Group III-V ferromagnetic/non-magnetic semiconductor heterojunctions and magnetodiodes

Abstract

Ferromagnetic Group III-V semiconductor/non-magnetic Group III-V semiconductor heterojunctions, with a magnetodiode device, to detect heterojunction magnetoresistance responsive to an applied magnetic field.

Claims

1. A magnetodiode device comprising a p-type ferromagnetic semiconductor component comprising a compound of a formula
(In.sub.1-xMn.sub.x)As wherein x is greater than zero and less than about 0.2; an n-type non-magnetic semiconductor component comprising a compound of a formula InAs; a diode coupled to said ferromagnetic semiconductor component and said non-magnetic semiconductor component; and a voltage source for application of voltage across said device, said device positioned in an applied magnetic field, whereby magnetoresistance thereof responsive to said applied magnetic field is positive under said magnetic field applied from about parallel to about perpendicular to direction of a current flow through said diode of said device.

2. The device of claim 1 wherein said ferromagnetic component is an epitaxial film on said non-magnetic component.

3. The device of claim 1 wherein x is less than about 0.04.

4. The device of claim 1 dimensioned to increase current density and magnetoresistance responsive to an applied magnetic field.

5. A magnetodiode device comprising a p-type ferromagnetic semiconductor component comprising a compound of a formula
(In.sub.1-xMn.sub.x)As wherein x is greater than zero and less than about 0.04; an n-type non-magnetic semiconductor component comprising a compound of a formula InAs; a diode coupled to said ferromagnetic semiconductor component and said non-magnetic semiconductor component; and a voltage source for application of voltage across said device, said device positioned in an applied magnetic field, whereby magnetoresistance thereof responsive to said applied magnetic field is positive under said magnetic field applied from about parallel to about perpendicular to direction of a current flow through said diode of said device.

6. The device of claim 5 dimensioned to increase current density and magnetoresistance responsive to an applied magnetic field.

7. The device, of claim 5 wherein said ferromagnetic component is an epitaxial film on said non-magnetic component.

Description

BRIEF DESCRIPTIONS OF THE DRAWINGS

(1) FIG. 1. Current density-voltage characteristics for an In.sub.0.96Mn.sub.0.04As/InAs junction at 78 and 294 K.

(2) FIG. 2 A-B. Forward bias J-V characteristics of an In.sub.0.96Mn.sub.0.04As/InAs junction at 78 K (A). The ideality factor as a function of voltage, calculated after subtracting the linear leakage current from the J-V characteristics, is shown in (B).

(3) FIG. 3. The ideality factor as a function of voltage for an In.sub.0.96Mn.sub.0.04As/InAs junction at varying temperatures.

(4) FIG. 4. The longitudinal magnetoresistance (%) as a function of voltage for an In.sub.0.96Mn.sub.0.04As/InAs junction at 78, 180, and 295 K.

(5) FIG. 5 A-B. The magnetic field dependence of the percent magnetoresistance (A) in the low bias, leakage regime and (B) in the high bias (0.335 V), diffusion regime for an In.sub.0.96Mn.sub.0.04As/InAs junction at 78 K. The solid line in (a) shows the H.sup.1.7 dependence, while the solid line in (B) shows the linear magnetic field dependence.

(6) FIG. 6 A-B. Schematic diagrams of magnetodiodes, illustrating two types of InAs contact, in accordance with this invention. Component dimensions are illustrative, only.

(7) FIG. 7. Schematic representation of a of the p-In.sub.0.96Mn.sub.0.04As/n-InAs mesa diode. As above, component dimensions are non-limiting.

(8) FIG. 8. Forward bias I-V characteristics at 295 K in the presence of magnetic fields ranging from 0-8 T.

(9) FIG. 9 A-B. Percent magnetoresistance as a function of magnetic field at 25, 78, and 295 K (A). Room temperature voltage as a function of magnetic field at I=5, 11, and 15 mA (B).

(10) FIG. 10. Magnetoresistance as a function of forward bias at 295 K and 1 T. The solid line shows a fit to the data using eq. (5).

DETAILED DESCRIPTION OF CERTAIN EMBODIMENTS

(11) This invention can be directed to new magnetoresistive devices, related composites and component materials and/or methods relating to the use of a Group III-V ferromagnetic semiconductor and a Group III-V nonmagnetic semiconductor heterojunction. For instance, such magnetodiodes exhibit a large junction magnetoresistance that is linearly dependent on the applied magnetic field at room temperature. As a result, this invention has applications in magnetic field sensors, gaussmeters, and other magnetoresistive devices, and can be readily integrated into current III-V semiconductor circuitry, enabling e.g., magnetic imaging. Accordingly, such devices can, without limitation, be used in a range of applications in which a continuous change in current is desired with increasing magnetic field. This advantage can be utilized, for instance, in conjunction with high field gaussmeters.

(12) Without restriction to any one embodiment, a magnetodiode device of this invention can comprise a ferromagnetic p-type In.sub.1-xMn.sub.xAs thin film deposited (e.g., by metalorganic vapor phase epitaxy) on an n-type InAs substrate. (See, e.g., FIG. 6.) When such a device is operated in the turn-on region in forward bias, the diode current is strongly dependent on the presence of magnetic fields. Magnetic fields applied either parallel (longitudinal) or perpendicular (transverse) to the flow of current through the devices lead to a positive magnetoresistance. This allows for one device/sensor to measure magnetic fields in all directions, whereas with conventional magnetodiodes at least two sensors would be required for multi-directional sensing. The observed positive magnetoresistance is in contrast to the negative magnetoresistance that was predicted in the art to arise in ferromagnetic/nonmagnetic semiconductor diodes. Magnetoresistance is linearly dependent on the magnetic field for fields greater than 1000 Oe and 2500 Oe, for certain devices operated at 78 K and 295 K, respectively. Such a linear response indicates this invention can be used in conjunction with magnetic field sensors, gaussmeters, and other magnetoresistive devices. Device structures and configurations illustrated herein are provided only for purpose of illustration. Other such structures and component configurations would be understood by those skilled in the art made aware of this invention.

(13) The electronic and magnetotransport properties of epitaxial p-(In,Mn)As/n-InAs heterojunctions were examined in conjunction with this invention. As described below, the junctions were formed by depositing ferromagnetic (In,Mn)As films on InAs (100) substrates using metal-organic vapor phase epitaxy. The current-voltage characteristics of the junctions were measured from 78 to 295 K. At temperatures below 150 K, ohmic currents appear to dominate transport at low bias, followed by defect-assisted tunneling current with increasing bias. At high forward bias, junction transport appears dominated by diffusion current. The magnetoresistance of the junctions were measured as a function of forward bias and applied magnetic field. Without limitation, the magnitude and field dependence of the longitudinal magnetoresistance appear to depend, at least in part, on the junction transport mechanism. Under high bias, a magnetoresistance of 15.7% at 78 K and 8% at 295 K in a 4400 Oe field was measured in an In.sub.0.96Mn.sub.0.04As/InAs junction. At 78 K, the high bias magnetoresistance increases linearly with magnetic field from 1000 to 4600 Oe.

(14) The current density-voltage characteristics of an In.sub.0.96Mn.sub.0.04As/InAs junction at 78 and 294 K are shown in FIG. 1. Diode behavior is observed. At temperatures below 220 K, a linear leakage current is observed under reverse bias. This current also dominates the junction characteristics at low forward bias, as will be discussed below.

(15) The forward bias current density-voltage characteristics of the In.sub.0.96Mn.sub.0.04As/InAs junction measured at 78 K are shown in FIG. 2(a). The total forward bias current density is given by:
J=J.sub.diff+J.sub.gr+J.sub.tun+J.sub.leakage,(1)
where J.sub.diff is the diffusion current, J.sub.gr is the generation-recombination current, J.sub.tun is the tunneling current and J.sub.leakage is the leakage current The measured J-V characteristics were fit to the expression,

(16) $\begin{array}{cc}J=J_0\exp \left(\frac{\mathrm{qV}}{k_{B}T}\right),& \left(2\right)\end{array}$
where is the ideality factor defined by the expression,

(17) $\begin{array}{cc}=\left(\frac{q}{k_{B}T}\right)\left(\frac{\mathrm{dV}}{d\ln I}\right),& \left(3\right)\end{array}$
At low bias, a linear J.sub.leakage component dominates junction transport, indicating the presence of parallel conduction paths shunting the junction. These parallel conduction paths are believed to arise from surface states created during the mesa etching process. After subtracting this ohmic component from the total current, the ideality factor was obtained. Ideality factors are commonly used to determine the origin of diode currents: junction conduction is dominated by diffusion when =1, generation and recombination (or diffusion under high injection conditions) when =2, and defect-assisted tunneling when >2. The ideality factor for the In.sub.0.96Mn.sub.0.04As/InAs junction as a function of bias is shown in FIG. 2(B). The ideality factor increases from 2.2 to 10.6 as the bias is increased from 0.145 to 0.255 V, respectively. This suggests that the conduction mechanism changes from generation and recombination to defect-assisted tunneling as the bias is increased. Conduction from the diffusion current component begins to increase at biases greater than 0.28 V, leading to a decrease in the ideality factor. The ideality factor reaches a local minimum at 0.32 V of =2.6. This ideality factor is greater than the high injection value of =2 presumably due to series resistance and continued tunneling currents, both of which can increase the ideality factor. The increase in ideality factor at biases greater than 0.32 V can be attributed to series resistance.

(18) The temperature dependence of the ideality factor for the In.sub.0.96Mn.sub.0.04As/InAs junction is shown in FIG. 3. The maximum value of the ideality factor decreases with increasing temperature from 10.6 at 78 K to 2.9 at 140 K. The minimum value of the ideality factor arising from diffusion currents also decreases with temperature from 2.6 at 78 K to 1.8 at 140 K. These results indicate that the diffusion current component increases as the temperature increases. Similar transport behavior was also observed in In.sub.1-xMn.sub.x,As/InAs junctions with x=0.035 and x<0.01. At room temperature, the ideality factors of the all diodes measured ranged from 1.3 to 1.7, indicating that the diffusion and generation-recombination currents dominate junction transport.

(19) The current-voltage characteristics have been measured in the presence of an applied magnetic field. FIG. 4 shows the percent magnetoresistance as a function of voltage for the In.sub.0.96Mn.sub.0.04As/InAs junction with a 4400 Oe field applied parallel to the direction of current flow. The percent magnetoresistance is given by 100(R(H)R0)/R0; a decrease in the junction current corresponds to a positive magnetoresistance. The magnitude of the magnetoresistance is directly dependent on the junction transport mechanism. At 78 K, a constant magnetoresistance of 1.7% is observed under low bias from 0.01 to 0.15 V, where the transport is dominated by ohmic leakage currents. The magnetoresistance begins to increase with the onset of tunneling currents at 0.16 V, reaching 6% at 0.27 V. The magnetoresistance remains at roughly 6% until the bias is increased past 0.31 V, at which point the magnetoresistance increases rapidly with increasing forward bias. The onset of the second increase in magnetoresistance corresponds to the local minimum in ideality factor and accordingly to the diffusive transport regime. The large observed magnetoresistance is equal to 15.7% at 0.4 V. Similar magnetoresistance behavior was also measured in a junction with x<0.01. At 180 K, only two magnetoresistance regimes are present as tunneling currents are not observed. The magnetoresistance is constant from 0.01 to 0.1 V. A rapid increase in magnetoresistance is observed as the bias is increased past 0.1 V, which corresponds to the onset of diffusion current. A magnetoresistance of 8% at 0.15 V is observed at room temperature.

(20) The field dependence of the magnetoresistance is given in FIG. 5, as measured at 78 K in an In.sub.0.96Mn.sub.0.04As/InAs junction. The magnetoresistance is subquadratic and follows a H.sup.1.7 power law dependence in the leakage current regime, as shown in FIG. 5(A). At high biases, where diffusion currents dominate transport, the magnetoresistance is linearly dependent on the applied magnetic field from 1000 to 4600 Oe, as shown in FIG. 5(B). A linear magnetoresistance in the diffusive current regime is also observed at 78 K in the junction with x<0.01, while the low bias magnetoresistance exhibits a H.sup.18 dependence.

(21) The magnetoresistance observed at low bias, in the ohmic conduction regime, is attributed to Lorentz scattering of carriers throughout the diode. See, R. A. Stradling and P. C. Klipstein, Growth and characterisation of semiconductors (Hilger, Bristol, England; New York, 1990). As the magnetic and electric fields are parallel a negligible magnetoresistance might be expected. However, longitudinal magnetoresistance is often observed in semiconductors due to defects, dislocations or electrical inhomogeneities that scatter carriers. The measured field dependence of H.sup., where ranges from 1.7 to 1.85, is in relative agreement with the expected H.sup.2 dependence of magnetoresistance due to Lorentz scattering. See, S. M. Sze, Physics of semiconductor devices (Wiley, New York, 1981).

(22) The origin of the unusual linear dependence of the magnetoresistance at high bias is currently under study. Nevertheless, the fact that the magnitude and field dependence depends on the specific injection mechanism suggests that the behavior is junction related. The measured magnetoresistance is not attributed to an increased resistance in the bulk InAs substrate or the InMnAs film. The longitudinal magnetoresistance of an InAs substrate and an InMnAs film was measured in the same configuration as the (In,Mn)As/InAs junctions. In both cases, the measured magnetoresistance was less than 1% and independent of applied voltage.

(23) The observed positive magnetoresistance is counter to the magnetoresistance predicted for a ferromagnetic/nonmagnetic semiconductor junction for diffusive transport. A negative magnetoresistance is predicted with the increase in current proportional to exp[((H)(H=0))/k.sub.BT], where (H) is the conduction band-edge splitting due to a magnetic field H. The fact that a positive magnetoresistance is observed suggests that the conduction band splitting in (In,Mn)As is small and that another magnetoresistance mechanism, the origin of which is presently unknown, plays a greater role in magnetotransport.

(24) With reference to examples 3-5, the measured magnetoresistance can be described using the diode equation and adding a series magnetoresistance term R(H) to the argument. The I-V characteristics of the In.sub.0.96Mn.sub.0.04As/InAs junction are modeled using the equation:

(25) $\begin{array}{cc}I=I_0\exp \left(\frac{V_A}{k_{B}T}\right)\left(\frac{-{\mathrm{IR}}_0}{k_{B}T}\right)\left(\frac{-\mathrm{IR}\left(H\right)}{k_{B}T}\right),& \left(4\right)\end{array}$
where is the ideality factor, R.sub.0 is the zero-field series resistance, and R(H) is the magnetic field dependent series resistance. From the I-V characteristics, the experimental values of I.sub.0, R.sub.0, R(H) and were obtained. The ideality factor and R(H=1 T) were fixed at the experimentally obtained values of =1.41 and R=1.14, respectively, while I.sub.0 and R.sub.0 were varied to achieve the best fit. FIG. 10 shows the fit to the magnetoresistance as a function of applied bias at 1 T and 295 K. (See, example 4 and eq. 5). The values of I.sub.0 and R.sub.0 obtained from the fit, 910.sup.4 A and 1.37 respectively, are in good agreement with the experimentally obtained values of 5.910.sup.4 A and 1.12.

(26) One possible origin of the magnetoresistance is carrier scattering due to fluctuations and clustering of the Mn ions at or near the junction. Parish and Littlewood predicted that mobility disorder can result in a non-saturating linear magnetoresistance in semiconductors. Without restriction to any one theory or mode of operation, an inhomogeneous distribution of magnetic Mn ions could cause fluctuations in the scattering rate of injected electrons, giving rise to such local variations in carrier mobility. While the model of Parish and Littlewood has not been applied to p-n junctions, it should be noted that in Ag.sub.2+Te a linear magnetoresistance emerged when the electron and hole concentrations in the material were equivalent. This is consistent with the present results in that the linear magnetoresistance in p-(In,Mn)As/n-InAs junctions occurs only under high injection conditions when the electron and hole concentrations near the depletion region are equivalent. Further, extended x-ray absorption fine structure (EXAFS) measurements have provided evidence that Mn ions cluster into nearest neighbor cation sites forming dimers and trimers in MOVPE grown InMnAs films.

EXAMPLES OF THE INVENTION

Example 1

(27) P-type In.sub.1-x Mn,As films were deposited at 480 C. by atmospheric pressure metal-organic vapor phase epitaxy on nominally undoped, n-type InAs (100) substrates to form the heterojunctions. Precursors used were trimethylindium (TMIn), 0.3% arsine (AsH.sub.3) in hydrogen and tricarbonyl(methylcyclopentadienyl)manganese (TCMn). A pre-growth anneal at 510 C. was carried out under an arsine overpressure in order to remove surface oxide from the InAs substrate. A detailed description of the growth conditions for (In,Mn)As films has previously been reported. See, A. J. Blattner, J. Lensch, and B. W. Wessels, Journal of Electronic Materials 30, 1408 (2001). Manganese concentrations were determined using standards-based energy dispersive x-ray spectroscopy (EDS). X-ray diffraction was used to verify phase purity of the (In,Mn)As films. Film thickness, as determined by profilometry, ranged from about 100-about 525 nm. Room temperature hole concentrations of (In,Mn)As films are on the order of 10.sup.18-10.sup.19 cm.sup.3. The room temperature electron concentration of the InAs substrates was 1.410.sup.17 cm.sup.3. Magneto-optical Kerr effect (MOKE) measurements indicated that the In.sub.0.96Mn.sub.0.04As and In.sub.0.965Mn.sub.0.035As layers were ferromagnetic at room temperature.

Example 2

(28) Mesa diodes were fabricated from the epitaxial structures using conventional photolithography and a citric acid wet etch. See, G. C. Desalvo, R. Kaspi, and C. A. Bozada, J. Electrochem. Soc. 141, 3526 (1994). The mesa diameter was 250 m. Ti/Au (15/175 nm) ohmic contacts were evaporated to the (In,Mn)As films. Silver paste was used to make contact to the InAs substrates. The diodes were wire bonded to a non-magnetic chip carrier with gold wire. Low-temperature measurements were carried out in a Janis ST-100 cryostat with a non-magnetic sample holder. A Keithley 2400 source-meter was used to source voltages and measure currents. Magnetoresistance measurements were made in fields of up to 4600 Oe applied parallel to current flow across the junction.

(29) Using a similar approach, high field, longitudinal magnetoresistance measurements were taken on a p-In.sub.0.96Mn.sub.0.04As/n-InAs junction. A nonsaturating linear magnetoresistance at fields greater than 1.5 T was observed at room temperature. The measured magnetoresistance is well described by a modified diode equation with a magnetic field-dependent series resistance.

Example 3

(30) For these studies, a 500 nm thick p-In.sub.0.96Mn.sub.0.04As layer was deposited on an undoped n-InAs substrate using atmospheric pressure metal-organic vapor phase epitaxy under conditions previously described. (A. J. Blattner, J. Lensch, and B. W. Wessels, J. Electron. Mater. 30, 1408 (2001).) The room temperature electron concentration of the InAs substrate is 210.sup.16 cm.sup.3, while the room temperature hole concentration of the p-InMnAs film is on the order of 10.sup.18-10.sup.19 cm.sup.3. The diodes were patterned into 250 Lm diameter mesas using photolithography and wet etching. Ti/Au was used as the p-In.sub.0.96Mn.sub.0.04As ohmic contact, while Ag was used as the n-InAs ohmic contact. Magnetoresistance measurements were made with the magnetic field applied parallel to the flow of current through the mesa diode using a Quantum Design Physical Properties Measurement System. A schematic of the experimental setup is shown in FIG. 7.

Example 4

(31) FIG. 8 shows the room temperature I-V measurements made with applied magnetic fields over the range of 0 to 8 T. The junction becomes more resistive with increasing magnetic field and shows no sign of saturating with field. The magnetoresistance, MR, defined as

(32) $\begin{array}{cc}\mathrm{MR}=\left(\frac{\mathrm{dV}\left(H\right)}{\mathrm{dI}}-\frac{\mathrm{dV}\left(H=0\right)}{\mathrm{dI}}\right)/\left(\frac{\mathrm{dV}\left(H=0\right)}{\mathrm{dI}}\right),& \left(5\right)\end{array}$
is shown in FIG. 9A as a function of applied field for a constant current of 11 mA at 25, 78 and 295 K. The room temperature magnetoresistance increases linearly with field from 1.5 to 9 T. The magnitude of the linear magnetoresistance increases with increasing applied current from 360% (5 mA) to 710% (15 mA) in a 9 T field. The field dependence of the magnetoresistance becomes sublinear with decreasing temperatures. At 78 K the magnetoresistance is proportional to H.sup.0.84, while at 25 K the magnetoresistance is proportional to H.sup.0.64.

Example 5

(33) Due to its large room temperature magnetoresistance, this device holds promise for magnetic sensing applications. (H. P. Baltes and R. S. Popovic, Proc. IEEE 74, 1107 (1986); R. S. Popovic, Hall Effect Devices (Institute of Physics, Bristol; Philadelphia, 2004).) The change in voltage as a function of magnetic field for three operating currents at room temperature is shown in FIG. 9B. For all currents the change in voltage is linear with increasing magnetic field from 1.5 to 9 T.

(34) As shown above and in the accompanying figures, the current-voltage characteristics of p-(In,Mn)As/n-InAs heterojunctions have been measured over the temperature range of 78 to 295 K. Three forward bias transport components are observed in the junctions at temperatures below 150 K. Ohmic leakage currents are observed at low bias, defect-assisted tunneling currents at intermediate bias and diffusion currents dominate at high bias. A positive magnetoresistance is observed with the application of a longitudinal magnetic field. The magnitude and field dependence of the magnetoresistance depend on the junction transport mechanism. A magnetoresistance of 15.7% and 8% is observed under high bias at 78 and 295 K, respectively. At 78 K, the high bias magnetoresistance increases linearly with increasing magnetic field from 1000 to 4600 Oe.

(35) The high field magnetoresistive properties of a p-In.sub.0.96Mn.sub.0.04As/n-InAs junction were also measured. Under forward bias, a large, nonsaturating magnetoresistance is observed at temperatures from 25 to 295 K in fields up to 9 T. At room temperature, the magnetoresistance increases linearly with magnetic field from 1.5 to 9 T and is greater than 700% at 9 T. As shown, the magnetoresistance can be simulated using a modified diode equation, including a field-dependent series magnetoresistance.