MX2007009000A - Magnetic memory composition and method of manufacture. - Google Patents

Abstract

A non-volatile magnetic memory device having one or more memory cells, each of the memory cells includes a magnetic switch including a magnetic component and a write coil located proximate the magnetic component, the write coil coupled to receive a current sufficient to create a remnant magnetic polarity in the magnetic component, and a Hall sensor, positioned proximate the magnetic component, to detect the remnant magnetic polarity indicative of a stored data bit.

Description

COMPOSITION OF MAGNETIC MEMORY AND MANUFACTURING METHOD The present invention is a Continued-In-Part application of US Patent Application Serial No. 11 / 189,822 filed on July 27, 2005 and claims the benefit of US Provisional Patent Applications numbers 60 / 647,809 filed on January 31, 2005 , and 60 / 752,035 filed on December 21, 2005, which are incorporated herein by reference.

BACKGROUND OF THE INVENTION Field of the Invention The present invention relates to a memory device, and more particularly, to a memory device using magnetic memory elements.

Description of Related Art The rapid growth in the market for portable consumer products (including products for computing and portable communications) is directed to the need to lower the energy consumption of non-volatile memory devices, with its ability inherent to retain information stored without energy.

The main technology currently available in the market for these applications is the EEPROM technology (Programmable Read Only Memory, which can be erased electrically), which depends on the loading (recording) or unloading (erasing) of the floating gate of a transistor type Oxide-Metal Semiconductor (N-type) using the tunneling called Fowler-Nordheim through the ultra-thin oxide layer of these structures. The loading of the door creates results in an electron inversion channel in the device producing conductivity (constituting a state of memory 1). By unloading the floating gate (that is, by applying a negative polarization) it separates the electron from the channel and returns the device to its initial non-conducting state (that is, memory status 0). A serious limitation of this technology refers to the tunneling that limits the duration of the erase / etch cycle and can induce a catastrophic failure (after a maximum of approximately 106 cycles). Moreover, the required charging time - which is of the order of 1 ms is relatively long.

To improve performance, the technology called FeRAM (Ferroelectric Random Access Memory) has been developed. The FeRAM memory cell consists of a bi-stable capacitor, and is comprised of a thin ferroelectric film containing polarizable electric dipoles. These dipoles, analogous to the magnetic moments in a ferromagnetic material, respond to an applied electric field to create a net polarization in the direction of the applied field. A hysteresis loop to sweep the applied field from positive to negative field defines the characteristics of the material. By removing the applied field, the ferroelectric material can retain a polarization known as the remaining polarization, which serves as the basis for storing information in a non-volatile form. The FeRAM seems to be a promising technology with good future potential because relatively low voltages are required (practically around 5V) to connect compared to approximately 12 to 15V for the EEPROM. Moreover, the FeRAM devices show recording resistance cycles of 108 to 1010 compared to about 106 for EEPROM, and the electric polarization ignition needs only about 10 ns compared to about 1 ms to charge an EEPROM. However, the need for an additional cycle to return to a bit determined to its original state for reading purposes aggravates the problems of dielectric fatigue. This, in turn, is characterized by degradation in the ability to polarize the material. In addition, due to the behavior of these materials around their Curie temperature, as well as the stability of the composition (and changes associated with the Curie temperature), even moderate thermal cycles promote accelerated fatigue. Finally, the uniformity and control of the manufacturing process is still a challenge.

Now, MRAM (Magnetoresistance Random Access Memory) - whose development began about 20 years ago - seems to be the biggest promise for existing technologies in terms of read / write and speed cycles. The technology depends on a recording process using the hysteresis loop of a ferromagnetic strip, while the reading process involves the effect of anisotropic magnetoresistance. Basically, this effect (based on the interaction of the spin orbit) refers to the variation of the resistance of a magnetic conductor, dependent on an externally applied magnetic field. The bit consists of a strip of two ferromagnetic films (for example, NiFe) interspersed with a poor conductor (eg, TaN), placed below an orthogonal conductive strip line (ie, known as the word line). To write, a current passes through the interleaved strip and when it is aided by a current in the orthogonal strip-line, the highest ferromagnetic layer of the interleaved strip is magnetized either in one direction or the opposite. The reading is made by measuring the magnet-resistance of the interleaved structure (ie, passing a current). The magnet-resistance ratio of only about 0.5% is typical, but they have allowed the fabrication of a 16Kb MRAM chip that operates with 100 ns recording times (and 250 ns reading times). A 250Kb chip was later produced by Honeywell.

The discovery of the so-called Giant Resistance -Iman (GMR) in 1989, implemented by sandwiching a copper layer with a magnetic thin film allowed further improvement in the performance of the memory device. The GMR structures showed a magnet-resistance of about 6%, but the interchange between the magnetic layers limits how fast the magnetization can change direction. Moreover, the magnetic undulation from the edge of the strip imposes a limitation in the reduction of the size of the cell, or incrustation.

Promising results were obtained with the so-called Pseudo-Spin Valve (PSV) cell made of a structure interspersed with two unadjusted magnetic layers, so one layer tends to connect the magnetization to a lower field than the other. The soft film is used to detect (by the resistance-magnet effect) the magnetization of the hard film - this last film constitutes the storage medium, having magnetized up or down (ie, states 0 or 1). PSV structures are receptive for embedding but the reported fields needed to connect the magnetic hard layer are still too high for high density integrated circuits. These devices seem to potentially represent a replacement for EEPROM.

Further improvements in magnet-resistance (ie up to 40%) are obtained with the spin-dependent tunneling devices (STD). These devices are made of an insulating layer (ie, the tunneling barrier) sandwiched between two magnetic layers. The operation of the device depends on the fact that the tunneling resistance, in the direction perpendicular to the stack, depends on the magnetization of the magnetic layers. The highest resistance is obtained when the magnetization of the layers is anti-parallel, and the parallel case provides the lowest resistance. The variation of the density of the state of rotation (ie, up or down) between the two magnetic layers explains this behavior. One of the layers is fixed with pins while the second magnetic layer is free and is used as the information storage means. The SDT promises for high performance non-volatile applications. There have actually been some reported values for recording times as small as 14 ns with this proposal. However, controlling the uniformity of resistance (ie the thickness and quality of the tunneling barrier), and therefore controlling the behavior of the bit-by-bit switch remains a real challenge that has yet to be overcome when implemented. . What is needed is a non-volatile memory device that is fast, reliable, relatively simple in design, economical and robust.

Compendium of the Invention Accordingly, the present invention is directed to a magnetic memory device that substantially avoids one or more of the problems due to limitations and disadvantages of the related art.

An object of the present invention is to provide a magnetic switch used to form a memory cell in a non-volatile magnetic memory device.

Another object of the present invention is to provide a simplified method for manufacturing a non-volatile magnetic memory device.

Additional features and advantages of the invention will be set forth in the description that follows and, in part, will be apparent from the description, or may be learned by practice of the invention. The objects and other advantages of the invention will be understood and attained by the structure pointed out particularly in the written description and the claims hereof, as well as by the accompanying drawings.

To achieve these and other advantages in accordance with the purpose of the present invention, as it is incorporated and widely described, a non-volatile magnetic memory device having one or more memory cells, each of the memory cells includes a magnetic switch which includes a magnetic component and a recording coil located next to the magnetic component, the recording coil is coupled to receive a sufficient current to create a magnetic polarity remaining in the magnetic component, and a Hall sensor, placed close to the magnetic component, to detect the remaining magnetic polarity indicative of a bit of stored data.

In another aspect of the invention, a method for manufacturing one or more memory cells of a non-volatile magnetic memory device, the steps include forming a magnetic switch including a magnetic component and a recording coil located next to the magnetic component, the coil of recording coupled to receive a sufficient current to create a magnetic polarity remaining in the magnetic component, and to form a Hall sensor, placed proximate the magnetic component, to detect the remaining magnetic polarity indicative of a stored data bit.

In still another aspect of the invention, a method of manufacturing a magnetic switch of a memory cell in a non-volatile magnetic memory device, the steps include forming a recording coil, forming a coaxial magnet point with the coil of recording, and electroplating a magnetic material at the magnet point to form a magnetic component located near the recording coil, the recording coil is coupled to receive sufficient current to create a magnetic polarity remaining in the magnetic component.

It is understood that both the general description and the following detailed description are exemplary and explanatory and attempt to provide further explanation of the invention as claimed.

Brief Description of the Drawings The accompanying drawings, which are included to provide a greater understanding of the invention and are incorporated into and constitute a part of this specification, illustrate the embodiments of the invention and together with the description serve to explain the principles of the invention. In the drawings: Fig. 1 shows a plan view of an exemplary embodiment of a memory cell according to the present invention; Fig. 2A shows a top view of an exemplary embodiment of a magnetic switch according to the present invention; Figs. 2B-2C show a side view of the exemplary embodiment of the magnetic switch shown in Fig. 2A; Y Figs. 3A-3B show conceptual views of an exemplary embodiment of a tuneable magnetic switch in accordance with the present invention.

Fig. 4 shows a graph illustrating the hysteresis loop for determining the reverse magnetization of the magnetic switch of the present invention.

Figs. 5A-5H shows several exemplary manufacturing steps for an exemplary sensor in accordance with the present invention.

Fig. 6 shows a scanning electron microscope (SEM) image of an exemplary sensor manufactured in accordance with the present invention.

Figs. 7A-7D show several exemplary manufacturing steps for isolating an exemplary sensor in accordance with the present invention.

Fig. 8 shows an exemplary embodiment of an electroplasty system in accordance with the present invention.

Figs. 9A-9D show several exemplary steps of a manufacturing process (i.e., carburizing) for an exemplary coil and magnet point according to the present invention.

Fig. 9E shows an SEM image of a sensor manufactured in accordance with the manufacturing process of the present invention.

Figs. 10A-10D shows several exemplary manufacturing steps for depositing a magnetic material at a magnet point according to the present invention.

Fig. 11 shows an SEM image of a magnetic switch manufactured in accordance with the present invention.

Figs. 12A-12E show several exemplary stages of an alternative manufacturing process (ie, attack direct chemical) for an exemplary coil and magnet point according to the present invention Fig. 12F shows an SEM image of an exemplary sensor manufactured in accordance with the alternative manufacturing process of the present invention.

Detailed Description of the Preferential Modalities Reference will now be made in detail to the preferred embodiments of the present invention, examples of which are shown in the accompanying drawings.

The present invention is directed to a magnetic memory device. In particular, Fig. 1 shows an exemplary embodiment of a memory cell of a magnetic memory device according to the present invention. The memory cell 10 according to an exemplary embodiment of the present invention includes a magnetic switch 120 and a sensor 130.

The magnetic switch 120 includes a component or magnetic material 122 and a coil 124 for storing the data. The sensor 130 includes an Effect sensor Hall 132 and output terminals 136 connected to a voltage detector (not shown) for detect the data stored in a magnetic switch 120.

In particular, the magnetic switch 120 includes a magnetic component 122. The magnetic component 122 can be a permanent magnet or a ferromagnetic material (for example, nickel or nickel-iron magnet). A coaxial coil 124 (connected to a current source, not shown) is positioned around the magnetic component 122. The coaxial coil 124 is made of a conductive material, such as Ti / Au metal. However, any other suitable conductive material (e.g., Ti / Cu / Ti) can be used without departing from the scope of the present invention. Although it is shown that the magnetic component 122 has a general shape. Cylindrical for purposes of illustration, any suitable shape (e.g., square, rectangular, horseshoe) can be used without departing from the scope of the present invention. Furthermore, for illustration purposes it is shown that coaxial coil 124 has six (6) turns around magnetic component 122. However, any suitable number of turns can be used without leaving the field of the present invention.

The Hall Effect sensor 132 includes a geometrically defined semiconductor structure with terminals of input 134 connected to an energy source 138 and output terminals 136 placed perpendicular to the direction of current flow. Although it is shown that the Hall Effect sensor 132 has a "Greek cross" shape for purposes of illustration, any suitable shape (eg, rectangular) can be used without departing from the scope of the present invention.

In general, the Hall Effect sensor responds to a physical amount to be detected (ie, magnetic induction) through an input interface and, in turn, outputs the detected signal to an output interface that converts the electrical signal from the Hall Effect sensor to a designated indicator. In the present case, when the Hall effect sensor 132 is subjected to a magnetic field (H) from a magnetic component 122, a potential difference appears across the output terminals 136 in proportion to the output terminals 136 in proportion to the resistance of the field. When the Hall Effect sensor 132 is subjected to an equal and opposite magnetic field, an equal and opposite potential difference appears through the same output terminals 136. The Hall Effect sensor 132 thus acts as a sensor of the magnitude and direction of an externally applied magnetic field.

In general, the shape and material used for the magnetic switch 120 determines the magnetization resistance (M) responsible for generating a magnetic field (H) around the sensor 130. The number of turns of the coil 124 around the magnetic component 122, together with the current applied to the coil 124, it determines the resistance of the induced magnetization (H) generated around the magnetic component 122 to establish the direction and intensity of the magnetization (M). The direction of the magnetization (M) of the magnetic component 122 determines the value of the stored magnetic data (i.e., "0" or "1" in the magnetic switch 120. The Hall Effect sensor 132 is characterized by the voltage signal VHa ?? which is generated in response to the magnetic field (H) emanating from the magnetic switch 120 detected at point P.

A current (I) (for example, current pulse) is sent through coil 124 in such a way as to generate a magnetic field H-coil. The magnitude of the current is chosen to be sufficient to change (i.e., back off) the magnetization of the magnetic component 122. The magnetic field generated by the magnetic component 122 needs to be sufficient for the sensor 130 to detect it at the detection point P. After detecting it, the sensor 130 needs to generate a response (VHa ??) greater than a balanced voltage signal Veq. A balanced voltage Veq is the threshold that must be overcome before generating any useful signal. More specifically, the magnetic field (H) generated by the magnetization (M) of the magnetic switch 120 must be strong enough at the point P to generate an induced voltage at the sensor 130 greater than Vdesp before the stored data can be detected reliably A magnetic field that generates a voltage signal smaller than the balanced voltage can not be detected by the sensor 130 under the present conditions of DC biasing.

Fig. 2A shows a top view of an exemplary embodiment of a magnetic component surrounded by a coil. For purposes of illustration only, Fig. 2B shows a side view of a magnetic component 222 having an initial magnetization direction (M) oriented downward. Fig. 2C shows that after sending a sufficiently high current (I) through the coil 224, the magnetic component 222 retains an induced magnetization whose direction is oriented upwards. In this case, the magnetic induction near the surface of the magnetic component 222, at the detection point P, is the field generated by the magnetic component 222. This field causes the sensor 130 to generate a voltage signal that should have a magnitude greater than the voltage signal VdeSp and a sign indicating the magnetization direction (for example, a positive voltage towards (up "). If an upward magnetization is designated as" 1"," then the sensor 130 detects the stored data as "1 Then to achieve a downward magnetization (ie, "0"), a suitable current (eg, pulse of current in the opposite direction) is sent again through coil 224 to generate a magnetic field -Hobina (is say, with the opposite orientation to Hbobina) sufficient to change (ie back) the magnetization of the magnetic component 222. After the pulse, the magnetic component 222 retains a magnetization that may have a smaller magnitude or whose direction is oriented downward . In this case, the magnetic field at the detection point P is the magnetic field generated by the magnetic component 222. The induction detected at the point P causes the sensor 130 to generate a voltage signal having a smaller magnitude or opposite sign indicating the direction of the magnetization (for example, a negative voltage for "down"). If a down or smaller magnetization is designated "0," then the sensor 130 detects the stored data as "0." In another embodiment of the invention, a tunable magnetic switch in accordance with the present invention ensures operational reliability of the manufactured magnetic memory device. In particular, the balanced voltage threshold Veq described above may be larger than expected. The displacement of the sensor is caused by things such as non-uniformity of the device and misalignment that occurs during manufacturing. The magnetic induction (B) generated by the magnetization (M) of the magnetic switch 120 must be strong enough at the point P to generate an induced voltage at the sensor 120 before the stored data is reliably detected. Once the memory device containing an array of cells 10 is manufactured, the internal components can not be rearranged to reduce the balanced operating threshold Veq. To solve this problem, a tunable magnetic switch in accordance with the present invention ensures the operational reliability of the manufactured magnetic memory device enabling that the detected magnetic field is tuned after the manufacturing process, as described below.

Figs. 3A and 3B show an exemplary embodiment of a tuneable magnetic switch in accordance with the present invention. For purposes of illustration, Fig. 3A shows a tuneable magnetic switch 320 that includes two magnetic components 322 and 330. The magnetic component 322 is coupled to a three (3) turn coil. However, any suitable suitable number of turns can be used without leaving the field of the present invention. The magnetic component 322 may be a soft cylindrical bar magnet made of a ferromagnetic material (eg, nickel-iron magnet). The magnetic component 330 may be a hard permanent magnet made of ferromagnetic material (eg, nickel, cobalt, and other related alloy magnets). Although the magnetic components 322 and 330 are shown to have a particular shape for purposes of illustration, any suitable form can be used without departing from the scope of the present invention.

As shown in Fig. 3B (ie, side view), magnetic switch 320 is exposed to external magnetic polarization field Hpo? Ar? provided by the magnetic component 330. Once a Hpo? ari polarization field is established on the magnetic switch, a current (I) (for example, a current pulse) is sent through the coil in such a way as to generate a magnetic field (H) having the same direction and orientation of the Hpo? Ari polarization field. The magnitude of the pulse of the current is chosen to be sufficient to direct the magnetic component 322 to its saturation magnetization value.

For purposes of illustration only, the magnetization direction (M) of the magnetic component 322 is shown as initially oriented downward, in the same direction of the constant polarization field Hpo? Ari. After the current (I) is sent through the coil 324, the magnetic component 322 retains a high magnetization. In this case, the magnetic field near the surface of the magnetic component 322, at the detection point P, is the combination of the polarization field Hpo? Ari and the field generated by the magnetic component 322. This combined field results in a very high magnetization state, generating a voltage signal much greater than the balanced voltage Veq. Therefore, the sensor 130 easily detects the stored data as "1," for example, assuming that the The downward direction of the magnetization (M) is designated as a high state (ie, "1").

To achieve a lower state (i.e., "0"), an adequate current (I) (i.e., a current pulse) is sent through coil 324 to generate a magnetic field -Hobin in the opposite direction to the field of polarization Hpo? ar? enough to generate a total magnetic field (ie, Hbob? Na + Hpo? Ar?) Which demagnetizes the magnetic component 322. After sending the current through the coil 324, the magnetization (M) will reverse following the line of recoil, explained later with reference to Fig. 4, providing a magnetic component 322 with very low magnetization. If the current is strong enough, the magnetization (M) can still be oriented in the opposite direction. In this case, the magnetic field at the detection point P will be that of the polarization field Hpo? Ari combined with the magnetic field generated by the magnetic component 322, which can be very low or in the opposite direction of the polarization field Hpo ? ari • In any circumstance, the total magnetic induction at point P will be considerably lower than that corresponding to the case of high level, non-existent, or even in the direction contrary. Accordingly, a definite low level state (ie "0") can be detected by the sensor 130.

The behavior of the switch shown schematically in Figs. 3A and 3B can be explained using the hysteresis loops of the magnetic component 322 as shown in Fig. 4. First, the intersection of the induction load line of the induction hysteresis loop defines a point "a" with Bi induction. . The point "a" can be used to determine the corresponding "b" point on the magnetization loop. The magnetization charge line can then be removed. This load line is then transferred via Hbob along the axis of the magnetic field to establish a new intersection at point "e" on the magnetization hysteresis loop. The corresponding point "f" on the induction loop can then be established. After removing Hb0bma (ie, the current pulse is removed), the magnetic component 322 will retract. Using the "f" point and recoil permeability, the line of recoil can be extracted. Finally, the intersection point "g" of the return line and the magnetization charge line can be determined, providing the B2 induction. The induction B2 is then established as the induced magnetization (M) which is stored in the magnetic component 322 once the current (I) is removed to establish the low state (ie, "0").

The manufacturing process will now be established with reference to Figs. 5-10. The manufacturing process of the memory cell 10 (as shown in Fig. 1) can be divided into two parts: (1) the manufacture of the sensor 130, and (2) the manufacture of the magnetic switch 120. For the switch tuneable magnet, an additional process is included to make the polarization magnet.

The Hall Effect sensor 132 is manufactured with highly mobile materials, such as III-V materials (ie, compounds formed from the elements of groups III and V of the periodic table). Materials III-IV include, but are not limited to, GaAs, InAs, InSb, and the two-dimensional electron gas (2DEG) related structures. A 2DEG structure based on a GaAS / AlGaAs hetero structure can be formed at the hetero linkage interface of a doped hetero-structure by modulation between an AlGaAs material of wide doped broadband (ie, barrier) and GaAs material of narrow band width not contaminated (ie, well). Ionized carriers (from the dopant) they are transferred to the well, forming the 2DEG. These carriers are separated from their ionized precursor impurities, and therefore, allow high mobility of the carrier and a large Hall Effect.

Although only materials III-IV are described herein, other materials (e.g., silicone) may be used to fabricate the Hall Effect sensor 132.

Figs. 5A-5D show the different steps of manufacturing the Hall Effect sensor 132 according to an exemplary embodiment of the present invention. A suitable insert is used, such as a semi-insulated GaAs insert with an active n-type thin film of GaAs 539 (approximately 0.5-0.6 μm). A protective layer 540 (for example, 4% of 950K PMMA) is centrifuged on the insert 538. The following centrifugation conditions can be used: spin speed = approximately 4000 rpm (thickness = 0.5-2 μm); Cooking temperature = 160 ° C; Cooked time = 7 minutes; exposure energy = 25kV; exposure dose = 150 μC / cm2; Developer = MBIK / IPA mix (1: 3), development time = 25 seconds. The protection layer 540 has the pattern through EBL (ie, electron beam lithography); however, any suitable pattern technique may be used (eg, photolithography with protection standard type AZ). A table etching process is then carried out to isolate the sensor. The chemical attack process involves wet chemical attack with, for example, a standard solution of H202 / H3P04 / H20.

After the etching process, the input terminals 134 and the output terminals 136 (Fig. 1) are deposited through a cementing process. As shown in Figs. 5E-5H, the cementing process involves spinning one every 542 made of a double layer / PMMA copolymer (at 4000 rpm). The cementing profile (ie, under chemical attack) provided by the difference in sensitivity between the copolymer and PMMA during the development process and after exposure to an electron beam. A contact layer 544 of suitable material, such as gold-germanium (AuGe), is evaporated on the insert 538 to a thickness of approximately 400 nm to form ohmic contacts 134 and 136 for use at the input and output terminals of the sensor 130. You can add a layer of nickel to the AuGe 544 layer to improve contact performance.

After the evaporation step, the cementing process is completed by placing the insert 538 in acetone to eliminate any unnecessary part of the AuGe 544 layer. After cleaning them properly, the contacts (ie the AuGe 544 layer) are subjected to thermal firing fast (RTA). The cooking is carried out at about 340 ° C for about 40 seconds in an RTA chamber filled with nitrogen flow (N2). The cementing process is completed by placing the 538 insert in acetone to remove any unnecessary part of the AuGe 544 layer. Fig. 6 shows the GaAs Greek Cross Hall effect sensor with the AuGe contacts. It also shows the alignment marks 546 included in the pattern.

Although the 4% PMMA protection is used in the previous example, any suitable protection, such as 2% PMMA can be used. Moreover, HMDS, an adhesion promoter, can be used as necessary. When 2% PMMA is used as protection, the following parameters of the lithography process can be used: PMMA (2%); Exposure energy = 15kV; exposure dose = 150 μC / cm2; developer = MBIK / IPA mixture (1: 3); Development time = 25 seconds.

Once the Hall Effect sensor 132 is manufactured, an insulating layer 748 is spun on the Hall Effect sensor 532. The insulating layer 748 is made of a suitable material, such as a dielectric polyimide, which can be processed as a typical protection (ie, centrifuged on a plate and cooked in an oven or on a hot plate). An example of a dielectric polyimide is the HP PI2545 Microsystem (a high-temperature, inter-metallic polyimide used in several microelectronic applications). It has a high glass transition temperature (ie, approximately 400 ° C) and can have the pattern with positive protection. Moreover, the cured film is ductile and flexible with a low CTE, and is resistant to chemicals for common dry and wet processes. Other suitable materials include silicone oxide and silicon nitride, which can be deposited through the Plasma Enhanced Chemical Vapor Deposition (PECVD) at low temperatures.

For illustrative purposes only, Figs. 7A-7D show an insulation layer 748 of PI2545 spun on the Hall 532 effect sensor at a rate of approximately 6000 rpm and then cooked smoothly on a hot plate. The temperature rises from 25 ° C to 170 ° C to 240 ° C / h. Once the temperature of 170 ° C is reached in the oven or hot plate, the temperature is kept constant for 9 minutes (ie soaking period). After the soaking period, the hot plate is cooled to room temperature by means of natural convection. When the insulating layer 784 is baked in an oven or hot plate at a temperature of about 140 ° C or 170 ° C, it develops a good chemical resistance to boil acetone, which is subsequently used to remove a protective layer.

Once the insulating layer 748 is deposited, a positive protection layer 750 (e.g., 4% PMMA or AZ5206) is spun onto the insulating layer 748. For purposes of explanation, 4% PMMA is used. The protective layer 750 is baked in a hot oven or dish at a temperature of 160 ° C for two (2) minutes, with a ramp rate of 6 ° C / minute and a soaking period of 6 minutes. A cooking temperature of 160 ° C is the minimum safe cooking temperature for PMMA (for example, PMMA cooked at 120 ° C may exhibit some adhesion failures).

Then, the insert is placed in an EBL camera, where it is exposed to a 25kV electro beam. The protective layer 750 has the pattern in such a way that it makes openings on the ohmic contacts of the Hall Effect sensor and alignment marks (if any). For a design of size 9 X 10 μm 2, a suitable dose may be in the range of 165 - 182 μC / cm 2; for a design of size 17 X 17 μm 2, a suitable dose may be in the range of 149-163 μC / cm 2; and for a design of size 100 X 112 μm2; a suitable dose may be in the range of 132-143 μC / cm2.

After exposure, the protective layer 750 is developed in a suitable solution, such as MIBK / alcohol (1: 3), for an appropriate amount of time (eg, about 40-55 seconds). The insert is then rinsed in alcohol and de-ionized water. Once the insert is clean, a diluted solution of PPD450 (1: 5) is used to chemically attack the insulating layer for an adequate amount of time (eg, approximately 6-14 minutes or even more). The degree of dilution and agitation and the times of development and chemical attack can be changed as necessary. Boiling acetone is used to remove the protective layer 750 (ie, PMMA). Finally, to complete the manufacture of the insulating layer 748, the insulating layer 748 is hard-baked at about 200 ° C using a temperature rise as described above. The insulating layer can be baked hard at a temperature as high as 400 ° C. However, that high temperature can create unwanted diffusion in the Hall Effect sensor.

Once the sensor 130 is manufactured, the magnetic switch 120 is made on the insulating layer 748. The general proposal for manufacturing the magnetic switch 120 is to first manufacture the coil 124, and then manufacture the magnetic component 122. Traditional methods for manufacturing magnetic materials (eg, Alnico and Martensitic steel) involve synthetic routes that include, for example, melting different components, straining and thermally processing high temperature (typically, above 800 ° C) (for example, tempered).

Other synthetic routes include sintering and extrusion. These methods are incompatible with micro-technology or plate-scale processing due to the extremely small sizes of the components.

Electroplasty, on the other hand, allows the relatively good definition of the shapes of the elements with fewer defects on the walls of the element. It is also an economical and relatively simple method to implement. Three-electrode systems can be used to monitor the stoichiometry of the deposited alloys.

The electroplasty will be used to explain the manufacturing process of the magnetic switch 120; however, any suitable synthetic route can be used. As shown in Fig. 8, an electroplasty system 800 includes an electroplasty cell 810, a computer 820, and a computer-controlled potentiostat / galvanostat 830. The computer 820 is connected to the electroplate cell 810 to through potenciostat / galvanostat 820 to control the electroplasty process. The potenciostat / galvanostat 830 can function as a potentiostat or as a galvanostat.

First, the coil and a magnet point or area within the coil where the magnetic component is to be deposited are formed on the sensor 130. A first exemplary process for forming the coil and the magnet point involves a titanium cementation process /gold. Figs. 9A-9D show several stages of manufacture according to the Gold cementation process according to the present invention.

The insulating layer 748 (of Fig. 7D) is first covered with a double protection layer 954 (e.g., copolymer / PMMA). For this, a layer of the Eli copolymer is first centrifuged on the insert. After, the copolymer layer is baked at 160 ° C for 5 minutes on a hot plate with a temperature rise as described above. The hot plate is allowed to cool to room temperature by means of natural convection. Then, a layer of 4% PMMA in anisole is centrifuged on the plate and baked at 160 ° C for 5 minutes using the defined temperature rise. The hot plate is allowed to cool to room temperature again by means of natural convection.

The insert is placed towards the EBL chamber, where the double protective layer 954 is exposed to an electron beam to design the coil 924 and the magnet point 923, with an exposure of 25 kV and several doses: for a fine design of coil, an appropriate dose is 150 μC / cm2; for the magnet point, an appropriate dose is 120 μC / cm2; for alignment marks (if any), an appropriate dose is 195 μC / cm2. The brands of Alignment can be included in the design to assist in the location of the magnet point. The double protective layer 954 is then revealed in a suitable solution, such as MIBK / alcohol, for approximately twenty (20) seconds.

After the design step, the insert is placed in an electron beam evaporator, where the titanium layer 952a and the gold layer 952b of 25 nm and 150 nm, respectively, are deposited on the designs to form the Ti / layer Au 952. The titanium layer 952 is used as an adhesion layer. Finally, the insert is separated from the evaporator and immersed in acetone for about one hour to remove the double protective layer 954 and any unwanted Ti / Au 952 layer. As shown in Fig. 9F, coil 924 and magnet point 923 are obtained. In this exemplary embodiment, only a single-turn coil 924 is used. However, different numbers of turns may be used as appropriate. without leaving the field of the invention.

After depositing the coil 924, the magnet point 923, and the alignment marks (not shown), the magnetic component 1229 is subjected to electroplating on the magnet point 923 through a mold that it provides the shape and dimensions of the magnetic component 122. As shown in Figs. 10A-10C, to make that mold, EBL is used to design a thick layer 1058 (for example, about 10 μm) of protection (e.g., AZ4620) on coil 924, magnet point 923, and alignment marks (not shown) The protective layer 1054 is fired at about 95 ° C for about 4 minutes. Then, the protective layer 1058 is placed in the chamber for EBL, where the areas where the alignment marks are located are exposed to an electron beam. After this exposure, the protective layer 1058 is revealed in a suitable solution, such as PPD450, and is separated from the areas where the alignment marks are located. The plate is cleaned with de-ionized water and blow-dried with N2. Then, using EBL (and the alignment marks as a guide), the magnet point 923 is designed and the protection layer is revealed for one second to obtain a 1060 well. Well 1060 functions as a vessel in which a material magnetic is subjected to electroplasty to form the magnetic component.

The insert with the protection template is then placed in the electroplasty cell 810 (Fig. 8), where the pulsed deposition (with, for example, a 2% strong cycle, where ton = 1 ms, t0ff = 49 ms, and the maximum current is about 1.4 mA) is used to deposit the magnetic material (eg, nickel or nickel iron) on the protection template forming the well on the magnetic point to, by this means, form an array of magnetic components 122. Pure materials in general are easier to deposit. However, alloys can also be used. Examples of materials that can be deposited include cobalt, iron, nickel, nickel-iron (NiFe), and cobalt-nickel-iron (CoNiFe). Different catalysts can be used to increase the coercivity of these materials if necessary.

For illustrative purposes, a solution based on nickel chloride with two additives, called saccharin (which acts as a stress releasing agent) and sodium lauryl sulfate (which acts as a surfactant), is deposited in well 1060 A current, such as a DC current, is used to make the magnet component. For an even smaller one, the larger relationship structure between dimensions, pulsed electro-deposition can be used (with, for example, an occupation cycle of 2%) can be used to deposit the magnetic material (e.g., nickel or nickel-iron) on the shield template to form a magnetic component array 122. The electroplating conditions are controlled by the computer-controlled potenciostat / galvanostat 830. Although the shape of the magnet is cylindrical , any shape can be developed (e.g., rectangle, frame) using the prior art. After electro-deposition, the mold (i.e., thick protective layer 1058) is separated using a suitable solution, such as acetone. Fig. 11 shows a magnetic switch developed using the above process.

Once the magnetic switch 120 has been completed, the other processing steps for making the tuneable magnetic switch as shown in Figs. 3A and 3B. For example, an insulating layer 748 is deposited on the upper part of the magnetic switch 120. Next, a hard permanent magnet, for example, is added to the upper part of the structure by means of hybrid integration of the prefabricated micro-magnets or by electroplating medium of hard ferromagnetic material, such as cobalt or selected alloys, on the insulating layer 748.

Although EBL is used as the exemplary method for manufacturing the mold, any suitable method, such as photolithography, can be used. For example, when photolithography is used, the mold is formed by exposing the protective layer (ie, AZ4620) to UV light through a prefabricated hard mask.

Another proposal to manufacture coil 924 and magnet point 923 involves chemical attack directly to seed layer 952 to obtain coil 924 and magnet point 923 in the same process step as shown in Figs. 12A-12E. An important concept is to use the seed layer 925 for the growth of the magnetic component 122 and, at the same time, to manufacture the coil 924. First, the carrier plate of the seed layer 952 (i.e., the Ti layer 952A, the Cu layer). 952b, layer Ti 952c) are designed through, for example, EBL. This design step can incorporate the use of a positive protection layer 1210 and a wet chemical attack. Again, the design includes a single-loop coil around a central metal point, with a metal path joining it electrically to a common electrode used for electroplasty. However, any suitable number of turns can be used. The plate is dried by means of cooking on a hot plate for approximately 30 minutes at approximately 150 ° C. A protective layer 1210, for example, AZ5206E) is centrifuged on the insert. The resistance layer 1210 is baked smooth, starting at about 95 ° C and then lowered to about 80 ° C, the change in temperature time being from about six (6) to seven (7) minutes. The protection layer 110 is then exposed (for example, exposure energy = approximately 10 kV, dose = approximately 6μC / cm2). After the exposure, the insert is revealed in a suitable solution, such as PPD450. The plate is then cleaned with de-ionized water. After the cleaning step, the plate is baked hard for about 10 minutes at about 125 ° C. The layers of titanium (Ti) and copper (Cu) are chemically etched with suitable solutions. For example, the Ti 952a and 952c layers can be chemically etched with a highly diluted solution of HF / HN03 / H20, while the copper layer 952b can be chemically etched with a solution of HC1 / H202 / H20. The insert is then cleaned to separate the shield 1210. The cleaning step may include, for example, boiling acetone, boiling alcohol, and rinsing with de-ionized water. Once coil 924 and 923 magnet point have been chemically attacked directly in the seed layer 952, the insert is subjected to the process to create the mold for the electroplating of the magnetic component as described above.

The magnetic memory device according to the present invention was described in relation to a magnetic switch on a Hall Effect sensor. In particular, the advantages of a magnetic component that can retain a magnetic field without any energy supplied to it and a simple sensor to read the stored magnetic field provide a non-volatile memory device that consumes very little energy for its operation compared to the devices of memory based on electricity currently in use.

Additionally, the tuneable magnetic switch according to the present invention was described. The advantages of the tuneable magnetic switch according to the present invention are numerous. First, because the magnetic component retains the induced magnetization (M) from the induction coil, the tuneable magnetic switch according to the present invention can function as a switch with non-volatile memory.

Second, the tuneable magnetic switch in accordance with the present invention provides a field high enough for the partial Hall Effect sensor or even fully compensates for the displacement of the sensor. In the case of the previous one, the tunability of the magnetic switch according to the present invention, that is, the polarization field can be adjusted relatively to the displacement of the sensor, allowing a greater tolerance to manufacturing constraints, making fabrication easier, and increasing the reliability of the devices. This is a considerable asset for miniaturization since the displacement of the sensor increases as the size of the devices is scaled downwards.

Another significant advantage of this proposal is that the tunable magnetic switch according to the present invention allows the use of magnets with low aspect ratio, which are easier to manufacture, because the polarization field compensates for the demagnetization of the magnetic component of the memory cell. The tuneable magnetic switch according to the present invention was described in relation to a magnetic memory device using Hall effect sensors. However, the switch magnetic tuneable according to the present invention can be applied with other magnetic memory devices since the bias magnetic field for tuning the magnetic switch can be applied to any magnetic component and sensor configuration.

The magnetic memory device according to the present invention has several applications including, but not limited to, radio frequency identification (RFID) tags, personal digital assistants (PDA), cell phones, and other computing devices. For example, the magnetic memory device according to the present invention can be used for aerospace / defense, sensors, and RFID applications. The magnetic random access memory of the present invention has been developed for difficult applications of low density radiation. The magnetic random access memory of the present invention is non-volatile, read / write directed, and made of materials of intense radiation. Relevant and emerging markets include aerospace and defense applications, such as rad-hard military and radar systems, satellite and security, sensors and RFID. The sensors in automotive applications, medium equipment such as bioelectronics, biosensors, and gas / liquid / energy meters, and seismic monitoring for oil and gas exploration, for example, are all planned as potential uses of the present invention. Future growth and technical evolution are anticipated in computing, PDA, and the display markets as well.

It will be apparent to those skilled in the art that various modifications and variations can be made to the tuneable magnetic switch of the present invention without departing from the spirit and scope of the invention. Thus, it is intended that the present invention cover the modifications and variations of this invention by including them within the claims and other equivalents.

Claims (44)

1. A non-volatile magnetic memory device having one or more memory cells consisting of: a magnetic switch with a magnetic component and a recording coil located close to the magnetic component, the recording coil coupled to receive a sufficient current to create a magnetic polarity remaining in the magnetic component, and a Hall sensor, placed next to the magnetic component, to detect the remaining magnetic polarity indicative of a stored data bit.

2. The device of claim 1, wherein the magnetic component is one of a permanent and ferromagnetic material.

3. The device of claim 2, wherein the magnetic component is made of nickel or nickel-iron alloy.

4. The device of claim 1, wherein the recording coil is a single turn coil.

5. The device of claim 4, wherein the recording coil is made of one of Ti / Au and Ti / Cu / Ti.

6. The device of claim 4, wherein the magnetic component is coaxial with the recording coil.

7. The device of claim 1, wherein the Hall sensor is a cross-shaped Hall sensor Greek

8. The device of claim 7, wherein the Hall sensor includes: a sensing part having a current flow direction responsible for the remaining magnetic polarity of the magnetic component; an input line, coupled to the sensing part, for feeding current to the sensing part; an output line, coupled to the sensing part, for receiving current from the sensing part; and a harvester, coupled to the outlet line and to the detection part, along one side of the Detection part between the output line and the input line, to collect current from the detection part and conduct the current to the output line to indicate a value of the stored data bit.

9. The device of claim 1, wherein the non-volatile memory device is incorporated into one of the radio frequency identification (RFID) devices, a cellular phone, and a personal digital assistant (PDA).

10. The device of claim 1, wherein the non-volatile memory device is made of materials of intense radiation.

11. A method for manufacturing one or more memory cells of a non-volatile magnetic memory device, comprising the steps of: forming a magnetic switch including a magnetic component and a recording coil located next to the magnetic component, the recording coil is coupled to receive a sufficient current to create a magnetic polarity remaining in the magnetic component; Y forming a Hall sensor, placed close to the magnetic component, to detect the remaining magnetic polarity indicative of the stored data bit.

12. The method of claim 11, wherein forming the magnetic switch includes the steps of: forming the recording coil; forming a coaxial magnet point with the recording coil; and elect plasty of a magnetic material on the point of magnet.

13. The method of claim 12, further includes the step of forming a magnet mold on the magnet point before electroplating, the step to form the magnet mold includes the steps of: forming a thick layer of protection on the coil of recording and the point of magnet; cook the protection; and design the protection to form a well around the magnet point.

14. The method of claim 13, wherein the protection is AZ4620 and the protection layer is cook at about 95 ° C for about 4 minutes.

15. The method of claim 11, wherein the magnetic material is subjected to electroplating on the magnet point by means of pulse deposition with an occupation cycle where Von is approximately lms, Vdesp is approximately 49 ms, and the maximum current is of approximately 1.4 mA.

16. The method of claim 12, wherein the magnetic material is one of the following: nickel, nickel-iron, cobalt, iron and CoNiFe.

17. The method of claim 12, wherein the recording coil and the magnet point are formed simultaneously during the same process.

18. The method of claim 17, wherein forming the recording coil and forming the magnet point includes the steps of: forming a double layer of copolymer and polyimide; design a double layer; depositing a conductor by means of electroplasty; Y Separate the double layer, thus forming the recording coil and the magnet point.

19. The method of claim 18, wherein the copolymer is Eli and the polyimide is PMMA.

20. The method of claim 19, wherein forming the double layer includes the steps of: cooking the double layer at about 160 ° C for about 5 minutes; and cool to room temperature.

21. The method of claim 19, wherein the double layer is designed by means of electron beam lithography wherein the exposure energy is approximately 25 kV, the dose of the recording coil is approximately 150 μC / cm2, and The dose of magnet point is approximately 120 μC / cm2.

22. The method of claim 18, wherein the conductor is formed of a titanium layer and a gold layer.

23. The method of claim 22, wherein the titanium layer is about 25 nm and the gold layer is about 150 nm.

24. The method of claim 17, wherein forming the recording coil and forming the magnet point includes the steps of: forming a seed layer of conductive material; form a protective layer on the seed layer; cook soft protective layer; design the protection layer; hard cooking designed protection; and attack with guimics the seed layer, thus forming the recording coil and the magnet point.

25. The method of claim 24, wherein the seed layer is formed of Ti / Cu / Ti.

26. The method of claim 25, wherein the Ti layers are chemically attacked with an HF / HN03 / H20 solution and the gold layer is chemically etched with a solution of HC1 / H202 / H20.

27. The method of claim 24, wherein the protection is AZ5206.

28. The method of claim 27, wherein the protective layer is baked smooth at about 95 ° C and lowered to about 80 ° C over about 6 to 7 minutes, and the design of the protective layer is baked hard at about 125 ° C for about 10 minutes.

29. The method of claim 27, wherein the protection layer is designed by means of electron beam lithography wherein the exposure energy is about 10 kV and the dose is about 6 μC / cm2.

30. The method of claim 11, wherein the step to form the Hall sensor includes: forming an active layer on a substrate; forming a protective layer on the active layer; design a protective layer; form a part of detection by means of chemical attack; forming input and output terminals on the detection part; and forming an insulating layer on the Hall sensor.

31. The method of claim 30, wherein the substrate is a semi-insulating GaAs insert and the active layer is a thin nano-active GaAs film.

32. The method of claim 31, wherein the GaAs film is from about 0.5 μm to about 0.6 μm.

33. The method of claim 30, wherein the protection is 4% PMMA, and the protection layer is designed with electron beam lithography where the exposure energy is about 25 kV, the exposure dose is about 150 μC / cm2, the developer is an MBIK / IPA mixture of approximately 1: 3 ratio, and the development time is approximately 25 seconds.

34. The method of claim 30, wherein the protection is 2% PMMA, and the protection layer is designed with electron beam lithography where the exposure energy is about 15 kV, the exposure dose is about 150 μC / cm2, the developer is an MBIK / IPA mixture of approximately 1: 3 ratio, and the development time is approximately 25 seconds.

35. The method of claim 30, wherein forming the protective layer includes the steps of centrifuging the protection at a speed of centrifugation of about 4000 rpm to form a thickness of about 0.5 μm to about 2 μm; and baking the protective layer at approximately 160 ° C for about 7 minutes.

36. The method of claim 30, wherein the inlet and outlet terminals are formed by means of the cementing process including the steps of: forming a double layer of copolymer and PMMA; form a contact layer; and apply rapid thermal annealing to the double layer.

37. The method of claim 36, wherein the contact layer is made of AuGe with a thickness of about 400 nm and nickel on the AuGe layer.

38. The method of claim 36, wherein the rapid thermal annealing step is performed at 340 ° C at 40 seconds with nitrogen flow.

39. The method of claim 30, wherein the protection is a PI2545 dielectric polyimide centrifuged at a centrifugation speed of about 6000 rpm and gentle cooking at about 25 ° C to about 170 ° C to 240 ° C / h.

40. A method for manufacturing a magnetic switch of a memory cell in a non-volatile magnetic memory device, the steps consist of: forming a recording coil; forming a coaxial magnet point with the recording coil; and electroplating a magnetic material on the magnet point to form a magnetic component located next to the recording coil, the recording coil coupled to receive a sufficient current to create a magnetic polarity remaining in the magnetic component.

41. The method of claim 40, wherein forming the recording coil and forming the magnet point includes the steps of: forming a double layer of copolymer and polyimide; design the double layer; depositing a conductor by means of electroplasty; and separating the double layer, thus forming the recording coil and the magnet point simultaneously.

42. The method of claim 41, wherein the conductor is formed of a titanium layer and a gold layer.

43. The method of claim 40, wherein forming the recording coil and forming the magnet point includes the steps of: forming a seed layer of conductive material; form a protective layer on the seed layer; cook soft protective layer; design the protection layer; hard cooking designed protection; and chemically attacking the seed layer, thereby forming the recording coil and the magnet point simultaneously.

44. The method of claim 43, wherein the seed layer is formed of Ti / Cu / Ti.