration n-type region, so that a large parasitic capacitance is formed.
If a suicide layer 21 is formed on the substrate surface, a distance between the suicide layer and the pn junction becomes short, forming the reason of leak current. Since the deep source/drain regions 14 are formed, it is possible to suppress an increase in leak current even if the silicide layer 21 is formed.
The present invention has been described in connection with the preferred embodiments. The invention is not limited only to the above embodiments. For example, process parameters can be changed in various ways in accordance with the design. A plurality type of transistors and different type of elements such as passive elements can be integrated. It will be apparent to those skilled in the art that other various modifications, improvements, combinations, and the like can be made.
The above-described embodiments are suitable for semiconductor integrated circuit devices of high integration degree.
This application is a continuation application of an international patent application, PCT/JP2003/006898, filed on Mary 30, 2003, the entire contents of which are incorporated herein by reference.
Number | Date | Country | |
---|---|---|---|
Parent | PCT/JP03/06898 | May 2003 | US |
Child | 11169666 | Jun 2005 | US |