BACKGROUND OF INVENTION
1. Field of the Invention:
The invention relates to a technique for the controlled transmission of radiant energy. More particularly, the invention is directed to an apparatus and systems in which the transmission of a flux of light or other radiant energy as it passes through an appropriate medium is controlled by a second flux of radiant energy, or the evanescent field thereof, such that transistor-like gain is achieved. Contemplated applications of the apparatus of this invention include fiber optics, optical integrated circuits, macroscopic optical devices for communications, sensing and control, optical computing, image processing, etc.
2. Description of the Prior Art:
A significant part of modern technology is concerned with apparatus and systems in which radiant energy is in some way altered during transmission through a medium. Communications systems, for example, may make use of a carrier wavelength of radiation energy which is somehow altered in analog or digital manner so as to represent intelligence. An ever-increasing part of technology depends upon communications systems of increasing sophistication which again generally depend upon a variation of some characteristic of energy with complex computations being the cumulative result of a multiplicity of such variations, possibly in discrete control elements. Such systems have, in the past, depended on relatively low frequency energy, perhaps DC, while later developments make increasing use of higher and higher frequency energy. This trend has gained impetus through the invention and development of the laser oscillator, variations of which may now produce CW or pulsed radiation at wavelengths from the far infrared through the visible spectrum and into the ultraviolet.
It is known to use induced absorption within a medium for one or more specified wavelengths of energy. Induced absorption permits operations such as those of switching and modulating, for example, for the interposition of variations representing information. U.S. Pat. No. 4,209,690 assigned to Bell Telephone Laboratories discloses an apparatus which can operate as an extremely rapid shutter, a switch, a modulator or a pulsed sharpener.
However, in the design of fiber or integrated optical circuits there remains a need for an optical amplifying system which can perform in a manner analogous to the transistor in an electronic circuit. While optical amplification based on photon multiplication via stimulated emission or parametric amplification is known, it is an object of this invention to disclose an optical amplifier which utilizes excited state absorption to produce a variable optical resistance or absorption. The result is a device characterized by transistor-like gain in switching capabilities.
It is another object of this invention to provide an optical transistor in which the photons from a control flux of radiant energy modulate the intensity of a signal flux of radiant energy in such a way that a change of one photon in the control flux causes a change of more than one photon in the signal flux.
It is yet another object of this invention to provide an optical transistor which utilizes a first and a second control flux of radiant energy such that a one photon increase in the first control flux causes a decrease of more than one photon in the signal flux, while a one photon increase in the second control flux causes an increase of more than one photon in the signal flux.
SUMMARY OF THE INVENTION
The invention provides a method for controlling the transmission of radiant energy through a light transmissive medium and an apparatus whereby the method can be effected. The method includes the steps of introducing a first flux of photons, or the evanescent field thereof, into said medium, said photons having energy E, and said medium having energy level characteristics such that the energy, E, of said first flux of photons matches an energy gap between an energy level designated as energy level 2, and another energy level designated as energy level 3 of said medium, and said first flux is substantially unattenuated as said first flux of photons, or the evanescent field thereof, passes through said medium; and selectively introducing a second flux of photons, or the evanescent field thereof, into said medium to attenuate said first flux of photons, said second flux of photons causing the population of energy level two of said medium, wherein a portion of said first flux is absorbed, incducing a population in level 3, a portion of which population in level 3 returns to level 2 and is available for further attenuation of said first flux, whereby the control of the transmission of said first flux of photons by said second flux of photons is more efficient than if the return of excited states from level 3 to level 2 did not occur.
BRIEF DESCRIPTION OF THE DRAWINGS
The above as well as other features and advantages of the present invention can be appreciated through consideration of the detailed description of the invention in conjunction with the several drawings in which:
FIGS. 1A and 1B are generalized energy level diagrams illustrating the mechanism by which the optical transistor according to this invention operates;
FIGS. 2A and 2B are generalized energy level diagrams illustrating the mechanism by which a first and a second control flux of radiant energy are used in an optical transistor according to this invention;
FIG. 3 is a simplified energy level diagram illustrating a complementary energy level scheme for a pair of associated optical transistors;
FIG. 4 is a schematic representation of an apparatus utilized in the demonstration of an optical transistor;
FIG. 5 is a cross-sectional view of an optical transistor incorporated into a fiber optic strand;
FIG. 6 is an embodiment of an optical transistor utilizing a fiber optic strand wherein the active transistor material is contained in a section of the fiber core;
FIG. 7 is an alternative embodiment of the optical transistor in which the active transistor material is contained in a section of the fiber optic cladding;
FIG. 8 is an alternative embodiment of the optical transistor in which the control flux is guided into an active cladding section by a separate fiber;
FIG. 9 is an alternative embodiment of the optical transistor in which the control flux source is independent of the active transistor material contained in the optical fiber;
FIG. 10 is an alternative embodiment of the active optical transistor in which the optical transistor material is incorporated into a thin film waveguide;
FIG. 11 is an alternative embodiment of a thin film optical transistor in which several signal fluxes are controlled by several control fluxes which can be selectively activated along the path of the signal fluxes;
FIG. 12 is an alternative embodiment of a thin film optical transistor according to this invention in which the signal and control fluxes are propagated in two different films; and
FIG. 13 illustrates an alternative embodiment in which two active transistor thin films are used.
DETAILED DESCRIPTION OF THE INVENTION
General Description
Basic Optical Transistor Mechanism
The invention, hereafter called the optical transistor, contains a suitable material, which may be crystalline, glass, liquid, vapor, organic or inorganic, such that the transmission of a flux of photons (hereafter called the signal flux) through the device is controlled by a second flux of photons (hereafter called the control flux or the evanescent field thereof). Furthermore, the interaction of these photon fluxes or their evanescent fields, with the material produces a "gain" effect such that a small number of control flux photons, n.sub.1, controls the transmission of a larger number of signal photons, n.sub.2. This control of many signal photons by few control photons is analogous to the modulation of large collector current by small base current in electrical transistor--hence the term optical transistor.
The mechanism by which the optical transistor operates requires a material in which induced absorption occurs among energy levels which have appropriate life-times, branching ratios and absorption cross sections. The generalized energy level diagram, FIG. 1, illustrates the process. For simplicity, only direct interactions of the photon fluxes with the material are shown, but interaction via the evanescent fields is also possible. Level 1 is the ground state and levels 2, 3 and 4 are not substantially populated at room temperature. The signal flux, S.sub.0, is matched to the energy difference between levels 2 and 3, therefore it normally passes through the material substantially unattenuated. The control flux, C.sub.0, populates level 4 which relaxes rapidly to a metastable state, level 2. Some decay from level 2 to level 1 occurs; however, the signal flux can now be absorbed and transfer excited states from level 2 to level 3. Some excited states in level 3 decay to level 2 and--if the transition rates into and out of level 2 are correct--are available to absorb more signal photons without the expenditure of additional control photons. If the induced transistion rate from level 2 to level 3 is much greater than the decay rate out of level 2, the introduction of n.sub.1 control photons causes the absorption of n.sub.2 signal photons with n.sub.2 >n.sub.1, and a gain effect is achieved.
Any of the discrete levels shown in FIG. 1 could be a multiplicity of levels or a continuum band. The downward transistions may be radiative or non-radiative in nature. Level 3 may lie at a higher energy than level 4, or they may be the same level, or level 4 may be removed with the control flux populating level 2 from level 1 directly. Also, more energy levels may exist in the system which are not directly involved in the transistor mechanism, or through which excited states may rapidly cascade in downward transistions. The gain of an optical transistor is defined as ##EQU1## the change in the final signal flux, dS.sub.f, for a given change in control flux, dC.sub.o. The absolute value is used since an increase in C.sub.o produces a decrease in S.sub.f. A simple steady-state rate-equation analysis of the energy level scheme described above yields ##EQU2## where S.sub.f is the final signal flux (photons/cm.sup.2 -sec), k is the fraction of the control flux which is absorbed in the transistor process, a.sub.31 is a materials parameter given by spontaneous decay rates in the system, .tau..sub.2.sup.-1 is the spontaneous decay rate of level 2, .sigma..sub.23 is the cross section for absorption from level 2.fwdarw.3, and .gamma.=1 for direct interaction of the flux with the material, and .gamma.<1 for interaction of the evanescent field, making (.gamma..sigma..sub.23) an effective cross section for absorption in the case of evanescent field interaction. The change in signal flux as it traverses the optical transistor, S.sub.o -S.sub.f, is given by
Controlled Population Mode
The response time of the device described in Section 2.1 is limited by .tau..sub.2, the lifetime of the metastable level--assumed to be the longest relevant lifetime in the system. This limitation can be overcome if a fifth level is available as shown in FIG. 2. Here, a second control flux, C.sub.o ' is matched to the energy gap between levels 2 and 5, and rapidly empties level 2 by stimulated emission.
"Complementary" Optical Transistors
The optical transistor concept described herein departs slightly from an electrical transistor analog in that the amplified output (final signal flux) is at a different wavelength than the controlling input (control flux). In the optical transistor one must deal with photons of two different wavelengths, whereas all the electrons in the collector and base currents of an electrical transistor are of the same kind. The problem can be overcome by using pairs of optical transistors made from materials with "complementary" energy level schemes as illustrated in FIG. 3. (In this Figure, level 4 is omitted for simplicity and the population of level 2 occurs directly from the ground state.) Here, the output signal flux from the first, S.sub.f is at wavelength .lambda..sup.1.sub.s. This output, S.sub.f (.lambda..sup.1.sub.s), becomes the control flux for the second transistor at wavelength .lambda..sub.c.sup.2, since .lambda..sub.s.sup.1 =.lambda..sub.c.sup.2. The complementary energy level scheme requires a signal in transistor 2 at the same wavelength as the control flux in transistor 1 so that, .lambda..sup.2.sub.s =.lambda..sup.1.sub.c also. Thus, the input and output of the two-transistor device are the same wavelength, .lambda..sup.1.sub.c. The significance of this concept is that a set of optical transistors paired in this manner can be used as a unit building block in more complicated analog and digital optical circuits with the photon output from one unit used directly as input for one or more other units. This optical unit is then a direct analog to the transistor used in electrical circuits.
Materials Requirements
The gain of the optical transistor is given by ##EQU3## From this expression we see that the maximum theoretical gain is obtained as k.fwdarw.1 (control flux fully absorbed) and .tau..sub.2 .fwdarw..infin., infinite lifetime in level 2. Thus, the maximum theoretical gain is
the ratio of the overall spontaneous decay rate from level 3, .tau..sub.3.sup.-1, to the direct decay rate from level 3.fwdarw.1, .tau..sub.31.sup.-1. (This development applies strictly to the energy level system described above, but the results are applicable in general with the obvious modifications for other energy levels and transition schemes included.) To optimize the maximum theoretical gain, .tau..sub.31.sup.-1 .fwdarw.O, or conversely, the return of excited states to level 2 rather than level 1 must be optimized. This is the primary materials requirement to maximize gain in an optical transistor.
The second term on the denominator, .tau..sub.2.sup.-1 /.gamma..sigma..sub.23 S.sub.f, will limit the intrinsic gain to values below the maximum theoretical gain. (In principle, one can arrange k.fwdarw.1 so that there is no fundamental material limitation introduced by the numerator of the expression for gain.) This term represents the ratio of spontaneous downward transitions out of level 2, .tau..sub.2.sup.-1, to induced upward transitions 2.fwdarw.3, (.gamma..sigma..sub.23 S.sub.f). As the upward transition rate becomes large compared with the decay rate, (.gamma..sigma..sub.23 S.sub.f)>>.tau..sub.2.sup.-1, the theoretical gain is approached.
Therefore, we require
and
for an optical transistor with an intrinsic gain
The foregoing describes the requirements for steady-state operation, i.e., modulation times much longer than any lifetime in the system. For a rapidly pulsed system, the pulse rate .tau..sub.p.sup.-1 is the limitting factor rather than the decay rate from level 2, therefore the relevant equation becomes .gamma..sigma..sub.23 S.sub.f >>.tau..sub.p.sup.-1. Also, the decay from level 3.fwdarw.2 must occur on a timescale much shorter than the pulse, .tau..sub.32.sup.-1 >>.tau..sub.p.sup.-1.
UO.sub.2.sup.2+ Embodiment
In this section we describe a specific embodiment of the optical transistor--a UO.sup.2+.sub.2 doped barium crown glass fiber--and set forth the device response for given system parameters. Demonstration of the optical transistor can be achieved with the system shown in FIG. 4.
The optical transistor itself is a 5 cm long fiber with a 10 .mu.m diameter core of UO.sup.2+.sub.2 doped barium crown glass. The glass composition is given in Table 1. This glass core is clad with Epo-tek 394 out of a total diameter of 30 .mu.m. An outer cladding of Epo-tek 394 and finely divided carbon powder (lampblack) makes the total diameter 125 .mu.m (FIG. 5). The outer absorbing cladding acts as a mode stripper by removing any modes propagating in the cladding instead of in the core. The fiber is mounted in an Amphenol 905 series fiber optic connector matched to the diameter of the fiber.
A signal flux at 530 mm is supplied by a doubled Nd:YAG laser and a control flux at 460 nm is supplied by a dye laser. These fluxes are focused independently by lenses on micrometer mountings before they encounter the beam splitter. This facilitates very accurate focusing of the portion of each beam which strikes the fiber, insuring that these incident beams are co-linear and overlap completely as they enter the fiber core. The output fluxes are detected by a conventional Si diode detector. When both fluxes are present during operation, the transmitted signal flux, S.sub.f, can be detected separately from any transmitted control flux, C.sub.f, by the use of a narrow band filter.
The steady-state gain of the UO.sub.2+.sub.2 optical transistor can be demonstrated with the system described above. Since gain is defined
the most straightforward way to measure it, in principle, is to measure the final signal with no control flux, S.sub.f, and then with the control flux present, S' so that ##EQU4## S.sub.f and S.sub.f ' are easily measured, however, the value of C.sub.o, the control flux which is launched into the fiber core and participates in the transistor process, is difficult to obtain. A small percent of any beam incident on the fiber will be reflected at the fiber surface, and some small portion may also be launched into the cladding instead of into the core. (The absorbing outer cladding prevents these modes from propagating to the detector.) Another small portion of the flux that is successfully launched into the fiber core may also be scattered or absorbed by defects in the fiber. These loss mechanisms cannot be measured directly, therefore an accurate value of C.sub.o is difficult to obtain. However, the final control flux, C.sub.f, is easily measured and can be related to the intrinsic gain by ##EQU5## where k is the fraction of the control beam that is absorbed. For the appropriate UO.sub.2+.sub.2 absorption cross section and the specified doping concentration and length of fiber, 99% of the control flux will be absorbed. Thus, for the specific UO.sub.2.sup.+2 optical transistor described in this section, ##EQU6##
S.sub.f ' and C.sub.f are not independent, but are related by the equations which describe the transistor action in the section entitled Basic Optical Transistor Mechanism. For example, an effective control flux, C.sub.o =6.81 .mu.W is required to produce a 2% modulation of a 0.751 mW signal. Under these conditions, one would measure S.sub.f =0.751 mW, S'.sub.f =0.736 mW, C.sub.f =0.068 .mu.W, yielding a gain of 2.2. Therefore, for realistic laser powers and photon densities in the fiber, an intrinsic gain>2 can be obtained with a UO.sub.2.sup.2+ optical transistor.
The scheme described above is a demonstration of the optical transistor effect in UO.sub.2.sup.+2 which has a rather low intrinsic gain, and operates at short wavelengths. Other materials with suitable properties will yield gain in the presence of extraneous losses in fiber optic or integrated optic systems and will operate at wavelengths supplied by convenient laser diode sources.
Other Embodiments
Several fiber optic embodiments of the optical transistor are given in FIGS. 6-9. In FIG. 6 the active transistor material is contained in a fiber core and the signal and control fluxes are propagated into the active region along the same fiber. FIG. 7 shows the active material comprising a section of fiber cladding. The energy of the two fluxes is propagated in the core but their evanescent fields extend into the cladding and activation of the cladding by the control flux will affect the transmission of the signal flux. FIG. 8 shows the control flux guided into the active cladding by a separate fiber. The control flux need not be a guided flux as FIG. 9 illustrates. Here the control flux is supplied by a lamp which activates a section of fiber core; alternatively, the same arrangement could activate a fiber cladding. FIGS. 6-9 illustrate fiber optic embodiments in which the active transistor material is comprised of solid materials. However, a capillary tube of liquid or gas could be substituted for the active fiber regions of FIGS. 6 and 9. The fluxes are not totally internally reflected in a capillary as they are in a clad fiber but capillary tubes guide light effectively for grazing incidence.
Optical transistor materials can also be incorporated into thin film waveguides. FIG. 10 illustrates a case in which the active material comprises the thin film. The example shows the signal flux edge coupled and the control flux prism coupled to the film, however, the fluxes may be interchanged in the drawing or both may be coupled in the same manner either separately or together such that the fluxes are coincident at the point of entry into the film. Also, the active transistor material may be restricted to a strip of thin film through which the fluxes propagate, or it may be in the base material and interact through the evanescent fields of the guided fluxes.
The planar area afforded by thin film embodiments may also be utilized by allowing several signal and control fluxes to interact on perpendicular paths as shown in FIG. 11 (only edge coupling is illustrated for simplicity). Now the signal fluxes are controlled not only by the magnitude and changes of a single control flux, but also by a number of control fluxes which can be selectively activated along the path of the signal fluxes. FIG. 11 illustrates the active transistor material comprising the thin film waveguide; again the waveguide itself may be passive and the active material may be in the base.
Also, the signal and control fluxes may be propagated in two different films, one the active transistor material and one passive. FIG. 12 illustrates this case for perpendicular groups of signal and control fluxes, however, it applies just as readily to co-linear cases such as FIG. 10. In FIG. 12 the signal fluxes propagate in the active film and the control fluxes in the passive film, however, this could be reversed. The relative position of the two films could also be reversed providing the upper film has the lower index of refraction.
The propagation of fluxes in more than one film provides a method for stacking optical transistors in the vertical dimension. FIG. 13 illustrates a film embodiment in which two active transistor films are used. In the lower film, control and signal fluxes C.sub.1 and S.sub.1 propagate. The transmission of S.sub.2 in the upper film is controlled by the evanescent field of S.sub.1. Therefore, S.sub.1 becomes the control flux for S.sub.2. This arrangement presents the possibility of device input, C.sub.1, and output, S.sub.2, being at the same wavelength which may be desirable for some applications.