Electrically-Erasable, Electrically-Programmable Read-Only Memory Cell

The novel features believed characteristic of the invention are set forth in the appended claims. The invention itself, however, as well as other objects and advantages thereof, will be best understood by reference to the following description of particular embodiments thereof, when read in conjunction with the accompanying drawings, in which:

FIG. 1 is a plan view of small part of a semiconductor chip having memory cells according to one embodiment;

FIGS. 2a-2e are elevation views in section of the semiconductor device of FIG. 1, taken along the lines a--a, b--b, c--c, d--d, and e--e of FIG. 1;

FIG. 3 is an electrical schematic diagram of the cells of FIGS. 1 and 2a-2e; and

FIGS. 4a-4d are elevation views in section, corresponding to FIG. 2a, of the device of FIGS. 1 and 2a-2e at successive stages in the manufacture thereof.

Referring now to FIGS. 1, 2a-2e, and 3, an array of electrically-erasable, electrically-programmable memory cells 10 is shown formed in a face of a silicon substrate 11. Only a very small part of the substrate is shown in the Figures, it being understood that these cells are part of an array of a very large number of cells. A number of wordlines/control gates 12 are formed by second-level polycrystalline silicon (polysilicon) strips extending along the face of the substrate 11, and bitlines 13 are formed beneath thick thermal silicon oxide layers 14 in the face. These buried bitlines 13 create the source region 15 and the drain region 16 for each of the cells 10. A floating gate 17 for each cell is formed by a first-level polysilicon layer extending across about half of a cell 10 and across one bitline and extending over onto another adjacent bitline 13. Two "horizontal" or X-direction edges of the floating gate 17 for a cell are aligned with the edges of a wordline/control gates 12. A tunnel area 19 for programming and erasing is formed near the source 15 of each cell 10, and silicon oxide at this window 19 is thinner, about 100 A, compared to the dielectric coating 20 of about 350 A for the remainder of the channel beneath the floating gate 17. Programming and erasing can be performed using a relatively low externally-applied voltage when the structure of the invention is employed, with Fowler-Nordheim tunnelling requiring very little current. The coupling between layer 12 and layer 17, compared to coupling between floating gate 17 and source 15 or substrate 11, is more favorable because the floating gate extends out across the bitlines 13 and isolating area 22. Therefore, a larger fraction of the programming/erasing voltages applied between control gate 12 and source 15 will appear between floating gate 17 and source 15. The cell 10 is referred to as "contact-free" in that no source/drain contact is needed in the vicinity of cell itself.

In contrast to the device of co-pending U.S. patent application Ser. No. 07/494,060 filed herewith, the regions between wordlines are implanted with P-type impurity to create doped isolation regions 21 that isolate cells from one another in the Y-direction. As in the device described in that application, strips 22 of LOCOS thick field oxide separate bitlines 13 between cells in the X-direction. Note that the array of cells is not of the "virtual-ground-circuit" type because there are two bitlines 13 or column lines (one for source, one for drain) for each cikymn (Y-direction) of cells, one bitline being a dedicated ground, and one being the data input/output and sense line.

The EEPROM cells of FIGS. 1, 2a-2e and 3 are programmed with a voltage Vpp applied to the selected wordline 12 of about +16 v to +18 v with respect to the source 15 of the selected cell 10. The source 15 of the selected cell 10 is at ground or other reference voltage. For example, in FIG. 3, if the cell 10a is selected to be programmed, then the wordline labelled WL1 is brought to +Vpp and the source labelled SO is grounded. The voltage +Vpp can be internally generated with charge pumps on the chip, with the externally-applied supply voltage having a relatively small positive potential, perhaps +5 v. The selected drain 16 (labelled D0 in this example) is allowed to float under these programming conditions so there is little or no current through the source-drain path. The Fowler-Nordheim tunneling across the tunnel oxide 19 (with thickness of about 100 A) charges the floating gate 17 of the selected cell 10a, resulting in a shift in threshold voltage of perhaps 3-6 volts after a programming pulse approximately 10 milliseconds in length.

A selected cell 10 is erased by applying a voltage Vee (internally generated) of perhaps -10 v on the selected wordline/control gate 12 and a voltage of about +5 v on the source 15 or bitline 13. The drain 16 (the other bitline 13) is allowed to float. During the erasure tunnelling, electrons flow from the floating gate 17 to the source 15 because the control gate 12 is negative with respect to the source 15.

When a "flash erase" is performed (all cells erased at one time), all of the drains 16 in the array are allowed to float, all of the sources 15 are at potential Vdd, and all of the wordlines/control gates 12 are at potential -Vee.

To prevent a write-disturb condition during the programming example (cell 10a being programmed), all of the sources 15 of non-selected cells, such as cell 10b, on the same wordline WL1 of FIG. 3 are held at a voltage Vb1, which is in the approximate range of 5-7 volts positive. The drains 16 of non-selected cells such as 10b are allowed to float, preventing any source-drain currents from flowing. The voltage Vb1 applied to the sources 15 prevents the electric fields across the tunnel oxides 19 of the cells, including example cell 10b, from becoming large enough to charge the floating gates 17.

Another condition to be avoided is the "bitline-stress", or deprogramming, associated with a high electric field across the tunnel oxide of a programmed cell when the source of the cell is at a potential near Vb1. To prevent this bitline stress condition, the non-selected wordlines/control gates WL0 and WL2 of FIG. 3 are held at a voltage in the approximate range of 5-10 volts positive, thereby reducing the electric field across the tunnel oxide 19 of each non-selected programmed cell. A programmed cell such as 10c has a potential of about -2 to -4 volts on its floating gate, so when the voltage Vb1 on the source S1 of such a cell 10c is in the range of 5-7 volts positive, the field across the tunnel oxide 19 could tend to deprogram the cell, but with voltage in the range of 5-10 volts positive on the wordline WL2, the field is reduced. This voltage on the wordline/control gate WL2 is not so great, however, as to cause a voltage threshold Vt change in a cell having no charge on its floating gate.

The cells described above can be read at low voltage. For example, a row of cells may be read by placing +3 v on the selected wordline/control gate, zero volts on all of the other wordlines/control gates, zero volts on all of the sources, and +1.5 v on all of the drains. In this condition, the source-drain path of a cell will be conductive in an erased or a non-programmed state (a cell with zero charge on its floating gate), i.e., storing a logic one. A programmed cell (programmed to the high-voltage-threshold state, with a negative charge on the floating gate) will not conduct, i.e., storing a logic zero.

A method for making the device of FIGS. 1 and 2a-2e will be described in reference to FIGS. 4a-4d. The starting material is a slice of P-type silicon of which the substrate 11 is only a very small portion. The slice is perhaps 6 inches in diameter, while the portion shown in FIG. 1 is only a few microns wide. A number of process steps would be performed to create transistors peripheral to the array, and these will not be discussed here. For example, the memory device may be of the complementary field-effect type in which N-wells and P-wells are formed in the substrate 11 as part of a prior process to create peripheral transistors. The first step related to the cell array of the invention is applying oxide and silicon nitride coatings 30 and 31 as seen in FIG. 4a, and patterning these coatings using photoresist to leave nitride over what will be the channel regions, the sources and drains, and bitlines 13 while exposing the areas where the thick field oxide 22 is to be formed. A boron implant at about 8.times.10.sup.12 cm .sup.-2 dosage is performed to create a P+ channel stop region beneath the field oxide 22. Then the field oxide 22 is grown to a thickness of about 9000 A by exposing to steam at about 900.degree. C. for several hours. The thermal oxide grows beneath the edges of the nitride 31, creating a "bird's beak" 22a instead of a sharp transition.

Turning now to FIG. 4b, the nitride 31 is removed and, in the area where the bitlines 13 are to be formed, an arsenic implant is performed at a dosage of about 6.times.10.sup.15 cm.sup.-2 at 135 KeV, using photoresist as an implant mask, to create the source/drain regions and bitlines. Next, another thermal oxide 14 is grown on the face to a thickness of about 2500 to 3500 A over the source/drain regions and bitlines 13, during which time a thermal oxide of about 300 A will grow over the channel areas (due to the differential oxidation occurring when heavily-doped and lightly-doped silicon are exposed to oxidation at the same time), to create the oxide layers 14 above the source/drain regions and bitlines 13. This oxidation is in steam at about 800.degree. to 900.degree. C. At the transition areas 18 where the bird's beak 22a has been formed, the edge of the originally-formed thermal oxide has masked the arsenic implant so the concentration is lower and so the oxide growth in that area is less than that of the oxide 14 of the oxide 22.

Referring now to FIG. 4c, a window 19 (also seen in FIG. 1) is opened in the gate oxide 20. This is done using photoresist as mask, and etching through the oxide 20 to the bare silicon, then growing a thin oxide to form the tunnel window 19. During oxidation of tunnel window 19, oxide 20 over the channel region will grow to approximately 350 A.

Referring now to FIG. 2a, first polysilicon layer is now applied to the face of the silicon slice, doped N+, and a coating 34 of oxide, or oxide-nitride-oxide, is applied to separate the two polysilicon levels. The first-level polysilicon is defined using photoresist to leave elongated strips in the Y-direction, parts of which will become the floating gates 17. An oxidation, performed after the first-level polysilicon, and also creates the gate oxide 35 for the series enhancement transistor 36. A second polysilicon layer is deposited, doped N+, and patterned using photoresist to create the wordlines/control gates 12. At the same time as the wordlines/control gates 12 are defined, the edges of the first-level polysilicon are etched, so that the elongated X-direction edges of the floating gates are self-aligned with the edges of the control gates.

A self-aligned ion implant step is performed, using the stacked polysilicon-1 and polysilicon-2 layers of wordlines/control gates 12 and floating gates 17 as a mask, to create the isolating regions 21. For this purpose, boron is implanted at a dosage of about 10.sup.12 cm.sup.-2 at about 70 KeV. After annealing and oxidation, this implant produces the P+ regions 21 very much like channel stop implants beneath field oxide.

The physical dimensions of P-type isolating element 21 may be less than the corresponding physical dimensions of LOCOS isolating element 21 of co-pending U.S. patent application No. 07/494,060 referenced above. Therefore, the memory arrays using the cell 10 of this invention may utilize less area on an integrated circuit chip.

Optionally, the junction profile on the channel side of source 15 may be tailored to make certain that it terminates under the 350 A gate oxide 20, extending over the entire lower surface of window 19 and thereby maximizing the field-plate breakdown voltage of the source junction. Extension 15a or 15b of source 15 extends past the window 19 area and greatly increases the possibility that erasure will be by purely Fowler-Nordheim tunneling and not by hot carriers. For example, extension 15a may be formed to extend source 15 completely under the lower surface of window 19 by implanting a N-type impurity in window 19 prior to or after growing the 100 A coating. An alternative procedure is to include phosphorus as one of the doping materials used to form source 15, then subjecting the slice to a temperature cycle that causes the phosphorus to diffuse laterally under window 19 to form extension 15b.

While the invention has been described with reference to an illustrative embodiment, this description is not meant to be construed in a limiting sense. Various modifications of the illustrative embodiment, as well as other embodiments of the invention, will be apparent to persons skilled in the art upon reference to this description. It is therefore contemplated that the appended claims will cover any such modifications or embodiments that fall within the true scope of the invention.