High Ion/Ioff SOI MOSFET using body voltage control

[0001] The present disclosure is related to a semiconductor device and, more particularly, to a partially-depleted silicon-on-insulator (SOI) MOSFET.

[0002] Manufactures of semiconductor devices continually strive for integrated circuits of higher density and devices of smaller dimensions in order to improve cost efficiencies and performance. However, these scaling efforts have resulted in some undesirable effects in leakage current, e.g., the off-state leakage current (Ioff) of a MOSFET.

[0003] Excess leakage current (Ioff) in these MOSFETs may result in high power consumption for devices of large-scale integrated circuits, such as memory, processors, controllers, communication and networking circuits, etc. The high power consumption, in turn, can adversely affect requirements for heat dissipation. Additionally, in the case of some hand-held products, the integrated circuits with excess leakage may reduce the operating-time that may be available within a given fixed-energy battery life. Accordingly, there may be a demand for low-leakage MOSFETs.

[0004] Of the bulk and silicon-on-insulator MOSFET, the SOI MOSFET can offer lower parasitic junction capacitance and superior sub-threshold voltage swing. Such benefits may assist high-speed, current drive applications.

[0005] In general, SOI MOSFETs can be characterized into two different categories: fully-depleted and partially-depleted SOI MOSFETS. Fully-depleted SOI (FD-SOI) MOSFETs conventionally comprise a very thin body region, which may be fully depleted of majority carries under the influence of a standard gate bias. Partially-depleted SOI (PD-SOI) MOSFETs, on the other hand, may have a thicker body that may not be fully depleted during normal gate bias. This remaining portion that is not depleted may be described as a “floating body” region. Typically, the partially-depleted versus fully depleted SOI MOSFET are more commonly used in integrated circuit designs, which may be due to a variety of considerations such as density constraints and their greater ease of manufacturing.

[0006] For some embodiments of the present invention, charge may be injected into a floating body of a partially-depleted SOI MOSFET during device operation. The charge may accumulate in the floating body to establish a voltage build-up that may influence an overall performance of the MOSFET. In other words, the floating body may “float” to an increased voltage magnitude based on the amount of accumulated injected charge. This attributed may be tailored for a desired performance.

[0007] The floating-body partially-depleted SOI MOSFET provides a number of features that may benefit a variety of different circuits. In one embodiment, a dynamic-kink effect—i.e., wherein the body-to-source voltage may increase in an N-channel device (and decrease in a P-channel device) as a result of a gate-to-body coupling—can help decrease the magnitude of a threshold voltage (|Vt|) for assisting drive current capability. This in turn can be used to reduce propagation delays for inverter applications. In another embodiment, a drain-to-body coupling can provide a dynamic-loading effect to decrease the body-to-source voltage for an N-channel device and increase the body-to-source voltage for a P-channel device, thereby effective to increase |Vt| and reduce the load (gate capacitance) of a given MOSFET. This might also be used to reduce propagation delays.

[0008] In accordance with another embodiment of the present invention, a partially-depleted SOI MOSFET may comprise source and drain regions over an insulator and a body region therebetween. The floating body region may receive injected carriers from an adjacent region of semiconductor material during its operation.

[0009] In another embodiment, the floating body region of a partially-depleted SOI MOSFET may receive a voltage. The voltage may be controlled to adjust the performance of the MOSFET device dependent on a particular application. For example, the voltage for the floating body region may be adjusted to establish a desired voltage threshold magnitude (|Vt|) for the MOFET dependent on its desired use, such as, e.g., for isolation, current drive, transparency, propagation characteristics, impedance, etc.

[0010] In another embodiment, a semiconductor device may comprise a memory cell and an SOI MOSFET to access the memory cell. The MOSFET may comprise a body between a source and drain. An additional region of semiconductor material may be disposed in contiguous relationship with a drain/source region of the MOSFET and may be formed with a conductivity of type opposite the drain/source. The body, drain/source and the additional region of semiconductor material may thus be described an inherent bipolar transistor.

[0011] In a further embodiment, at least one of the parameters of the bipolar device—e.g., the drain/source implant dose, its implant energy, the drain width, a thickness of a silicide over the drain, the placement of another silicide over the additional semiconductor region, a lifetime control implant, a halo implant or the gate electrode placement—may be structured to establish a gain for the inherent bipolar device of a magnitude sufficient to enable carrier injection into the floating body, yet small enough to prevent MOSFET inter-latching.

[0022] In the following description, numerous specific details are set forth to provide an understanding of exemplary embodiments of the present invention, wherein similar elements between the various embodiments may be annotated similarly.

[0023] Additionally, readily established circuits or elements of the exemplary embodiments may be disclosed in simplified form (e.g., simplified block diagrams and/or simplified description) to avoid obscuring an understanding the embodiments with excess detail. Likewise, to aid a clear and precise disclosure, description of known structures—e.g., sidewall spacers, gate oxides, hallo regions, conductive lines, contacts, vias, etc.—may similarly be simplified where persons of ordinary skill in this art can readily understand such structures and provisions by way of the drawings and present disclosure.

[0024] As used herein, “substrate” or substrate assembly may be meant to include, e.g., a portion of a semiconductor wafer. Such portion may have one or more layers of material that have been formed on or within the substrate. These layers and/or additional layers formed thereon, may be patterned and/or may comprise dopants to produce devices (e.g., transistors, diodes, capacitors, interconnects, etc.) for an integration of circuitry. In forming these devices, one or more of the layers may comprise or result in topographies of various heights. When referencing this integration of circuitry, it may be described as integrated together, or on or with the substrate.

[0025] Furthermore, those skilled in the art will recognize that although an example of the present invention may describe fabrication of an N-type MOS transistor, the dopant type(s) of the substrate in addition to the doped regions thereof may be reversed to form a P-type MOS device. And, in accordance with further embodiments, both N-type and P-type devices could be formed on a common substrate.

[0026] As used hereinafter, “floating” may be used to characterize the relationship of a particular portion of a body to a MOSFET relative to other portions thereof. The descriptor “floating”, in one sense, may characterize the relationship of the body relative to an underlying bulk substrate. It may also characterize a voltage potential thereof that may be “floating” separately relative to other regions of the MOSFET. Therefore, in accordance with particular embodiments disclosed herein, the floating body might obtain an electrical potential that may be different from the potential of a channel region of the MOSFET.

[0027] In exemplary embodiments of the present invention, a source/drain region of a partially-depleted SOI MOSFET may propagate charge across the source/drain to influence the bias of the non-depleted region. Despite such influx of charge thereto, the non-depleted region may still to be described as a floating body.

[0028] As referenced herein, portions of a transistor may be described as formed in, at or on a semiconductor substrate. Such alternative terms on/in/at may be used individually merely for purposes of convenience. In the context of forming semiconductor devices, such terms may collectively reference portions of a semiconductor element that may be within and/or on a starting material. For example, with reference to device 100 of FIG. 1A, a silicon-on-insulator (SOI) metal-oxide-semiconductor field-effect transistor (MOSFET) may comprise source 102 and drain 104 regions within semiconductor layer 106. Body region 105 may be disposed between the source and drain regions. Gate 112 (e.g., conductive material over a gate oxide (not shown)) may be disposed on the upper surface of semiconductor layer 106 and over surface 109 of body 105. Thus, the description of device 100 at a region of a semiconductor substrate may be understood, dependent upon its relative context, to comprise, e.g., source, body and drain regions within semiconductor layer 106 over insulator layer 108 and over bulk substrate 110.

[0029] Referencing FIG. 1A, in accordance with an embodiment of the present invention, a MOSFET 100 may be formed in a layer of semiconductor material 106 disposed over an insulator, e.g., buried oxide 108 on supporting substrate 110. In a particular example, the semiconductor material may comprise silicon of, e.g., P-type dopant. N-type dopant may be implanted into select regions of the layer of semiconductor material may define the source and drain regions about body region 105.

[0030] In operation, a positive potential may be applied to the gate 112 to establish an electric field within body region 105. This, in turn, may attract and accommodate minority carriers of the body to the surface region of the body beneath the gate. The accumulation may form a conductive inversion channel (N-channel) between the source and drain regions. Conversely, relative to FIG. 2A, a P-channel MOSFET 100′ might receive a negative voltage for biasing its gate to effect a P-channel within an N-type body between its source and drain regions. For the following discussion, an N-MOSFET and appropriate biases therefore may be used for description of various embodiments; it will be understood, however, that the same principles may also apply to a P-MOSFET, in which the roles of electrons and holes are interchanged, and positive biases replaced by negative biases.

[0031] Further referencing FIG. 1A, the partially-depleted SOI MOSFET 100 may further comprise another doped region 114 (drain2) formed in contiguous relationship with drain 104 (drain1). The extra region may contact drain region 104 on a side thereof opposite to body region 105. With the drain 104 disposed electrically in series between body 105 and the additional region 114, the collection together may be viewed as an inherent bipolar device. For an NMOS of N-type source, P-type body and N-type drain regions, an additional region of P-type material may be formed against one of the drain/source N-regions. The resulting device may be deemed a hybrid combination of a MOSFET 122 (source, body, drain1) and an inherent bipolar transistor 124 (see FIG. 1B). The body 105, drain1 104 and drain2 114 may be alternatively described as collector, base and emitter respectively of the bipolar device. In the case of an NMOS 122 (of a N-type source, P-type body and N-type drain) the additional region 114 of P-type conductivity may serve as an emitter region of a PNP bipolar device 124. Conversely, referencing FIGS. 2A–2B, for a PMOS device 132, the extra dopant region 114′ of N-type conductivity may establish emitter 114′ to an NPN bipolar device 134.

[0032] In operation, referencing FIG. 3 and in accordance with a particular embodiment, a positive bias Vdd may be coupled through a conductive line 116 to the extra dopant region 114 (e.g., drain2 of P-type conductivity). Minority carriers (holes) may be injected into the drain region 104 (drain1 of N-type conductivity). Traversing the drain region 104, these carriers 125 may then enter body region 105 of the MOSFET 122. Accordingly, the floating body of the MOSFET 122 may accumulate residual charge and build-up a potential with respect to the source 102. In this particular example, the source may be coupled through a conductive line 120 to a ground potential. With sufficient charge injection, the floating body may obtain a voltage level that may effectively decrease the transistor's threshold voltage Vt, which may enhance the resulting drive current. Although the extra dopant region 114 is illustrated in this explanation as being coupled to Vdd; it will be understood that alternative embodiments may bias or couple this region to a alternative node, supply, integrated member, conductor, or contact. Similarly, source region 102 could be configured for a bias voltage other than ground.

[0033] Referencing FIG. 4, the current (I) versus voltage (Vgs) characteristics 140 for a carrier injection enhanced (body-injection) MOSFET may be compared relative to the I-V characteristics 150 for a typical MOSFET. One performance parameter that may be taken from this curve is the subthreshold swing, which may be defined by an amount of gate voltage increase required to produce a decade change in drain current. For particular embodiments disclosed herein, this subthreshold swing for the carrier injection enhanced MOSFET may be substantially reduced, which may be due to an ever-increasing level of carrier injection into the floating body for enhancing its voltage magnitude as the gate voltage ramps upwardly. Such property for reduced subthreshold swing may enable the formation of circuits with enhanced switching sensitivity; wherein small changes in gate voltage may control switching of the device from an off-state condition of very low leakage (Ioff) to an on-state condition of high drive (Ion).

[0034] For a conventional bulk MOSFET, the subthreshold swing presented along curve 150 might be limited to around 75 mV/decade (theoretically, its lowest level would be limited to 60 mV/decade at room temperature). For a partially-depleted SOI MOSFET with the body-injection provisions, a kink effect 142 (caused by injection of carriers 125 at a given bias level Vx) appropriately positioned along the I-V curve by design can effect a low subthreshold swing for toggling between the on-state and off-state conditions. When placed appropriately, the kink can substantially reduce the subthreshold swing and the change in gate-to-source voltage required to control the operation of the MOSFET.

[0035] In particular embodiments, the gain (e.g., beta or β_PNP) for the carrier transport and injection by the inherent bipolar device (124 of FIG. 1B) may be used to control the placement of the kink effect 142. As one parameter of the device's many possible influences, an appropriately designed gain for the carrier transport and injection can serve to position the kink operatively along the I-V curve for simplified control and switching between states. For example, with the appropriate carrier transport gain through the drain region for injection into the floating body, a subthreshold swing may be obtained that is substantially less than the 60 mV/decade.

[0036] Various solutions may be used to establish the carrier transport gain, and likewise the amount of carrier injection enhancement for placement of the kink effect. Some exemplary embodiments determine or define a gain therefore of a magnitude sufficient to assist the carrier injection into the body region, yet small enough to guard against the inter-latching of the MOSFET. For example, one form of solution may comprise impacting the gain via the amount of dopant provided for the base region over a base integral length relative to the amount of dopant provided for the emitter region over an emitter integral length thereof, wherein a design for a target ratio therebetween may be guided or established by general principals of available Gummel equation relationships and understandings of bipolar transistor theory.

[0037] For the base region (drain), its thickness, overall width, and doping profile may each or collectively be designed to, at least in part, establish the desired gain. In one case, silicide 264 may be disposed over a portion of the drain 204 (see FIG. 5), which may have the effect of decreasing the effective thickness of the base region. Additionally, at the boundary between the silicide and silicon semiconductor material (the surface of the silicide), an infinite recombination rate may exist, thereby effectively reducing the overall bipolar gain. During operation, the percentage of holes that might otherwise traverse the base may be reduced as a result of the recombination occurring at the silicide interface.

[0038] In another particular embodiment, silicon layer 108 may comprise a thickness, e.g., between 200 and 5,000 angstroms or of about 1,000 angstroms. The dopants implanted therein may have been formed using energy of between 1 keV to 100 keV (for phosphorous) and between 1 keV and 300 keV (for arsenic). This may form a doping profile that may impart gain to the transport of carriers through the base for injection into the floating body. Again, the gain should be sufficient to enable the carrier injection while at the same time being small enough so as to guard against the risk of MOSFET inter-latching.

[0039] In the emitter region, further referencing FIG. 5, silicide 266 may also be disposed with a given outline over the additional region 214. An offset may be defined between the peripheral edge of the silicide 266 relative to a depletion region for the boundary between the base and emitter 204, 214. This offset may be set to establish an effective emitter integral length, and to define in part (per the Gummel concepts) the desired gain for the bipolar device.

[0040] Although the embodiment of FIG. 5 illustrates the silicide structures 264,266 over respective drain1 and drain2 regions; alternative embodiments may comprise just one of the silicide regions.

[0041] In accordance with a further embodiment, lifetime adjustment species may be implanted into at least one of the additional region 214 and/or drain region 204. These lifetime adjustment implants may serve to recombine with a portion of the minority carriers that might otherwise travel across the drain region, so as to affect, at least in part, the desired gain for carrier transport and injection into the floating body region.

[0042] In another embodiment, referencing FIG. 6, a semiconductor memory 300 may incorporate a body-injection, partially-depleted SOI MOSFET 100 (from an embodiment described above relative to FIGS. 1–5) as an access transistor to access a memory element. In this embodiment, the memory element may comprise a capacitively-coupled thyristor 302. Again, the properties—e.g., the doping, thickness, width, density, and/or silicide treatment, and/or lifetime adjustment implant, etc.—of the emitter region relative to the base region of the inherent bipolar device 324 may be designed to establish an carrier transport gain that may assist injection of carriers into the floating body region. The gain may be designed for a magnitude sufficient to facilitate the residual carrier injection into the floating body region of body 105, yet small enough to guard against inter-latching with the MOSFET.

[0043] Although showing one memory cell 100,302 in FIG. 6, it may be understood that semiconductor device 300 may comprise a plurality of memory cells across the substrate. For example, the illustrated memory cell 100,302 may be configured in repeated mirrored relationship about various mirror axis, such as mirror axis 340.

[0044] Herein above, the body-injection MOSFET was described in integrated relationship for a thyristor-based memory, which resulted in the N-P-N-P-N-P structure 202, 105, 204, 314, 316, 318 respectively as illustrated in FIG. 6. In alternative embodiments, the body-injection enhanced MOSFETs may likewise be integrated into alternative circuits such as, e.g., buffer circuits, latching units, processors, routers, switchers, decoders, and others.

[0045] In yet another embodiment, referencing FIGS. 6 and 7, bias circuit 700 may couple a voltage to the floating body region of a partially-depleted SOI MOSFET 400. The bias circuit 700, in particular embodiments, may be controlled by an external controller (not shown), that may adjust the voltage for biasing the floating body according to a pre-selected operation of the transistor. The selected operation may comprise, e.g., an operability for isolating high resistance for a low Ioff, transparent low resistance for high Ion, an intermediate resistance for given leakage and thyristor stability for memory retention, or simply for amplification. In further embodiments, the controller may cause the bias circuit 700 to bias the floating body portion of the MOSFET's body with a voltage level that may be based on a plurality of different input considerations—such as, e.g., the applied voltage levels of the first and/or second wordlines (WL1, WL2), the magnitude of the reference voltage (Vref), and/or the voltage level required for sensing data states on the bitline (BL).

[0046] To further assist an understanding of these particular embodiments, a proposed model for operation may again be described relative to a MOSFET's body region. Referencing FIGS. 6–7, the body region 105 may be influenced by the potentials of neighboring regions—including the voltage (Vgs) of gate electrode 112 and the voltage of the adjacent drain region 204. Electric fields may develop through the body between the gate electrode and the source region and may form an inversion channel in the surface region of the body. In forming the inversion layer, carriers may be pulled from deeper regions of the body and may thereby create a depleted zone that is stripped of carriers beneath the inversion channel layer. The depleted region may be characterized much like an intrinsic insulating region. A remaining portion of the body region, which has not been depleted, may thus be described as the floating body portion beneath the intrinsic or insulating barrier. This floating body may carry a voltage potential relatively isolated from the inversion channel. Further, one may theorize that perhaps the potential of the floating body might not begin to increase relative to the channel region until sufficient depletion for the insulating barrier therebetween has been established.

[0047] In a further embodiment, an electrical interconnect may be formed to contact the body region at a location outside the perimeter or outline of the gate electrode 112. In a particular application, this interconnect to the body region for biasing the floating body may be driven with a voltage related to that of the gate electrode. In another example, an electrical short may be provided between the gate electrode and the interconnect to the body, such that the gate and floating body regions may receive similar applied voltage levels. Accordingly, when the gate voltage is increased, the body voltage might also increase and a drain current may develop more rapidly than what might otherwise be available via the influence of the gate electrode alone.

[0048] In another exemplary application, further referencing FIG. 7, a controller with the bias circuit 700 may drive the floating body with a voltage appropriate for the selected state of the MOSFET. For example, assuming an NMOS device, the voltage for the floating body bias may be driven negative (e.g., −1V) for the off-state condition. This negative bias may act to limit leakage current (Ioff) of the transistor and to assist its isolation performance. Alternatively, when a high transparency or large drive current (Ion) may be desired, the voltage for the floating body may be driven positive (e.g., 1V).

[0049] In further possible aspects, the gate electrode may also comprise a silicide (not shown) over the patterned conductive material. The silicide may be formed of a refractory metal such as, e.g., titanium, cobalt or nickel.

[0050] While certain exemplary features of the embodiments of the invention have been illustrated and described herein, may modifications, substitutions, changes and equivalents may now occur to those skilled in the art. It is, therefore, to be understood that the appended claims are intended to cover all such embodiments and changes as fall within the spirit of the invention.