WO1991016747A1

WO1991016747A1 - High power laser diode

Info

Publication number: WO1991016747A1
Application number: PCT/RO1991/000002
Authority: WO
Inventors: Iulian Basarab Petrescu-Prahova; Sorina Lazanu
Original assignee: Institutul De Fizica Atomica
Priority date: 1990-04-20
Filing date: 1991-04-04
Publication date: 1991-10-31
Also published as: IL97898A0; RO102871B1; EP0478755A1

Abstract

High power laser diode consisting of a limiting, a main, an intermediate, an active, and an excitation region, which has a confining factor comprised between 1.5*10-4 and 1.5*10-3 and a resonator length comprised between 1.5 and 3 cm, wherein, due to a small refractive index difference between the main and limitation regions, only the fundamental transverse mode, which has the maximum of the flux density situated in the main region, can propagate.

Description

HIGH POWER LASER DIODE

FIELD OF THE INVENTION

This invention relates to high power laser diodes which can be used for laser emission at increased densities of the radiation flux and in the fundamental transverse mode.

BACKGROUND OF THE INVENTION

There is known that the increased radiation flux density operation of laser diodes is limited by the catastrophic degradation, phenomenon wherein the semiconductor material ot the active region, situiated at the semiconductor material-external medium interface, this interface being also the mirror of the Fabry- Perot resonator, is subjected to an excessive heating produced by laser radiation abεorbtion near the interface. Due to this excessive heating, the resonator mirror can be destroied by local melting. The destruction starts with the band to band absorbtion of the laser radiation flux near the surface, this absorbtion prevailing over the gain due to the decrease of the free carrier concentration toward the surface as the result of the fast surface nonradiative recombination phenomena. The absorbtion of the laser radiation flux, followed by the dissipation of the absorbed energy in nonradiative processes, produces an initial local heating which in its turn induces the decrease of the band gap and, further on, the increase of absorbtion and heating phenomena rates etc. When a critical value of this local heating is exceeded, the absorbtion and heating processes succeed each other so fast that they lead to the catastrophic degradation phenomenon, the intensity of which depends on the magnitude of the surface recombination rates and on the magnitude of tne laser radiation flux density reaching the surface and being absorbed. The majority of laser diodes are fabricated from a succession of epitaxial, parallel layers, including the active layer, and the Fabry-Perot resonator mirrors are obtained by cleaving the semiconductor crystal perpendicular to the plane of the epitaxial layers, so that they inherently conserve the disadvantage of the catastrophic degradation at relatively small radiation flux densities, due to high value of the surface recombination at the active region-external medium interface.

There are also known laser diodes with transparent mirrors whereat the radiation flux density can be increased by diminishing the surface recombination rate and the influence of mirror absorbtion mechanisms. Such a type of laser diodes presents an active region terminated in the mirror vecinity with a transparent semiconductor material, with a larger than that of the active region band gap, so that the absorbtion mechanisms at the semiconductor material-external medium interface are completely eliminated and, at the active region-transparent medium interface, the interface recombination rate is reduced, the decrease of the carri er concentration i s also reduced, so that the interface absorbtion of the incident flux on the interface is also reduced.

Transparent mirror laser diodes have the disadvantage that for fabricating the transparent region microlithographic, etching and epitaxial growth processes are needed and the mirrors need to be produced by cleaving in a relatively narrow, 20 - 50 μm wide region.

Laser diodes (Great Britain Patent BB 2031644 A) formed from structures with 4 or 5 layers wherein one of the layers is used mainly as a waveguide and is adjoined to the active layer or is separated from this one by a carrier confinement layer so that the radiation flux is distributed both in the active layer and in the waveguide are also known. These diodes present the disadvantage that the thickness of the waveguide is not greater than 2 μm and that the operation in higher order transverse mode in the widest proposed waveguides is possible.

SUMMARY OF THE INVENTION

The aim of this invention is to fabricate laser diodes with as high as possible values for the flux density and for the total emitted flux.

The problem which is solved by this invention is to fabricate laser diodes formed by epitaxial, parallel between the two mirrors, layers, with the wavequide main region separated from the active one, diodes having a very small magnitude of the radiation flux in the active region.

The high power laser diode according to the invention resolves this problem using a structure consisting of a substrate, a limitation region, a main region, an intermediate region, an active region, on excitation region and a contact region, wherein the refractive index of the active region is the hiqhest in the structure, wherein the refractive index of the main region is the next lnferior one, wherein the product ot the active region thickness by the square root of the difference of the squares of the refractive indexes of active, respectively main regions is smaller than a quarter of the vacuum wavelength of the laser radiation, wherein excitation and intermediate regions have conductivity types opposite to each other, in order to opperate at high radiation flux densities and to allow only ths propagation of the fundamental transverse mode the maximum radiation flux of which i s located in the main region and the effective retractive index or which is comprised between the refractive index values of the limitation and of the main regions, the product of the main region thickness by the square root of the difference of the squares of retractive indexes of the main, respectively limitation regions is smaller than half the vacuum wavelength of laser raoiation, the difference between the refractive indexes of the main and limitation regions is obtained only by different dopping concentrations of these regions, the difference in carrier concentrations being comprised between 7*10⁺¹⁷ and .9*10⁺¹⁷ cm^-3, correlated to the values for the main region thickness between 2 and 5, μm, the confinement factor is comprised between 1-5*10^-4 and 1.5*10^-3 and the resonator lenth is comprised between 1.5 and 3 cm.

BRIEF DESCRIPTION OF THE DRAWINGS

Other features of the diode according to the invention will become apparent from the detailed description which follows, reference being made to FIG. 1...3 in which:

FIG. 1 is a perspective view of the laser diode;

FIG. 2 is a graph of the radiation flux density as a function of the coordinate perpendicular to tne waveguide structure;

FIG.3 is a graph of the refractive index of the semiconductor materials as a function of the coordinate perpendicular to the waveguide structure.

SPECIFIC DESCRIPTION

The laser diode according to the invention, reference being made to FIG. 1, consists of the following layers; a substrate 1, a limitation region 2, a main region 3, an intermediate region 4, ar active region 5, an excitation region 6 and a contact region 7. The regions 1...7 consists of semiconductor materials. The external region 1 is covered with a metallic contact 8 and the external region 7 with a metallic contact 9. The metallic contacts 8 and 9 allow the flow of an electrical exciting current through the laser diode when an external voltage V_ext, with the proper polarity, is applied. The waveguide, consisting of regions 2...6, terminates at one side with a semiconductor material- external medium interface 10 and at the other side with a semiconductor material-external medium interface 11. These two interfaces are plan parallel to each other and perpendicular to the waveguide structure and represent the Fabry-Perot resonator mirrors.

The spatial disposal of the laser diode presented in FIG. 1 is refered to an orthogonal coordination system Oxys. For simplicity a laser diode with a very large width w in the y direction is considered and the laser radiation along the y direction is supposed uniform. The radiation propagates along the waveguide in z direction. The mirrors 10 and 11 are parallel to the xy plane.

A cross-section parallel to the xy plane is considered. In this cross section, the radiation flux density has a dependence determined by a I=I(x) function on the x coordinate perpendicular to the waveguide structure. In FIG. 2 a graph 12 which represents the I(x) function in a (I,x) coordinate system is shown. The shape of the graph is the same in every transverse cross-section parallel to the xy plane. A flux df is passing through the w*dx area of the transverse cross-section:

df=I(x)*w*dx (1)

The limitation region 2 extends from a coordinate x ₁ to coordinate X₂, the main region 3 from the coordinate X₂ to coordinate x₃, the intermediate region 4 from the coordinate x₃ to a coordinate x₄, the active region 5 from the coordinate x₄ to a coordinate x₅, the excitation region from the coordinate x₅ to a coordinate x₆ and the contact region from the coordinate x₆ to a coordinate x ₇. The active region thickness is d_a=x₅-x₄ and the main region thickness is d_m=x₃-x₂. The function I(x) has a maximum value of the radiation flux in the active region, I_max, and a maximum value of the radiation flux density in the active region I_m.a..

The sum of the flux density in the whole (x₁, x₆) interval of the laser diode structure determines a value of the total flux, F, and the sum of the flux density in the (X₄,x₅) interval determines a value f_a of the flux passing through the active region cross-section. The emitted flux passing through the active region, f_a, is proportional to the double hatched area in FIG. 2 and the total radiation flux F passing through the whole cross- section is proportional to the hatched area, including the double hatched one. The confinement factor T is the subunitar ratio:

= f_a/F (2)

The I=I (x) function represented by the graph 12 is characteristic for the fundamental transverse mode propagation in the guide. For this mode the function has the main maximum value inside the main region and asymtoticaly decreases to zero in the excitation 6 and limitation 2 regions. The operation in the fundamental transverse mode is important both for applications wherein laser diode radiation is transmited through optical fibers and for those wherein it is transmited into atmosphere through collimating systems.

FIG. 3 presents a graph 13 of the dependence on the coordinate x of the refractive index n of the semiconductor materials constituting the epitaxial structure in a coordinate system (n,x). For simplicity, the refractive indexes are supposed constant in every region. In the following an algorithm for choosing the n₁...n₆ is presented. The active region 5 has a value n₅ of the refractive index which is the maximum value. The main region 3 has a value n₃ which is next inferior to the n₅ value. The wave propagation speed, v, along the z direction in the waveguide is determined by an effective index n_eff such that v=c/n_eff, where c is the light propagation speed in vacuum. A relation between the active region 5 thickness, d_a, and the refractive index difference (n₅- n₃) is imposed so that the propagation of any mode with an effective refractive index value n_eff between n₃ and n₅ values is not possible. In connection with the vacuum laser radi ation wavelength λ_o, the conditioning relation for the cut off of the modes with n₅<n_eff<n₃ is:

(3) or, approximately:

(4)

The intermediate region 4 has a refractive index value n₄ and the excitation region 6 a refractive index value n₆, values firstly determined by the energy band gaps of the semiconductor materials constituting the respective regions, energy band gaps which should be high enough compared to the energy band gap of the active region 5, to build effective potential barriers for the nonequilibrium carriers, electrons and holes, in their paths from the active region 5 toward the adjoining regions and to provide the carrier confinement effect.

The regions adjoined to the active region 5, namely the intermediate region 4 ana the excitation region 6 have conductvity types, n or p, opposite to each other, so that between them there is a p-n junction for the active region 5 excitation. The active region 5 has the conductivity type either n or p, or can consist of undoped semiconductor material, or can include the p-n junction.

If the semiconductor material of the main region 3 has the band gap close to the band gap of the active region 5, then the intermediate region 4 should be thick enough in order to avoid the nonequilibrium carrier tunneling. The thickness of an effective potential barrier dependes on the carrier effective mass and on the barrier height which in its turn depends on the difference between the intermediate region 4 and the active region 5 energy band gaps.

For example, in order to have a tunneling transmission coefficient through the intermediate region 4 barrier smaller than 10^-7, in the case of the carriers with the effective mass equal to 0,07 m_e and of a barrier height equal to 140 meV, the intermediate region thickness should be larger than 0,03 μm.

Greater the intermediate region 4 thickness, (x₄-x₃), greater is the radiation attenuation from the main maximum, situated in the main region, toward the active region and greater is the ratio r of the absolute maximum radiation flux densityI_max to the maximum radiation flux density in the active regionI_m.a., r=I_max/ I_m.a.. I n the mean time, toghether with a high r value, a high over unity ratio of the total radiation flux F to the active region flux f_a, F/f_a=1/

can be obtained. In this way, toghether with a great ratio between the main and active regions thicknesses, the intermediate region plays an essential role in obtaining small values for

The approximate relation between the total flux F and the maximum density of the radiation flux in the active region is:

F=k*I_m.a*_.d_a*w

where K is a numerical coefficient between 0.5 and 1. The valuek=0.5 can be used when the double natched area on FIG. 2 is approximated by a triangle area, if the radiation does not penetrate in the excitation region, and the value k=1 can be used if the radiation is extending in the excitation region, due to a relatively small difference between the n₅ and n₆ values.

The l_m.a. value is limited by catastrophic degradation phenomena. Informations about the maximum admissible value of I_m.a. can be obtained from laser diode wherein the radiation flux is almost completely confined in the active region (T =1). For the AlGaAs materials system and for continuous wave operation, I_m.a. values of the order of 10⁶ W/cm² are known. If I_m.a. value is limited, to obtain as high as possible values for F, as low as possible values for

are needed. Values for

from 1,5*10^-4 to

1.5*10^-3 are proposed in this invention.

The refractive index n₂ of the limiting region 2 must be sufficiently close to the n₃ value in order to avoid the propagation of other than the fundamental transverse mode.

The condition for higher order transverse mode cut off is expressed as a function of refraction indexes n₂, n₃, and n₅ and of the active region 5 and main region 3 thicknesses, d_a and d_m, by the relation:

(6 )

Using the previous relation (3), for the main region thickness d_m a condition is imposed:

(7)

The main region 3 thickness, d_m, is choosed from 2 to 5 μm, in relation to the desired width of the laser beam and to the desired total flux. For these thicrnesses, 2 μm and 5 μm, the differences between the retractive indexes n₃ and n₂ should be

7*10^-3, respectively 1.1*10^-3, for a 0.9 μm wavelength and

5.6*10^-3, respectively 0.9*10^-3, for a 0.8 μm wavelength. So small differences between refractive indexes can be reproducibly produced by free carrier concentration variation rather than by composition variation. The concentration variation should be

7*10^-17 cm^-3, respectively 1.1*10¹⁷ cm^-3, for the 0.9 μm wavelength and 5.6*10¹⁷ cm^-3, respectively 0.9*10¹⁷ cm^-3, for the

0.8 μm wavewlength.

When the semiconductor material of the active region 5 is excited by the current flow through the p-n junction, in the active region 5 nonequilibrium carriers, electron and holes, appear and they recombine by photon emission. The stimulated emission determines a gain coefficient of the active region, g_a.

The cut off condition for modes with n_{ef f} between n₃ and n₅ determines the spreading of the electromagnetic radiation outside the active region 5, mainly in the waveguide main region 3. Due to this spreading the wave in the guide has a modal gain coefficient 6 proportional to the intrinsec active region gain coefficient g_a and to :

G=g_a*

(8)

Due to the low confinement factor values, a very high

excitation of the active region 5 is needed in order to obtain a modal gain coefficient to exceed the losses. Values in the order of magnitude of 1000 cm^-1 for the active region gain coefficient g_a are proposed in this invention so that the active region 5 is highly excited by high concentrations of the nonequilibrium carriers, radiative recombination of which produces the stimulated emission.

At the semiconductor material-external medium interfaces 10, respectively 11, nonradiative surface recombination processes are produced, supplementary to the bulk, radiative ones so that there is a decrease toward the surface of the excitation level of the active region 5, expressed by the noneguilibrium carrier concentrations. This decrease determines the replacement of the wave amplification by its attenuation, due to absorbtion. The absorbtion generates nonequilibrium carriers in the vecinity of the semiconductor material-external medium interfaces 10 and 11, carri ers, which nonradiatively recombine on the surface and produce the initial local heating. The initial local heating is relatively small due to the separation, produced by the intermediate region 4, of the main region 3, wherein the almost entire laser radiation is nonabsorbely emitted at interfaces 10 and 11, from the active region 5, wherein, at interfaces 10 and 11, recombination and absorbtion processes occure. The local heating of the active region 5 is proportional to the radiation flux density in this region, which is small compared to the maximum radiation flux density. The local heating reaches the critical value for producing the catastrophic degradation at a critical value I_m.a.cr. of the maximum flux density in the active region 5, critical value which is approximately equal to the radiation flux density which produces the catastrophic degradation for a diode wherein the maximum of the flux density is in the active region.

The maximum flux which can be obtained before the appearance of catastrophic degradation can be derived from a formula analogous to (5):

F_cr=k*I_m.a.cr.*d_a*w/

(9) and smaller the

value greater the F_cr value.

The aim of the invention is to provide as great, as possible laser radiation total flux using as small as possible

values. Increasing in this way the limit for the catastrophic degradation obviously appears another limitation, namely that due to active region heating. The active region heating is produced by the dissipation of spontaneous recombination energy necessary to reach the threshold condition and by the Joule effect. The operation of the laser diode at relatively low current densities

J_s is needed in order to limit this heating. So, for a good electric to radiant energy conversion and for high radiation fluxes emission, low current density and high total current operation is needed, what implies the usage of a very long resonator length L. In this invention a resonator length between

1,5 and 3 cm is proposed, the length value being limited only by technological considerations.

When the resonator length increases the transmission losses decrease. Resonators with the back mirror reflectivity R₁=1, obtained by dielectric or dielectric-metals coatings, will be considered firstly. The other mirror reflectivity R₂ can be the natural one or, by the mentioned coating methods, can be varied between large limits, for example between 0.01 and 0.62. Mirror losses can be expresed by a loss coefficient β , given by the relation:

(10)

For example, for L=2.3 cm, R₁=1 and R₂=0.01 β =1 cm^-1, and for R₂=0.62 β=0.1 cm^-1.

There is known that greater than the wave attenuation coefficient α in the cavitiy is the mirror loss coefficient β , greater is the efficiency for the extraction of the stimulated emission. A good engineering value for their ratio is:

β/α =3 (11)

Free carrier absorbtion lasses in the active region will be considered as the main wave attenuation mechanism inside the waveguide. Other possible mechanisms are the attenuation due to free carrier absorbtion in the other waveguide regions and the attenuation due to the light scattering on inhomoqenitIes at waveguide interfaces or inside the waveguide.

Free carrier absorbtion in the active region 5 determines an intrinsec absorbtion coefficient α _a. Due to this absorbtion, the attenuation of the waveguide is characterised by an attenuation coefficient α . The relation between α and α _a is analogous to (8):

(12)

Considering the relation (11) the laser diode according to the invention should work at α values between 0.033 and 0.33 cm^-1. Such low values are possible only due to low

values and assuming that the free carri er absorbtion in the active region is the main loss mechanism.

With these proposed values for α (0,033-0.33 cm^-1) and for

(1.5*10^-4 -1.5*10^-3) one should consider that for the high excitation of the active region, generating free carri ers in conduction and valence bands. the intrinsec absorbtion coefficient in the active region should not exceed 8 safety value of approximately 200 cm^-1. This value should be related with free carri er concentration in the active region necessary to obtain a sufficient gain to surpass losses.

Total losses coefficient, due to mirror transmission and losses inside the cavity, α + β , has a value between 0.133 and

1.33 cm^-1. In stimulated emission conditions, total losses equal the modal wave gain 6 in the waveguide. So, G values are in the

0.133-1.33 cm^-1 interval and considering P values in the 1.5*10 - ⁴ -1.5*10^-3 interval, the intrinsec active region gain g_a=G/

should be approximately 890 cm^-1. Such a value for the active region gain can be obtainec if the volume current density J_vth of the threshold recombination current is in the order of magnitude of 3*10⁴ A/cm²/μm and if, correspondingly, free carrier concentration is 2-3*10¹⁸ cm^-3, what, corresponds to absorbtion coefficients of 40-60 cm^-1, so that the mentioned value 200 cm^-1 is indeed a safety value.

To obtain high efficiencies, the volume current density J_v should be few times greater than the volume current density at threshold, J_vth. The increase in the volume current density J_v is limited by the surface current density J_s, the value of which determines the active region heating, and by the active region thickness by the relation:

J_v=J_s/d_a (13)

J_s maximum admissible values for the continuous, respectively quasicontinuous (0.2 ms, 100 Hz pulses), regimes are

4*10³ A/cm², respectively 10⁴A/cm². Smaller are the d_a values, greater can be the J_s values. The radiation flux obtained in an active region of length L, width w and thickness d_a is:

F= (hv/e)*(J_s/d_a-J_vth)*L*w*d_a*β/(α+β) (14) where hv represents the emitted photon energy and e the electron charge.

The emitted flux should be to a safety factor s=2 smaller than the flux which produces catastrophic degradation and which is expressed by the formula (9). The optimum thickness of the active region can be determined from the formulae (9) and (14). Such optimum thicknesses of the active region are 0.045 μm and

0.11 μm for the continuous, rspectively quasicontinuous, regimes for laser diodes with F =1.5*10^-4. For P =1.5*10^-3 case, the optimum thicknesses of the active region are 0.11 μm, respectively 0.26 μm. For very short (100 ns) pulse regime, the active layer thickness can be greater, but can not exceed 0.32 μm, as will be further shown.

As shown by the relations (3) and (4) the active region thickness depends on the difference of refractive indexes n₅ and n₃. The values of the refractive indexes n₃ and n₅ are related to the corresponding energy band gap values of the semiconductor materials of the main region 3 and active region 5, and these values determine the emitted photon energies in the active region 5 and the band to band absorbtion coefficient for these photons in the main region 3. The n₃ and n₅ values and the corresponding band gaps should be so that the mentioned absorbtion coeficient is as small as possible, what means that the main region 3 is as transparent as possible for the radiation emitted in the active region. From this point of view it is prefered the main region to consists of an n type semiconductor and the energy gap of the main region 3 to be at least 120 meV greater than the energy of the emitted photons, to avoid band tail absorbtion. In the AlGaAs system to this energy difference corresponds a refractive index difference n₅-n₃ equal to 0.07 and a maximum active region thickness equal to 0.32 μm.

Also, the mam region 3 impurity doping should be kept to low concentration values to reduce free carrier losses and to avoid band tail absorbtion. Free carrier absorbtion in the main region should be smaller than 0.033 cm^-1, respectively 0.33 cm^-1, and the free carrier concentration should be smaller than 10¹⁶ cm^-3, respectively 10¹⁷ cm^-3, for laser diodes with

equal to

1.5*10^-4, respectively 1.5*10^-3. To reduce tree carrier absorbtion losses, the doping levels should be kept under 10¹⁷ cm^-3 in the intermediate region 4 and excitation region 6, too.

A first example of laser diode according to the invention is a structure in the AlGaAs system, with difterent values of the composition indexes and layer thicknesses. Al_yGa_{1- y}As system is choosed to illustrate the invention since the material constants are better known. All numerical examples above mentioned are related to this system, but other ternary and quaternary A_IIIB_V and A_IIB_VI compounds can as well be utilised using the same ideas: a structure with a separation layer between the active region and the main region of the waveguide, with a 2-5 μm wide main region, having the maximum of the radiation flux inside, with the radiation confined by a small difference between the limitation region and the main region refractive indexes, produced by a doping difference in the order of 10 cm^-3, with a radiation confinement factor in the order of 10^-4-2*10^-3 and with a long resonator length of 1.5-3 cm. Can be used ternary compounds: Al_xIn_{1- x}P, Al_xln_{1- x}As, Al_xGa_1-xSb, Al_xIn_{1- x}Sb, Ga_xIn_1- _xP, Ga_xIn_1-xAs, Ga_x In_1-xSb, GaP_xAs_1-x , GaAs_xSb_1-x, InP_xAs_1-x, InAs_xSb_{1- x}; and quaternary: Al_xGa_1-xAs_yP_1-y, Al _xGa_1-xAS_ySb_1-y, Ga_xIn_1-xP_yAs_1-y, Ga_xIn_1-xAs_ySb_1-y, (Al _xGa_1-x)_yIn_1-yP, (Al _xGa_1- _x ) y In_1-yAS, (Al _xGa _{1 - x} ) _y In_{1 -y}Sb , In (P_x As_1-x) _ySb_1-y.

In the first example, a laser diode with a 2.3 cm resonator length and 1, respectively 0.62, mirror reflectivities is considered. The refractive indexes and the layers thicknesses are presented in tab. 1:

The confinement factor ot this structure is 1.5*10^-4. Due to the relatively high active region thickness this structure is suitted for the quasicontinuous regime. The estimated total flux emitted without the appearance of the catastrophic degradation is

3.8 kW per mm of the laser diode width in the p-n junction plane.

The second example is a laser diode wherein the main, intermediate and active region thicknesses are modified, according to tab. 2:

The confinement factor of this structure is 1.5*10^-4, as in the previous example. The smaller active region thickness indicate this structure for the continuous regime. The total flux emitted without the appearance of the catastrophic degradation is

1.5 kW per mm of tne diode width in the p-n junction plane. The smaller than in the first example value of the emitted flux is related to a smaller value for the active region thickness.

The third example refers also to a laser diode with a 2.3 cm Iength but with 1, respectively 0.01, reflectivities. Such a diode can wort if the confinement factor is greater than in the previous cases, for examρle 1.5*10^-3. Refraictive indexes and the layer thicknesses are pesented in tab. 3:

This diode is indicated for the continous regime and the total flux emitted without the appearance of the catastrophic degrdration can be estimated to be 0.38 kW per mm of the diode width in the p-n junction plane. The smaller flux compared to that of the second example structure is related to the higher confinement factor

value.

The forth example refers to laser diode wherein the main, intermediate and active region thicknesses are modified comparatively to the previous example, according to the tab. 4:

This diode is indicated for the quasicontinuous regime. The flux emitted without the appearance of the catastrophic degradation is aproximately 0.95 kW per mm of the diode width in the p-n junction plane.

The active region, with thicknesses from 0.045 to 0.26 μm can be realised from succesive quantum wells, tor example quantum wells 7 nm wide separated by 3 nm wide barriers. If the wells with the mentioned width are made from GaAs and the barriers from Al_cGa_1-cAs with the composition index c=0.23, then the energy of the photons emitted from the active iaver is greater than the GaAs band gap and equal to 1.466 eV in the particular considered case. The active region refractive index has for this photon energy a mediated value approximately equal to 3.57. The band gap of the main region is 1.64 eV, 0.17 eV greater than the energy of the emitted photons. To this value of the photon energy corresponds a refractive index value equal to 3.53. On the basis of an active region consisting of quantum wells, the fifths example is constructed, described in tab. 5:

The confinement factor of this structure is 1.5*10^-3. Since the active region consists of quantum wells the diode will work with a reduced threshold current compared to that of the third example to which the diode has closed layer thickness and

coefficient values. The flux emitted without the appearance of the catastrophic degradation is 0.38 kW/mm as in the third example.

Structures without the intermediate region can be realised too in the frame of this invention. In such structures the band gap of the main region should be at least 200 meV greater than the active region band gap and the active region thickness should be smaller than 0.25 μm. When the active region thickness is only

0.035 μm the confinement factor is 3.5*10^-3 , respectively 1.1*10^-4, for structures with the main region thicknesses of 2, respectively 5 μm.

The laser diodes according to the invention present the advantage that they work at hign densities o f the radiation fluxes and also at high total flux, being constructed from simple structures with parallel between the two mirrors epitaxial layers.

Claims

WE CLAIM

1. A high power laser diode consisting of a substrate (1), a limitation region (2), a main region (3), an intermediate region (4), an active recuon (5), an excitation region (6) and a contact region (7), wherein the refractive index of the active region (5) is the highest in the structure, wherein the refractive index of the main region (3) is the next interior one, wherein the product of the active region (5) thuclness by the square root of the difference of the squares of the refractive indexes of active (5), respectively main regions (3) is smaller than a quarter of the vacuum wavelength of the laser radiation, wherein excitation (6) and intermediate (4) regions have conductivity types opposite to each other, in order to opperate at high radiation flux densities and to allow only the propagation of the fundamental transverse mode the maximum radiation flux ot which is located in the main region (3) and the effective refractive index of which is comprised between the retractive index values of the limitation ( 2 ) and of the main (3) regions, the product of the main region (3) thickness by the square root of the difference of the squares of refractive indexes of the main (3), respectively limitation (2) regions is smaller than half the vacuum wavelength of laser radiation, the difference between the refractive indexes ot the main (3) and limitation (2) regions is obtained onIs by different dopping concentrations of these regions, the difference in carrier concentrations being comprised between 7*10⁺¹⁷ and 0.9*10⁺¹⁷ cm^-3, correlated to the values for the main region (3) thickness between 2 and 5 μm, the confinement factor is comprised between 1.5*10^-4 and 1.5*10^-3 and the resonator lenth is comprised between 1.5 and 3 cm. region (3) has n type conductivity and has the energy band gap at least 0.12 eV greater than the energy of the photons emitted in the active region (5) and wherein the active region (5) thickness is maximum 0.32 μm.

3. A laser diode according to claims 1 and 2, wherein the main region (3) has a reduced doping level, under 10¹⁶ cm^-3 , respectively 10¹⁷ cm^-3, depending on the mentioned confinement factor values and the intermediate region (4) and the excitation region (6) have an impurity doping level under 10¹⁷ cm^-3.

4. A laser diode according to claims 1...3, wherein the active region consists of quantum wells.

(56) Bibliography: Great Britain Pattent 20311644 A.