PATENTSCOPE sera indisponible durant quelques heures pour des raisons de maintenance le lundi 03.02.2020 à 10:00 AM CET
Recherche dans les collections de brevets nationales et internationales
Une partie du contenu de cette demande n'est pas disponible pour le moment.
Si cette situation persiste, contactez-nous auObservations et contact
Note: Texte fondé sur des processus automatiques de reconnaissance optique de caractères. Seule la version PDF a une valeur juridique


Technical Field
The invention is an integrated photodetector-, amplifier semiconductor device.
Background of the Invention
The rapid development of optical technology, especially optical communication technology, has increased the importance of various kinds of optical devices.
Particularly important among these optical devices are optical detectors. High sensitivity and high speed are among characteristics of importance for optical detectors used in optical communication systems. Also of importance are high signal gain and high reliability as well as ease in maintaining the optical detector device under typical commercial operating conditions.
A number of references are useful as background in this area of technology. Particularly illuminating is an article entitled "Recent Developments in Monolithic
Integration of InGaAsP/lnP Optoelectronic Devices" by
U. Koren et al , IEEE Transaction on Microwave Theory and Techniques, MTT-30 (10), October 1982, pp. 1641-1649.
Another article useful in understanding the invention is entitled "A Self-Aligned InQ -.,Ga0 47As Junction
Field-Effect Transistor Grown by Molecular Beam Epitaxy" by D . Wake et al , IEEE Electron Device Letters, EDL-5 ( 7 ) , July 1984, pp. 285-287.
Summary of the Invention
The invention is an integrated photodetector-amplifier structure featuring a p-region, intrinsic region, n-region photodetector (PIN photodetector) and a self-aligned junction field-effect transistor (JFET)
incorporated in the same integrated circuit. The
photodetector is back illuminated, features a buried p-n junction so as to avoid surface leakage, vertical
integration so that partial regrowth is not needed, and an air bridge to isolate the capacitance of the PIN electrode (the p-contact on the PIN structure) from the JFET
structure. Such integrated photodetector devices have extraordinary properties including high sensitivity, high response frequency and ease of fabrication. The invention includes integrated structures with multiple elements; that is, more than one PIN and/or more than one JFET element. Indeed, a preferred embodiment of the invention is an integrated PINFET structure featuring a dual gate to minimize parasitic capacitance from n-contacts. Another embodiment features a PINFET integrated with a
transi pedance amplifier to provide additional stages of amplification and to provide negative feedback to the input stage.
Brief Description of the Drawing
FIG. 1 shows a side view of an integrated PINFET structure;
FIG. 2 shows a prospective view of the same structure as shown in FIG. 1;
FIG. 3 shows a side view -of an epitaxial Layer semiconductor structure useful in the fabrication of PINFET structures;
FIG. 4 shows a variation of the PINFET structure described in FIGS. 1 and 2.
FIG. 5 shows a circuit diagram useful in the practice of the invention; and
FIG 6 shows a circuit diagram useful in the practice of the invention.
Detailed Description
The invention is based on the discovery that a vertically integrated photodetector-amplifier could be made using epitaxial growth techniques and semiconductor
processing techniques. This photodetector-ampli ier structure has a number of features that are highly
advantageous. First, it is made by putting down epitaxial layers without any melt-back or regrowth steps. The junction is buried under the epitaxial layers so that surface leakage is not a problem. There is an air bridge between the p-contact of the PIN photodetector and the gate electrode of the junction field-effect transistor (JFET) . The structure is self-aligned to facilitate the fabrication of fine-line gate structures. The structure is nearly t planar (does not have sudden, sharp steps or ridges) so that conventional fabrication techniques can be used to make the integrated detector-amplifier device.
The invention is best illustrated by a detailed description of one embodiment of the invention; namely, a detector-amplifier combination usually called a PINFET made for radiation detection at wavelengths less than 1.65μm. It should be recognized that the PINFET is the basic unit of the invention but more elaborate circuits (usually containing more JFET structures) may be "used in the
practice of the invention. Also, reasonable variations of this embodiment which use the basic principles of the invention are also included in the invention. This
structure is useful in illustrating the principles of the invention although in practice more elaborate
configurations are often used.
First is described a side view of the PINFET to show the epitaxial layer structure, the p-diffusion region and air bridge from the p-contact to the PIN photodetector structure to the gate electrode of the JFET structure.
FIG. 1 shows a side view of the PINFET cut through the PIN detector part and the gate width of the field-effect transistor. The PINFET J_0 is made up of layers of III-V semiconductor compounds beginning with a substrate 11 of
N -InP generally doped in the range from 1- 5 x 10
atoms/cm2 with tin or sulfur. This layer is generally about 150 μm thick. On top of this layer is an undoped

(intrinsic) layer 12 of InGaAs (approximate composition

In0.53Gao 47As*) lattice-matched to InP. The
thickness of this layer is about 5 μm. The bandgap of this layer (0.75 eV) is such as to ensure absorption of the incident radiation (radiation with energy greater than 0.75 eV or wavelength shorter than 1.65μm) in the layer and the doping level is low to ensure extension of the electric field gradient throughout the layer. Generally, impurities make the doping n-type with a concentration of about 5 x

10 atoms/cc.
On top of this layer 12 is a thin layer 13 of semi-insulating material (often called the channel
confinement layer), generally semi-insulating InP (bandgap of 1.35 eV) or undoped InAlAs (bandgap of 1.4 eV) . The thickness is generally about 1μm. This layer serves to isolate the JFET electrically from the remainder of the circuit and provide an insulating layer against which the channel layer can be pinched off. The composition of
InAlAs for lattice-matching to InP is approximately
In0_ .5_-2Al0- .4.8QAs. The semi-insulating - InP is
usually Fe doped either by ion implantation or grown by MO-CVD metallo-organic chemical vapor deposition.
Above the insulating layer are two more III-V semiconductor layers used to form the n. and p_' layers of the JFET. These layers are n-type InGaAs 14 (lattice-matched to InP) and p-type InGaAs 15 -(lattice matched to InP).' Other materials (principally Inp or InGaAsP). may be used for these layers. Typical thicknesses are 0.4 μm for the n-InGaAs and 0.6 μm for the p-InGaAs. These layers are loosely referred to as the n-type JFET layer (or channel layer) and the p-type JFET layer. These layers will be discussed in more detail below.
Various metallizations are also part of the PINFET structure. For example, attached to the N+-InP layer 11 is an n-type contact 16 (typically Au-Sn) with an opening to admit radiation. The surface of the N+-InP layer 11 exposed by the opening in the n-type contact 16 is covered with an anti-reflection (AR) coating 17 to reduce
reflections at the surface.
Also shown are some p-contacts generally made of Cr-Au alloy. For example, such an alloy forms the p-contact 18 for the PIN part of the structure and the p-contact 19 for the JFET gate electrode. The same metallization forms the air bridge 20 connecting the PIN part of the circuit to the gate 19 of the JFET. The p-contact layer 18 is on top of the PIN part of the
structure, and the p-contact 19 layer is on top of the JFET structure, more specifically layer 18 is on top of region 15B of layer 15 and layer 19 is on top of region 15A of layer 15. Not shown in FIG. 1 (but see FIG. 2) are source electrode 24 and drain electrode 25.
Also shown is the P+-type region 21 which forms the £ region of the PIN structure. This is generally introduced by diffusing zinc through the top three layers (13, 14 and 15) prior to putting down the p-contact 18 of the PIN structure. Doping concentration is usually in the range 10 18-1019 atoms/cm- ~3. Various passivating
layers may also be used (e.g., SiNχ) but are not shown.
Because the section shown goes down the center of the gate electrode, the source and drain electrodes are not shown in this figure.
Two .important features of ' the invention are worthy of emphasis; namely, gap 22 under the air bridge 20 which isolates the PIN electrode from the gate electrode of the JFET and the mesa isolation trenches 23 which surround the PIN structure and isolate it from the JFET structure.
The salient features of this structure are as follows:
1. The P-N junction of the PIN structure is buried inside the structure giving excellent passivation and avoiding surface leakage.
2. This is a vertically integrated structure where regrowth is not necessary. Further, it is a self-aligned structure where the thickness of the p-InGaAs layer determines the extent of undercutting of the gate electrode as shown below.
3. The air bridge isolates the electrical
characteristics (e.g, capacitance) of the PIN structure from the JFET permitting higher frequency response and more rapid response times.

FIG. 2 shows a perspective view of the same structure made with the p-contact 18 on the PIN
photodetector, the air bridge 2C and the gate electrode 19 of the JFET structure. Also shown are the p-type InGaAs 15 and n-type InGaAs layer 14 which are present both in the PIN part of the structure (as regions 15B and 14B
respectively) and the JFET part of the structure(as regions 15A and 14A, respectively). These layers are removed between the PIN and JFET parts of the structure (under the air bridge 20) so as to provide electrical isolation between photodetector and amplifier parts o.f the structure. The insulating layer (or channel confinement layer) 13 is found below these layers, again to electrically isolate the PIN part of the structure from the JFET part of the
structure and to serve as the confinement layer in the JFET structure. Next is the absorption layer 12 typically made of InGaAs lattice-matched to InP and the InP substrate 11. An n-type contact layer 16 serves as one of the electrical contacts of- the PIN structure.
Also shown in relief is the P+-tyρe region 21 which extends through the JFET layers 14, 15 and through the insulating layer 13 and partially into the absorbing layer 12. This forms the p-n junction of the photodetector and provides conduction to the p-contact 18 of the PIN photodetector. Thus, the p-n junction of the photodetector structure is buried inside the structure rather than on the surface. Also shown in FIG. 2 is a circuit diagram of the structure.
A particular advantage of the inventive structure is the ease of fabricating the structure in such a way as to obtain optimum operation. A convenient point to begin a description of the fabrication procedure is at the layered structure as shown in FIG. 3. This structure is made by various epitaxial layer growth techniques starting with an N+-InP substrate.
The layered structure consists of an N+-InP substrate 11 doped with sulfur or tin in the concentration range from 1-5 x 10 18 atoms/cir ~ . This layer is
eventually thinned down to a thickness of 150 μm generally later in the fabrication procedure. On top of the
substrate is an undoped layer 12 of InGaAs with a thickness of approximately 5 μ . This layer has approximate
comp cosition In0_ .5_3-Ga0..4.-7.As so that it is
approximately lattice-matched to InP. Impurities generally make this layer n-type with a doping concentration in the
14 -3
range of 1-30 x 10 atoms/cir . Next is found an
intrinsic layer 13 usually made of semi-insulating iron doped InP or undoped lattice-matched InAlAs. This is often called the channel confinement layer. The thickness of this layer is approximately 1 μm. Next come the two JFET

layers; first the n-type InGaAs layer 14 generally doped
16 with silicon in the concentration range of 4-8 x 10
atoms/cm3. The thickness of this layer, often called the channel layer, is generally between 0.3 and 0.5 μ .
'The other JFET layer is a p-InGaAs layer 15. and is typically doped with Be. Often the doping concentration is different in different parts of the p-InGaAs layer. For most of the thickness of the p-InGaAs layer (typically 0.5 to 0.6 μm) including at the interface with the n-InGaAs layer, the doping concentration is about 1-5 x 10 18
atoms/cm3; for a thin layer on top of the p-layer
(typically 50 nm) , the doping concentration is about 1-2 x

10 19 atoms/cm-3..
An exemplary process for fabricating this
structure is as follows; before carrying out any processing steps, the structure shown in FIG. 3 is oriented so that a

<110> direction in the [100] wafer is directed so that the long dimension of the gate (gate width) lies along this direction.
First, the p-diffusion procedure is carried out to form the p-n junction of the PIN structure. A SiN layer is deposited on the p-InGaAs. This is patterned to serve as a diffusion mask to define the active PIN area 21 in FIG. 2. The zinc vapor diffusion is carried out so that the
diffusion depth just penetrates the upper portion of the intrinsic InGaAs layer 12. After the diffusion, the SiN mask is removed. Alternatively the SiN mask may be used to etch away the £ and n_ InGaAs layers in the diffusion region, followed by zinc diffusion through the semi-insulating layer alone.
Then the p-contact metal patterns (including the PIN contact and gate contacts of the various FETS) are defined photolithographically using Cr-Au or other
suitable p-contact metals.
These metal contacts are made so they can serve as shadow masks when the source and drain of the JFETs are formed.
A note on nomenclature might be of use at this point. The dimension of the gate electrode in the plane of the channel layer pointing toward source and drain
electrodes is called the lengtn of the gate 'even though this is usually the short dimension of the gate electrode and the direction perpendicular to the length, usually extending between source and drain electrodes is called the width of the gate although this is often the long dimension of the gate electrode.
A critical part of the fabrication procedure is removal of the p-layer (including some undercutting of the gate electrode) without substantially affecting the n-layer or substrate. Generally, the exact procedure depends on the nature of the III-V semiconductor material in the n-layer and p-layer.
Particularly important in the practice of the invention is obtaining a suitable shape for the undercut post p-type material so that the gate electrode acts as an effective shadow mask for the evaporated source. The p-InGaAs layer is etched away using a 50 volume percent solution of citric acid with added aqueous hydrogen
peroxide and -an electrical monitoring system to measure the progress in the etching process. The current-voltage characteristics of two adjacent gate electrodes are
measured and the etch stopped when the resistive
characteristics disappear and only back-to-back diode .
characteristics remain. This indicates removal of the p-layer between adjacent gate electrodes with only the p-n junction under each gate electrode remaining. Removal of the p-layer also results in undercutting under the gate electrode by a distance approximately equal to the original thickness of the p-layer. The overhang of the gate
electrode over the p-layer post ensures self-alignment of the source and drain electrodes deposited later.
The n-contact metallization of source and drain of the JFET is put down by standard photolithographic
techniques using typically Ge-Au. This is a self-aligned process because of the overhang on the undercut gate. This yields very close spacings between drain and source
contacts (typically 2 μm) without the danger of contact with the p-layer.
Next, mesa isolation is carried out by covering the JFET structure and etching away the n-InGaAs including the bridge region 22. Typical etchant is 50 volume percent citric acid with aqueous hydrogen peroxide. At this point, the substrate (N+-InP) may be thinned to 150 μm.
On the back side of the substrate (N+-InP), an n-contact is put down except for openings directly
underneath the active PIN areas to allow admittance of incident light.
A number of variations may be used in the practice of the invention. For example, other III-V semiconductor materials may be used for the two JFET layers provided they are lattice-matched to the InP substrate (or other material lattice-matched to the InP substrate). The device described above and in FIGS. 1, 2 and 3 uses lattice-matched InGaAs (approximate composition
In0..5,3,Ga0..4.-7.As). Other materials that could be
substituted for InGaAs at either the n-layer or p-layer are as follows: InGaAsP lattice-matched to InP, InP, other * ternary or quaternary III-V compounds that can be lattice- matched to InP.
Typical combinations and etching systems are as 5 follows:
1. Both p-layer and n-layer are InP. Typical etches are hydrochloric-phosphoric acid (typically in the volume ratio of 1:4 using concentrated hydrochloric acid) sometimes with small amounts of hydrogen peroxide added, or 0 alternatively broπ-ine-methanol (usually dilute solution). The same monitoring technique can be used as described above. Typical etch rate is 300-600 nm/minute.
2. P-layer InGaAs (or InGaAsP) lattice-matched to InP and n-layer InP. A typical etch is citric acid — with 5 aqueous hydrogen peroxide mentioned above. The etch will remove the p-layer and stop at the n-type InP. The etch rate is about.80-120 nm/minute. The same monitoring procedure as described above can be used to observe the progress of the etching process. However, since the etch 0 rate greatly decreases on reaching the InP, this technique is largely self-stopping and thus electrical monitoring is not essential.
3. P-layer is InP and n-layer is InGaAs or
InGaAsP lattice-matched to InP. A typical etch is HCl; 5 phosphoric acid (composition described above) which etches InP but not InGaAs or InGaAsP. Again, the etching process can be monitored as described above, although once again this is not essential to successful processing.
FIG. 4 shows a variation of the PINFET structure 0 0_ described above, in which a double gate electrode is used to decrease the long dimension of the device. In addition, the double gate structure reduces the area of the drain pad needed, thus reducing the capacitance of the drain electrode. The device is very similar to the device 5 described above except for the double gate electrode and two outside source electrodes. The device is made up of substrate 41 made of N+-InP, an absorbing layer 42 made of instrinsic InGaAs, buffer or isolation layer 43 (usually called channel confinement layer) made of undoped AlInAs or iron-doped semi-insulating InP, channel layer 44 of n-type InGaAs and p+- layer of InGaAs 45 (gate p-layer) for the junction on the JFET. Again, we see the p-diffusion region 46 under the PIN electrode 47, the n-contact 48 to the InP substrate and the opening, with anti-reflection coating 49 under the PIN structure. The circuit diagram is much like the one for the simple-gate structure shown in FIGS. 1 and 2.
The basic PINFET structure is usually incorporated in various kinds of structures so as to integrate
photodetector, amplifier and various other circuit elements in one semiconductor chip. FIG. 5 shows a typical base circuit unit in which the invention is often used. Here, RB is a bias resistance and RL is a load resistance of the circuit.
In most circuit applications, the resistors (Rβ and RL) are usually made in the form of FETS
appropriately biased and placed in the semiconductor structure to produce the desired characteristics.
Generally, the load resistance is produced by a JFET structure close to identical to the JFET structures
described above so as to minimize capacitance and produce good electrical characteristics even when doping profile is not optimum. Often, such a FET has the gate electrode connected to the source electrode. The bias* resistance is also usually supplied by a FET structure but not
necessarily a structure as described above and has a large gate length compared to the JFETS described above. ,
More complex circuits are also useful in the -practice of the invention. FIG. 6 shows a typical t transi pedance amplifier circuit, including cascode stage useful in the practice of the invention.