Semiconductor devices and methods with tunnel contact hole sources and non-continuous barrier layer

[0002] This application claims priority from U.S. Provisional Patent Application No. 60/136,159, filed May 27, 1999, and said Provisional Patent Application is incorporated herein by reference.

[0004] This invention relates to semiconductor devices and methods. The invention has particularly advantageous application to III-V semiconductor light emitting devices, including semiconductor lasers and semiconductor light emitting diodes (LEDs).

[0005] In the U.S. Pat. No. 5,936,266 of Holonyak et al. there are disclosed III-V semiconductor devices and methods in which the amount of p-type material can be minimized, with attendant advantages in electrical, thermal, and optical performance, and in fabrication. A form of that invention is directed to a generally planar semiconductor device wherein a layer of p-type semiconductor material is disposed over (that is, directly on or with one or more intervening layers) a layer of n-type semiconductor material, and an electric potential is coupled between the p-type layer and the n-type layer, and wherein current in the device that is lateral to the plane of the layers is coupled into the p-type layer. A tunnel junction is provided adjacent the p-type layer for converting the lateral current into hole current. The tunnel junction can be an n+/p+ junction oriented with the p+ portion thereof adjacent said p-type layer. The lateral current can be electron current in the n+ layer and/or electron current in a further layer of n-type material disposed over the tunnel junction. An objective of the referenced invention was to minimize the amount of p-type material (with low mobility hole conduction) and, to the extent possible, employ only n-type layers (electron conduction) to carry the device current. In addition to electrical and thermal performance advantages from reducing the amount of lossier p-type material, optical advantage can also accrue since p-type material of the same conductance as n-type material will generally be more absorptive of the light being generated in semiconductor light emitting devices. Conversely, it was noted that since the tunnel contact junction is highly doped, it should be kept relatively thin to avoid undue light absorption.

[0006] It is among the objects of the present invention to provide devices of the general type described above, but with novel structural modification that can result in further operational improvement.

[0007] A form of the invention set forth in the above-referenced U.S. Pat. No. 5,936,266 utilizes a quantum well lower gap barrier layer, for example of In_yGa_1−yAs, between the p+ GaAs and n+ GaAs layers to reduce energy gap and carrier mass, and increase tunneling probability.

[0008] In accordance with a feature hereof, the thin In_xGa_1−xAs layer can be non-continuous. The In_xGa_1−xAs layer could be “cut up” into parallelepiped boxes or quantum boxes or quantum dots (as in quantum dot injection devices, e.g., lasers) but, for the present purpose, to provide an array of thin tiny box-like or dot-like tunnel barriers. An electron tunnels with high probability via the dot. Thus, for example, in a VCSEL driven with lateral electron current, sufficient injection is expected (via the tunnel dots or islands) for VCSEL operation, but with reduced absorption of the recombination radiation (photons) in the laser operation directed vertically through the tunnel region. Thus, a tunnel contact of reduced absorption is possible in spite of the lower gap of the quantum dot barriers.

[0009] Quantum dot injection devices, such as quantum dot lasers, are known in the art, such as for the purpose of lowering threshold currents in laser diodes. As described, for example, in K. Eberl, “Quantum Dot Lasers”, Physics World, September, 1997, a quantum dot confines the motion of electrons in all three spatial directions such that electronic states are squeezed into discrete transition energies so that fewer carriers are needed to create a population inversion. Temperature dependence is also reduced due to the discrete energy spectrum and large separation between energy levels. For examples of quantum dot devices and techniques, see H. Saito et al., “Room Temperature Lasing Operation Of A Quantum Dot Vertical Cavity Surface Emitting Laser”, Appl. Phys. Lett. 69 (21), Nov. 18, 1996; M. Maximov et al., “High Power Continuous Wave Operation InGaAs/AlGaAs Quantum Dot Laser”, J. Appl. Phys., 83, 10, May, 1998.

[0010] In the present invention, the non-continuous layer, in the form of quantum dots or other forms, is used to advantage in the barrier layer of a tunnel junction of a device that employs lateral electron flow into the tunnel junction.

[0011] In accordance with an embodiment of the invention, there is provided a semiconductor light emitting device that includes: a semiconductor active region disposed between first and second semiconductor layers, the first semiconductor layer being p-type, and the second semiconductor layer being n-type; tunnel junction means disposed over the first semiconductor layer, the tunnel junction means including a tunnel barrier region that is a non-continuous layer; means for coupling electric potential between the tunnel junction means and the second semiconductor layer; and means for causing lateral electron flow into the tunnel junction means.

[0012] In a preferred embodiment of the invention, the tunnel junction means comprises an n+/p+ junction oriented with the p+ portion thereof adjacent the first semiconductor layer, and wherein the n+/p+ junction includes the tunnel barrier region that is a non-continuous layer. In this embodiment, the n+/p+ junction is a GaAs junction modified with an In_xGa_1−xAs tunnel barrier region which is the non-continuous layer. The non-continuous layer can be in the form of a pattern or can be stochastic in form and/or arrangement.

[0013] Further features and advantages of the invention will become more readily apparent from the following detailed description when taken in conjunction with the accompanying drawings.

[0014] FIG. 1 is a cross-sectional view of a light emitting device of the type disclosed in U.S. Pat. No. 5,936,266, as modified in accordance with the improvement of the invention.

[0015] FIG. 2 is a partially broken away cross-sectional view that illustrates an example of a non-continuous barrier layer in the device of FIG. 1.

[0016] FIG. 3 is a top partially broken away view of a portion of the non-continuous barrier layer of FIG. 2.

[0017] An embodiment of the above-referenced U.S. Pat. No. 5,936,266 is shown in FIG. 1. A light emitting diode (LED) includes an n-type substrate 1205 having an n-type confining layer 1215 disposed thereon. The substrate 1205 may be, for example, GaAs or GaP, and the lower confining layer 1215 may be, for example, n-type AlGaAs. An active region is represented at 1230, and is selected to obtain the desired light wavelength. As one example, the active region may be bulk AlGaInP or may include one or more quantum wells of GaAs or InGaP between layers of AlGaInP. An upper confining layer 1240 is a p-type layer, for example a layer of p-type AlGaAs or AlInP. Above the p-type upper confining layer 1240 is the p+ layer 1252 and an n+ layer 1254. Both of these layers may be heavily doped GaAs or, of particular interest herein, the n+ portion may include two n+ layers, such as a layer of n+ InGaAs with n+ GaAs thereon. The tunnel junction is labelled at 1250. As described in Richard, Chen, Sugg, Hofler, and Holonyak, Appl. Phys. Lett., 63, (26), 1993, a basic GaAs p-n tunnel junction can be modified by employing a thin (e.g. ˜100 Angstrom) In_xGa_1−xAs strained layer in the barrier region to reduce the energy gap (ΔE_g=100-130 mev) and carrier mass and thus increase the tunneling probability without sacrificing the high injection barrier and voltage of GaAs. In essence, low gap tunneling was introduced into a high gap injection device. Metal contacts 1261 and 1265, which may be, for example Ti/Au or Ge/Au, are respectively applied to the top and bottom of the light emitting diode. A positive potential can be applied via the contacts, as shown. Further embodiments are set forth in the above-referenced U.S. Pat. 5,936,266 showing the same or similar principle being employed in laser operation, such as in a VCSEL (vertical cavity surface emitting laser), which can preferably have a collar formed of a native oxide of aluminum gallium arsenide, and which can employ lateral electron current confined at least in part by a native oxide.

[0018] In accordance with a feature hereof, the thin In_xGa_1−xAs layer can be non-continuous. The In_xGa_1−xAs layer could be “cut up” into parallelepiped boxes or quantum boxes or quantum dots (as in quantum dot injection devices, e.g., lasers) but, for the present purpose, to provide an array of thin tiny box-like or dot-like tunnel barriers. An electron tunnels with high probability via the dot. Thus, for example, in a VCSEL driven with lateral electron current, sufficient injection is expected (via the tunnel dots or islands) for VCSEL operation, but with reduced absorption of the recombination radiation (photons) in the laser operation directed vertically through the tunnel region. The reduced absorption would also be advantageous in LED operation, as in the simplified illustration. Thus, a tunnel contact of reduced absorption is possible in spite of the lower gap of the quantum dot barriers (which, it is noted, can be stochastic in form and arrangement).

[0019] FIG. 2 shows an example of a non-continuous barrier layer 1251 in the device of FIG. 1. Layers 1252 and 1254 are respectively the p+ and n+ layers (e.g. heavily doped GaAs layers) of the FIG. 1 device, and, in this example, the non-continuous barrier layer 1251 comprises quantum dots of n+ InGaAs in the form of truncated prisms or truncated cones. FIG. 3 shows an example of a portion of the layers 1252 and 1251, before deposition of the layer 1254. The dots, or other configuration, preferably cover less than half the area of layer 1252, and preferably cover a fraction of the area in the range 0.05 to 0.50, so that the available uncovered area fraction is preferably in the range 0.50 to 0.95, with substantial reduction in absorption of recombination radiation from the active region of the device.

[0020] The non-continuous layer can be obtained by any suitable technique (e.g. of the type used in existing quantum dot laser fabrication), for example by employing a tilted crystallographic plane for seeding of the dot regions, by droplet epitaxy, or by any other suitable technique known in the art.

[0021] It will be understood that GaAs and InGaAs are illustrative and that other III-V higher-gap lower-gap systems, for example AlGaAs-GaAs-InGaAs, InAlGaP-InGaP-GaAs, and other combinations.