Embodiments described below relate to a nonvolatile semiconductor memory device and a manufacturing method thereof.
In a nonvolatile semiconductor memory device, for example, a NAND type flash memory, there is a demand to increase an amount of charge that can be accumulated in a floating gate.
A nonvolatile semiconductor memory device of an embodiment described below comprises: a semiconductor layer; agate insulating film formed on the semiconductor layer; a floating gate formed on the gate insulating film and including silicon as a material thereof; a charge accumulation film formed on a surface of the floating gate; a block insulating film formed on the charge accumulation film; and a control gate formed on the block insulating film. An upper layer of the floating gate includes germanium and boron.
Next, the nonvolatile semiconductor memory device according to the embodiment will be described in detail with reference to the drawings.
First, the nonvolatile semiconductor memory device according to the embodiment will be described with reference to
[Schematic Configuration]
A plurality of the memory cells MC aligned in a bit line BL direction are connected in series by sharing between them source/drain diffusion layers as will be mentioned later. The plurality of memory cells MC connected in series configure one memory string. One end of the memory string is connected to the bit line BL via a drain side select gate transistor SG1. The bit line BL and the drain side select gate transistor SG1 are connected via a contact 22.
In addition, the other end of the memory string is connected to a source line SL not illustrated, via a source side select gate transistor SG2. The source line SL and the source side select gate transistor SG2 are connected via a source side contact 33.
A gate of the drain side select gate transistor SG1 is connected to a drain side select gate line (SGD) 13A arranged in parallel to the word line WL. In addition, a gate of the source side select gate transistor SG2 is connected to a source side select gate line (SGS) 13B arranged in parallel to the word line WL. Here, a direction in which the word line WL extends is defined as a word line direction (X direction), and a direction in which the bit line BL extends is defined as a bit line direction (Y direction).
As shown in
A plurality of the memory cells MC each comprise: a plurality of source/drain diffusion layers 12 disposed on the surface of the semiconductor substrate 10; agate insulating film 13 (tunnel insulating film) disposed on a channel region between these source/drain diffusion layers 12; and a floating gate 14 disposed on the gate insulating film 13. A film thickness of the gate insulating film 13 may be set to, for example, about 6 nm. Moreover, a film thickness of the floating gate 14 may be set to, for example, about 10 to 25 nm. This floating gate 14 includes in its upper layer a germanium (Ge)-segregated layer 14A into which Ge is implanted and in which a silicon density is lowered. This Ge-segregated layer 14A has boron (B) implanted therein in this region where the silicon density is lowered. This point will be described in detail later.
Note that in the case where a distance between the memory cells MC is small, it is also possible for the source/drain diffusion layer 12 to be omitted. This is because due to a so-called fringe effect, a conduction path penetrating the channel region of the plurality of memory cells MC can be generated even if there is no the source/drain diffusion layer 12.
Furthermore, this memory cell MC comprises a charge accumulation film 15 disposed on the floating gate 14. This charge accumulation film 15 has a function of accumulating a charge injected into the floating gate 14 via the gate insulating film 13 by a write operation, and is formed by, for example, silicon nitride (SiN). A film thickness of the charge accumulation film 15 may be set to, for example, about 2 nm. Existence of the charge accumulation film 15 makes it possible for an aspect ratio of the floating gate 14 to be reduced.
Formed on this charge accumulation film 15 is a block insulating film 16. As an example, this block insulating film 16 is configured by a first insulating film 16A formed from hafnium oxide (HfOx), a second insulating film 16B formed from silicon oxide (SiO2), and a third insulating film 16C formed from hafnium oxide (HfOx).
Deposited on this block insulating film 16 via a barrier metal 17 is a conductive film 18 acting as the word line WL. As an example, film thicknesses of the first insulating film 16A, the second insulating film 16B, and the third insulating film 16C may each be set to about 5 nm. In this example illustrated in
As an example, the barrier metal 17 includes a first metal film 17A formed from tantalum nitride (TaN) and a second metal film 17B formed from tungsten nitride (WN). Film thicknesses of the first metal film 17A and the second metal film 17B may each be set to about 5 nm. In addition, the conductive film 18 is formed by a metal such as tungsten (W).
Note that the block insulating film 16, although it has a three-layer structure in the illustrated example, is not limited to this structure, and the block insulating film 16 may also be configured as a single-layer structure formed by a single material. Moreover, there may also exist a boundary layer between the charge accumulation film 15 and the block insulating film 16 and between the block insulating film 16 and the barrier metal 17.
Note that although omitted from illustration, patterning of the bit line (BL) is executed by depositing the conductive film 18 which is to be the word line WL and forming a film of a mask material thereon, and then processing the gate insulating film 13 through conductive film 18 to the semiconductor substrate 10 with a width of approximately 20 nm by photolithography and etching. Moreover, the source/drain diffusion layer 12 is formed by performing ion implantation of an n type impurity (phosphorus (P), and so on) in a self-aligning manner using the gate electrode formed in this way as a mask, and then diffusing the impurity by heat processing of 955° C. for 30 seconds.
The memory cell shown in
[Formation Method of Ge-Segregated Layer 14A]
A formation method of the previously mentioned Ge-segregated layer 14A will now be described with reference to the flowchart of
After forming a polysilicon film which is to be the floating gate 14, a surface of that polysilicon film undergoes ion implantation of germanium (Ge) with, for example, the implanted dose of 2.5×1015 cm−2 at the acceleration energy of 0.5 keV (step S1). In this step S1, a depth of a position where concentration of germanium ions becomes the maximum concentration depth (Rp) is approximately 2.5 nm. Note that the Rp is a depth from the surface of the floating gate 14. Moreover, a spreading width ΔRp of 1σ of the maximum concentration depth (Rp) is approximately 1.2 nm. Due to such a step, the germanium is distributed at a depth of within several nm from the surface of the floating gate 14.
Furthermore, density of germanium at the maximum concentration depth in this step S1 is approximately 7×1021 cm−3, and is a density exceeding 10 percent of silicon density (approximately 5×1022 cm−3). Due to this ion implantation of germanium whose atomic radius is larger than that of silicon, excessive knock-on of silicon atoms occurs in a vicinity of the maximum concentration depth (Rp) from the surface of the polysilicon film. This knock-on causes a condition where silicon density lowers extremely.
Next, after completion of step S1, a region where germanium ions have been implanted is further implanted with boron (B) (step S2). As an example, the ion implantation of boron (B) is performed with the dose of 5×1014 cm−2 at the acceleration energy of 0.2 keV. By so doing, the boron (B) can be implanted at high concentration into the above-mentioned region where silicon density has lowered. Furthermore, the boron (B) formed here is clustered (B12 cluster), and can be stably maintained due to the existence of the germanium.
In this step S2, a depth of a position where concentration of boron (B) becomes Rp′ (maximum concentration depth) is approximately 1.7 nm. Note that Rp′ is a depth from the surface of the floating gate 14. This depth substantially matches a position of the region formed by the germanium ion implantation where Si density has lowered. Furthermore, a maximum concentration of boron (B) is approximately 2×1021 cm−3, and sufficiently fulfills a density required for formation of metallic boron (B12 cluster).
After the ion implantation of boron, heat processing is performed at the temperature of 600° C., for the heating time of 1 hour, in nitrogen ambient atmosphere (step S3), and a step for recovering a damaged region caused by boron ion implantation is executed. Due to the above, formation of the Ge-segregated layer 14A is completed.
[Write Characteristics and Erase Characteristics of Memory Cell of Present Embodiment]
In the case of the floating gate 14 in which the Ge-segregated layer 14A is formed according to the steps of
Next, advantages of forming this Ge-segregated layer 14A will be described with reference to the energy band diagram of
[Advantages of Germanium Ion Implantation]
Next, advantages due to ion implantation of germanium will be described with reference to
In
On the other hand,
[Advantages of Implantation of Boron (B)]
Next, advantages of ion implantation of boron (B) will be described with reference to
Therefore, it is made possible to generate carrier density greater than or equal to the density of solid-solubility limit without performing heat processing.
However, the B12 cluster in the silicon is decomposed with increasing temperature of the heat processing temperature and increasing time, and its concentration eventually returns to the density of solid-solubility limit in the silicon. In order to enable the B12 cluster to exist stably in the silicon, restriction of thermal-budget is required, and application to a device is difficult.
On the other hand, when a first heat processing is executed at a low temperature and for a long time of 600° C. and 1 hour and then a second heat processing is performed at a high temperature and for a short time of 955° C. and 30 seconds, then, contrary to when only heat processing at a high temperature and for a short time is performed, a peak of the B is spectrum of boron appears in a vicinity of a binding energy of 188.1 eV. This means that when a low-temperature long-time heat processing is once performed and then a high-temperature short-time heat processing is performed, 5-coordinate boron (that is, the B12 cluster) can be formed to exist stably. This change is due to the Ge-segregated layer 14A in the surface layer of the floating gate 14. The temperature of the first heat processing is not strictly limited to 600° C., and as an example may be set in the range of 525° C. to 600° C.
On the other hand, a graph of the case where only high-temperature short-time heat processing is performed without low-temperature long-time heat processing being performed is shown by curve B in the graph of
These results show that performing heat processing at a low temperature and for a long time of 600° C. and 1 hour and then performing heat processing at a high temperature and for a short time of 955° C. and 30 seconds results in formation of the Ge-segregated layer 14A having germanium segregation in the surface of the polysilicon of the floating gate 14. These results also show that by having the B12 cluster incorporated into the Ge-segregated layer 14A, a thermally stable binding state is maintained. This is thought to be due to silicon atoms being excessively knocked-on by high-concentration germanium ion implantation, and silicon atom density lowering near the surface. In the low-temperature heat processing at 600° C., solid phase growth from the amorphous layer formed by ion implantation proceeds, but its diffusion coefficient is small. Hence, the state of lowered density of silicon atoms in the surface layer of the polysilicon layer is maintained, and Ge—Ge binding becomes dominant.
On the other hand, in the high-temperature heat processing at 955° C., it is thought that the diffusion coefficient also increases, hence diffusion of Si atoms proceeds and lattice strain is relaxed, and a change to SiGe binding proceeds.
Regarding binding of boron in germanium, the atomic radius of germanium is 0.122 nm, whereas the atomic radius of boron is 0.088 nm, which is 30 percent smaller than that of germanium.
By substituting germanium with boron, a strain occurs owing to the lattice contracts. Therefore, by having crystal recovery performed by an heat processing, boron whose atomic radius is small is thought to be clustered, thereby being incorporated into a lattice position and crystal strain relaxes. The size of a B12 cluster is 0.522 nm, which is different from the size of 0.566 nm of tetrahedral structure configured from five germanium atoms by approximately 8%. Therefore, it is thought that stable existence can be more easily achieved by clustered boron rather than substituting by monoatomic boron whose atomic radius is small.
A B12 cluster with a icosahedral structure consisting of 12-boron atoms, by being substituted for a 4-coordinate crystal lattice of silicon or germanium, provides a divalent electron-deficiency state. Therefore, carriers are generated at a ratio (carrier generation ratio) of 2/12=1/6 to the boron density. The carrier generation ratio is smaller than that of monoatomic boron, but the density of solid-solubility limit is large, hence a high carrier concentration can be obtained.
Moreover, in conjunction with the above, the graph of
The graph of the case where boron is implanted without performing germanium ion implantation and then heat processing at 700° C. for 30 seconds is performed is shown by curve P in
Moreover, the graph of the case where boron is implanted without performing germanium ion implantation and then heat processing at 955° C. for 30 seconds is performed is shown by curve Q in
On the other hand,
Therefore, considering that boron existing at or above the density of solid-solubility limit is easily clustered, a generation ratio of the B12 cluster is thought to be higher in germanium. Although the B12 cluster has a low activation ratio, i.e., at 1/6, 1/6 of carriers can be generated with respect to the boron density. Due to this, a super-saturation activation layer of boron over the density of solid solubility limit, i.e., 2×1020 cm−3 can be formed stably in silicon. This is what
In the region where silicon atoms are knock-on excessively due to high concentration germanium ion implantation to cause silicon-atom density to be lowered in a vicinity of the surface and germanium to be segregated (the Ge-segregated layer 14A), the ion implantation dose of germanium is important.
On the other hand, at the germanium implantation dose of 1×1015 cm−2 or more, heat-processing temperature dependency of sheet resistance is found to be hardly observed. Germanium density at the maximum concentration depth at this implantation dose is approximately 5×1021 cm−2. It may be estimated from this result that the density of germanium with which segregation of germanium is formed due to generation of the excessive knock-on silicon region and low-temperature heat processing is 5×1021 cm−3. Thus, it is thought that by performing ion implantation of germanium of the density of 5×1021 cm−3 or more, the stable B12 cluster is formed due to segregation of germanium due to generation of the excessive knock-on silicon region and low-temperature heat processing is formed. In other words, it is thought that by performing ion implantation of germanium of approximately 10% or more in density with respect to density of silicon according to the lattice constant thereof, the stable B12 cluster is formed due to segregation of germanium due to generation of the excessive knock-on silicon region and low-temperature heat processing is formed.
[Other]
While certain embodiments of the inventions have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the inventions. Indeed, the novel methods and systems described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the methods and systems described herein may be made without departing from the spirit of the inventions. The accompanying claims and their equivalents are intended to cover such forms or modifications as would fall within the scope and spirit of the inventions.
This application is based on and claims the benefit of priority from prior U.S. Provisional Patent Application No. 61/950,877, filed on Mar. 11, 2014, the entire contents of which are incorporated herein by reference.
Number | Date | Country | |
---|---|---|---|
61950877 | Mar 2014 | US |