WO1994028605A1

WO1994028605A1 - Improved frequency multiplier and method of producing the same

Info

Publication number: WO1994028605A1
Application number: PCT/US1994/006116
Authority: WO
Inventors: David Eger; Moshe Oron; Avigdor Zussman; Mordechai Katz; Adolf Shahna
Original assignee: Isorad U.S.A., Inc.; Soreq Nuclear Research Center
Priority date: 1993-05-31
Filing date: 1994-05-31
Publication date: 1994-12-08
Also published as: IL105863A

Abstract

A frequency-converted laser apparatus including a laser source (10) which generates primary radiation, a Bragg reflector (24) which reflects part of the primary radiation within a frequency range centered at a preselected fundamental frequency back along an optical path to the laser source (10) thereby to form an external laser cavity including the laser source (10) resonant at the preselected fundamental frequency, and a frequency converter (22) located within the external laser cavity on the optical path between the laser-source (10) and the Bragg reflector (24) and operative to convert part of the primary radiation into frequency-converted secondary radiation. There is also provided a method of forming on the surface of a crystal a frequency converting optical waveguide including a plurality of periodic segments comprising sections of a first kind, having a first electric polarity, and sections of a second kind, having a second electric polarity, including making a plurality of discrete regions on the surface of the crystal according to a preselected masking pattern, preparing an ion exchanging molten salt solution, controlling the concentration of oxides in the molten salt solution, immersing the masked surface ot the crystal in the molten salt solution for a preselected period of time and removing the masking material from the surface of the crystal.

Description

IMPROVED FREQUENCY MULTIPLIER AND METHOD OF PRODUCING THE SAME

The present invention relates to laser frequency conversion in general and, more particularly, to fre- quency multipliers for use with diode lasers.

There has been much effort directed toward the development of non-linear optical devices for laser frequency conversion. Particularly, attempts have been made to construct compact laser devices using diode lasers and frequency-converting wave-guides. In such devices, a fundamental frequency beam is emitted by a laser diode and guided through an optical non-linear wave guide in which the fundamental frequency is dou- bled, thereby generating a frequency-converted beam. Normally, the fundamental frequency is in the infrared range and the second-harmonic frequency is in the visible range, thereby producing a useful output of visible laser light. The aim of such frequency-doubling system is to construct compact laser sources which may be useful, for example, in medical and metrological instruments and for optical recording. In order to obtain reasonable results from fre- quency multipliers using optical non-linear devices, the phase velocities of the fundamental wave and the second-harmonic wave must be perfectly matched. Unfor- tunately, in most mediums the phase velocities of the two waves are considerably different. Somech et. al. in Applied Physics Lett. 21, p. 140, (1972), suggest using quasi-phase matching (QPM) which is based on periodi- cally changing the refractive index and the non-linear coefficient of the wave-guide. For efficient multipli- cation of the fundamental wave, using QPM, the wave- length of the fundamental wave must be within a narrow range, of no more than 1 or 2 A, centered at a wave- length which satisfies the specific QPM condition for matching of the fundamental and second harmonic fre- quency. Consequently, it is extremely difficult to match between the fundamental frequency generated by the diode laser and the frequency which the QPM wave- guide is adapted to double. Actual QPM wave-guide frequency converters have been constructed using raate- rials such as, for example, LiTa0₃ and LiNb0₃. U.S. Patent 5,028,107 to Bierlein et. al. de¬ scribes a frequency converter which includes periodi- cally segmented wave-guides formed of a KTiOP0₄ (KTP) crystal or the like. To form the wave guide pattern, first, a masking material is applied to a smoothed surface of the KTP crystal, providing a pattern of aligned regions along the surface. Then, a preselected amount of a molten salt containing cations such as Rb⁺ or the like is applied to the smoothed, partially masked, surface of the KTP crystal for a preselected time period and at a preselected temperature, resulting in a preselected amount of cation replacement at the unmasked regions of the surface. The cation replacement at the unmasked regions changes the index of refraction of a portion of the crystal under the unmasked regions. Finally, the masking material is removed and the edges of the crystal are polished to provide clean input and output faces. W.P. Risk et. al. and G.L. Bona et. al., in "Generation of 425nm Light by Waveguide Frequency Doubling of a GaAlAs Laser Diode in an Extended Cavity Configuration", a paper presented at "Compact Blue- Green Lasers", a Topical Meeting, on Feb. 2-4, 1993, and published in OSA Technical Digest Series, Vol. 22, P. 489, describes a method of coupling a diode laser to a QPM frequency converter using an external grating. According to this method, unconverted infrared light emanating from the output face of a wave guide is reflected by an external grating and fed back to the laser diode through the wave-guide. Due to this feed- back, the frequency generated by the laser is stabi- lized at a preselected level, controlled by the angle of the grating which is used for selecting the reflect- ed frequency. Unfortunately, such a system is not useful for most applications of compact visible-light lasers because it is complex and expensive and, espe- cially, because it is extremely sensitive to changes in ambient conditions. U.S. Patent 5,185,752 to Welch describes an ar¬ rangement for coupling a diode laser to a frequency doubling wave-guide. This Patent suggests the use of a fixed distributed reflective grating which is formed at the input face of the wave-guide. The grating reflects infrared radiation of a preselected fundamental fre- quency, determined by the spacing of the grating, back into the laser cavity, thereby stabilizing the output of the diode laser at the preselected frequency. The stabilized fundamental frequency is then doubled by a nonlinear wave-guide, downstream of the grating, in the manner described above. The proposed frequency-doubler of Patent '752 poses a number of problems. One problem is that the reflected portion of the fundamental wave is not uti- lized by the frequency doubler. It should be appreciat- ed that the infrared output power of the diode laser is reduced, due to loss of the reflected radiation, by a factor of (1-R)², wherein R is the reflectivity of the grating. For practical purposes, reflectivity R is preferably equal to at least 0.2 and, therefore, since the converted output power is substantially proportion- al to the square of the input power, the conversion efficiency of such a device will be reduced to no more than 64% of the original efficiency of the doubler wave-guide. The '752 Patent does not provide appropriate means for fixing or adjusting the wavelength of the reflected wave with the accuracy required for frequency doubling. A The method disclosed by Patent '752 for creating a first or second order grating on a substrate includes holographical exposure of the surface of the substrate through a selective mask, followed by etching of the exposed portions of the surface at the input face of the wave-guide. It should be appreciated that such a method has not, to date, provided the accurate wave- length matching, on the order of 1 - 2 A at 10,000 A wavelength, required for frequency doubling. Shinozaki et. al., in Applied Physics Lett. 59, p. 512, (1991), proposed a compact device including a periodically segmented waveguide, which provides fre- quency doubling and laser diode frequency locking. The periodically segmented waveguide is designed to reflect part of the input radiation and to convert the frequen- cy of some of the unreflected input radiation. The reflection from the wave-guide is used for stabilizing the frequency generated by the laser diode. This ar- rangement requires the use of a wave-guide which meets both the QPM, i.e. phase matching, and Bragg, i.e. frequency matching, conditions described above. In order to meet the Bragg condition, a considerable fraction of the fundamental wave is reflected and is not used by the frequency doubler. The energy carried by the wave decays exponentially as a function of distance traveled, within the boundaries of the wave- guide and, therefore, the power actually used for doubling and the power at the doubled frequency are thus reduced. The requirement that both the Bragg and QPM condi- tions should be met simultaneously imposes a substan- tial limitation on the choice of fundamental frequency. Furthermore, after such a frequency has been chosen, there are considerable technical limitations to the actual construction of a wave-guide adapted for the fundamental frequency. Firstly, the standard photolith- ographic methods used for producing segmented QPM wave- guides limit one to a set of discrete dimensions for the sections of the wave-guide. For example at 0.1 μm addressing, which is typically the case, all the sec- tions of the wave-guide must be integer multiples of 0.1 μm. Secondly, existing techniques have inherent inaccuracies which result in discrepancies between prescribed, design, values and actual, produced, values of the period length. Thus, according to existing techniques, it is very difficult to construct an actual wave-guide which is adapted for a prescribed frequency and meets both of the above mentioned conditions. Roelofs et. al., in "KTP Segmented Waveguides as Concurrent Bragg Reflectors and Second Harmonic Genera- tors", a paper presented at "Compact Blue-Green Lasers", a Topical Meeting, on Feb. 2-4, 1993, and published in OSA Technical Digest, Series 2, P. 485, proposed an improvement to the apparatus described in the preceding paragraphs, including the use of superpe- riod structures in the construction of the wave guide. The use of superstructures and temperature tuning allows some adjustment of the frequency reflected by the wave-guide, but it does not completely remove the limitation imposed on the choice of frequency. Another disadvantage of the apparatus proposed by Roelofs et. al., for which their disclosure suggests no solution, is the reflection of the second harmonic frequency, back towards the diode laser. It is appreci- ated that, in principle, when the QPM and Bragg reflec- tion conditions are both met simultaneously for the fundamental frequency, the Bragg condition is also met for the second harmonic frequency. Thus, a substantial amount of second harmonic radiation is reflected back towards the laser diode, thereby reducing the frequency doubling efficiency of the device. Optical waveguide structures for the devices described above and similar devices, for example the QPM frequency converter described in U.S. Patent 5,157,754 to Bierlein et al., are produced through a process which includes an ion exchange process. In patent '754 as in the other prior art patents, a molten salt solution is applied to the surface of a KTP wafer which is coated with a metallic, Ti, mask. The solution contains one of the monovalent ions, Rb⁺, Tl⁺ or Cs⁺, and one of the divalent ions, Ba⁺⁺, Ca⁺⁺ or Sr⁺⁺. Unfortunately, it has been found that, in prac- tice, such ion exchange processes rarely yield the desired optical waveguides, particularly frequency doubling waveguides. Typically, the resultant wave- guides feature either very poor transmission at the fundamental frequency or very poor frequency conversion efficiency, i.e. very poor generation of second harmon- ic radiation, or both. Furthermore, even in rare cases where reasonable frequency doubling waveguides are obtained, such reasonable results are non-reproducible.

It is, therefore, an object of the present inven- tion to provide an improved frequency-converted laser device. It is a further object of one aspect of the present invention, to provide an improved method and apparatus for producing frequency-converted laser devices. In accordance with one, preferred, aspect of the present invention there is thus provided a frequency- converted laser apparatus including a primary laser oscillator, operative to emit primary radiation, and a Bragg reflector. The Bragg reflector is operative to reflect part of the primary radiation, within a limited frequency range centered at a preselected fundamental frequency, back along an optical path into the primary oscillator. The apparatus also includes a frequency- converter, located on the optical path between the primary laser oscillator and the Bragg reflector, operative to convert part of the radiation at the fundamental frequency into frequency-converted second- ary radiation. This arrangement creates an external laser cavity, including the primary laser oscillator and the frequen- cy converter, which resonates at the preselected funda- mental frequency. In a preferred embodiment of the invention, the primary laser oscillator includes a reflective back facet and, thereby, the external laser cavity extends from the reflective back facet to the Bragg reflector. In accordance with a preferred embodiment of the present invention, the frequency converter and the Bragg reflector are both formed on a crystalline mate- rial and, preferably, both are formed on a single crystal. The crystalline material preferably includes a KTP crystal or other crystals of the same family such as KTA, RTP and CsTP, or any combinations of such crystals. Alternatively, other crystalline materials may be used, for example the crystals LiNb0₃, K θ3 and LiTa0₃ . 8 Further in accordance with a preferred embodiment of the invention, the fundamental frequency is in the infrared range and, more preferably, the primary oscil- lator is a diode laser emitting infrared radiation. The frequency of the secondary radiation is preferably higher than the fundamental frequency and, more prefer- ably, it is in the visible range. In a particularly preferred embodiment of the invention, the frequency of the secondary radiation is a second harmonic frequency equal to twice the fundamental frequency. In a preferred embodiment of the present inven- tion, the frequency converter includes a plurality of periodic segments including sections of a first kind, having a first index of refraction, and sections of a second kind, having a second index of refraction. Further, in a preferred embodiment, the sections of the first kind have an electric polarity opposite the electric polarity of the sections of the second kind. In a particularly preferred embodiment of the inven- tion, each periodic segment is a superperiod L-^ includ- ing k sub-periods and each sub-period includes one section of the first kind followed by one section of the second kind. Similarly, the Bragg reflector preferably includes a plurality of periodic reflecting segments including reflecting sections of a first kind, having a first index of refraction, and reflecting sections of a second kind, having a second index of refraction. In a particularly preferred embodiment of the invention, each periodic reflecting segment is a reflecting su- perperiod Q_j including j reflecting sub-periods and each reflecting sub-period includes one reflecting section of the first kind followed by one reflecting section of the second kind. The frequency converter preferably includes an input face for receiving the primary radiation. Accord- ing to this embodiment of the invention, at least one lens is located between the primary oscillator and the frequency converter for focusing the primary radiation onto the input face of the frequency converter. In accordance with another, preferred, aspect of the present invention, there is provided an integrated wave-guide unit including a frequency converting por- tion, adapted to receive primary radiation and, upon receipt of the primary radiation, to convert part of the primary radiation into frequency-converted second- ary radiation, and a Bragg reflector portion, adapted to receive radiation from the frequency-converting portion and to reflect part of the primary radiation, within a limited frequency range centered at a prese- lected fundamental frequency, back through the fre- quency converting portion. In a preferred embodiment of the invention, the frequency converting portion and the Bragg reflector portion are both formed on a single chip of crystalline material. Preferably, the single chip of crystalline material is a chip of KTP crystal. The integrated unit is preferably adapted to receive primary radiation in the infrared range which is, more preferably, generated by a diode laser. The frequency-converted secondary radiation is preferably in the visible range. In a preferred embodiment, the frequency of the secondary radiation is higher than the fundamental frequency and, more preferably, the fre- quency of the secondary radiation is a second harmonic frequency equal to twice the fundamental frequency. In a accordance with a preferred embodiment of the invention, the integrated unit includes a plurality of periodic segments including sections of a first kind, having a first index of refraction, and sections of a second kind, having a second index of refraction. The first kind of sections have an electric polarity oppo- site the electric polarity of the second kind of sec- tions. In a particularly preferred embodiment, each periodic segment is a superperiod L^→^ including k sub- periods wherein each sub-period includes one section of the first kind followed by one section of the second kind. Similarly, the Bragg reflector portion preferably includes a plurality of periodic reflecting segments including reflecting sections of a first kind, having a first index of refraction, and reflecting sections of a second kind, having a second index of refraction. Preferably, each periodic reflecting segment is a superperiod Q_j including j reflecting sub-periods, wherein each reflecting sub-period includes one re- fleeting section of the first kind followed by one reflecting section of the second kind. According to another, preferred, aspect of the present invention, there is provided a frequency-con- verted laser apparatus including a primary laser oscil- lator having a reflective back facet and a front facet, operative to emit primary radiation through the front facet, and an integrated wave-guide unit as described in the previous paragraphs. The integrated wave-guide unit preferably includes an input face, optically associated with the front facet of the primary oscilla- tor, for receiving the primary radiation. According to this aspect of the invention, an external laser cavity is formed between the reflective back facet of primary oscillator and the Bragg reflector portion of the integrated wave-guide. The external laser cavity reso- nates at the fundamental frequency. The apparatus preferably includes at least one lens, located between the front facet of the primary oscillator and the input face of the integrated wave- guide unit, for focusing the primary radiation onto the input face of the integrated wave-guide. Alternatively, no lenses are provided between the front facet of the primary oscillator and the input face of the integrated waveguide and, in such a case, the primary oscillator and the integrated waveguide are mounted very close together such that the front facet and the input face are only a few micrometers apart. Since the primary radiation is typically generated by the primary laser source as an elliptical beam, a preferred embodiment of the invention includes at least one optical element which shapes the frequency-convert- ed secondary radiation to produce a substantially circular output beam. The at least one optical element may include at least one cylindrical lens and/or at least one anamorphic prism. In a preferred embodiment of the invention, the apparatus further includes a thermal controller, asso- ciated with a surface of the integrated waveguide unit, which controls the temperature of the integrated wave- guide unit. Preferably, the apparatus further includes an output radiation detector which provides a feedback responsive to the intensity of the frequency-converted secondary radiation, wherein the thermal controller receives the feedback from the output radiation detec- tor and controls the temperature of the integrated waveguide unit in accordance with the feedback. Additionally or alternatively, the apparatus includes a laser control unit which receives the feed- back from the output radiation detector and controls the electric power provided to the laser source and/or the temperature of the laser source in accordance with the feedback. The present inventors have found that the perform- ance of frequency doubling waveguides is critically dependent on the concentration of oxides in the molten salt solutions which are used for stimulating the ion exchange process during production of the waveguides. It has been found that to obtain an efficient frequency doubler, the oxide concentration, [0 ], must lie within a narrow range, c_a≤[0²~]≤C_jD, wherein c_a and c^ are at least partially dependent on the type of KTP crystal used. The acceptable concentration range is typically very narrow, e.g. the difference between c^ and c_a may be on the order of 10 - 40 ppm. When [0²~ ]<c_a, the resultant waveguides show very poor frequency conversion efficiency. When [0²~]>c_b, both the frequen- cy conversion efficiency and the optical transmission efficiency of the resultant waveguides are poor. Therefore, according to another aspect of the present invention, there is provided a system and a method for controlling the [0²~] concentration of a molten salt solution which is subsequently applied to the surface of a crystal for ion exchange thereon. According to this aspect of the present invention, the ion exchanging molten salt solution is supplied with donors, i.e. additives which increase the concentration of oxides in the solution, and/or acceptors, i.e. additives which reduces the concentration of oxides in the solution, as necessary for maintaining the concen- tration of oxides within the range c_a≤[0* l≤C_jj. Exposure of the molten salt solution to active gasses and moisture in the air may have an uncontrolla- ble effect on the oxide concentration in the solution. In a preferred embodiment of the present invention, the molten salt solution is kept in a controlled environ- ment, preferably an inert gas environment, which raini- mizes undesirable interaction between the solution and the external environment. It has been found that when the solution is originally prepared using pure com- pounds, the oxide concentration in the solution is generally lower than the required minimum, c_a. This lower concentration, in the controlled environment, is then preferably increased to a desired concentration, within the range c_a≤[0²~]≤c_jj, by adding a preselected amount of donor solution to the molten salt solution. After the desired oxide concentration is reached, the controlled environment ensures that the desired concen- tration is maintained, allowing only minor fluctua- tions. If the oxide concentration is higher than C_jj, a preselected amount of acceptor compound is added to the molten salt solution. There is thus provided, in accordance with a preferred embodiment of the present invention, a method of forming on the surface of a crystal a frequency converting optical waveguide including a plurality of periodic segments comprising sections of a first kind, having a first electric polarity, and sections of a second kind, having a second electric polarity, includ- ing: selectively masking the surface of the crystal in accordance with a preselected masking pattern; preparing an ion exchanging molten salt solution; maintaining the concentration of oxides in the molten salt solution within a preselected, narrow range; immersing the selectively masked surface of the crystal in the molten salt solution for a preselected period of time; and removing the masking material from the surface of the crystal. In a preferred embodiment of the invention, the ion exchanging molten salt solution is maintained in a substantially inert environment while the masked sur- face is being immersed in the solution. The present inventors have also found that the performance of a waveguide depends on the ion exchange depth profile of the surface of the processed KTP crystal, where a waveguide has been formed. Thus, according to the present invention, the performance of the waveguide can be evaluated without actually using the waveguide, i.e. without coupling the waveguide to a laser diode, by analyzing the ion exchange depth pro- file of a preselected area of a given surface of the crystal. In a preferred embodiment of the present inven- tion, the expected performance of a waveguide formed on a given KTP crystal is evaluated by X-ray diffraction pattern analysis, preferably using an X-ray diffractom- eter operating in a double crystal rocking curve (DCRC) measurement mode. A processed, i.e. ion exchanged, surface of the given KTP crystal is scanned with an X- ray at a preselected wavelength and the resultant diffraction pattern is compared with a reference pat- tern which corresponds to the ion exchanged portions of a functionable waveguide. For convenience, the measure- ments are performed on the unpatterned Z⁺ surface (i.e the back surface) of the given crystal, rather than on the Z~ surface (the front surface) where the waveguide pattern is formed, since the Z⁺ provides a homogeneous simulation of the ion exchanged portions of the Z~ surface. Then the X-ray diffraction pattern of the waveguide does not sufficiently match the reference pattern, it is further possible according to the present invention to determine whether the concentration of oxides in the solution used for preparing the waveguide is too high, i.e. higher than the above mentioned maximum, c^, or too low, i.e. lower than the above mentioned minimum, c_a. It has been found that the X-ray diffraction pat- terns of crystals processed with an excessive concen- tration of oxides are distinctly different from the X- ray diffraction patterns of crystals processed with a deficient concentration of oxides. There is thus also provided, in accordance with a preferred embodiment of the invention, a method for evaluating whether a frequency converting optical waveguide formed on a surface of a crystal will effi- ciently convert an input frequency to an output fre- quency, comprising: irradiating an area of a surface of the crystal with X-ray radiation; measuring the magnitude of X-ray radiation dif- fracted by said area across a preselected range of angles to obtain a diffraction pattern of said area; and comparing the diffraction pattern of said area to a predetermined X-ray diffraction pattern indicative of efficient frequency conversion. It should be appreciated that although the above described X-ray diffraction analysis method is pre- ferred, the present inventors have found several other methods which are equally suitable for evaluating the functionability of waveguides. For example. Electron Microprobe Analysis, Auger Electron Spectroscopy, Secondary Ion Mass Spectrometry and optical dispersion measurements, have all been found suitable for deter- mining the ion exchange profile of the surface of the processed KTP crystal.

The present invention will be better understood from the detailed description of the preferred erabodi- ments of the invention taken in conjunction with the following drawings, of which: Fig. 1 is a schematic illustration of a frequency- converted laser apparatus, constructed in accordance with a preferred embodiment of the present invention; Fig 2A is a schematic illustration of an integrat- ed wave-guide unit useful in the operation of the frequency-converted laser apparatus of Fig. 1, con- structed in accordance with one, preferred, embodiment of the invention; Fig. 2B is a schematic illustration of an inte- grated wave-guide unit useful in the operation of the frequency-converted laser apparatus of Fig. 1, con- structed in accordance with another, preferred, erabodi- ment of the invention; Fig. 3 is a schematic illustration of a portion of a system for forming a frequency doubling optical waveguide on a crystal; Fig. 4A illustrates an X-ray diffraction pattern typical of an inoperable frequency doubling waveguide of a first type; Figs. 4B and 4C show linear and logarithmic rubid- ium density profiles, respectively, calculated from the pattern of Fig. 4A; Fig. 5A illustrates an X-ray diffraction pattern typical of an operable frequency doubling waveguide; Figs. 5B and 5C show linear and logarithmic rubid- ium density profiles, respectively, calculated from the pattern of Fig. 5A; Fig. 6A illustrates an X-ray diffraction pattern typical of an inoperable frequency doubling waveguide of a second type; and Figs. 6B and 6C show linear and logarithmic rubid- ium density profiles, respectively, calculated from the pattern of Fig. 6A. Reference is now made to Fig. 1, which illustrates a frequency-converted laser apparatus in accordance with a preferred embodiment of the present invention. A diode laser 10, including a back facet 12 and a front facet 14, emits primary radiation, preferably in the infrared range of 800nra - lOOOnm. In accordance with a preferred embodiment of the present invention, back facet 12 is highly reflective while the reflectivity of front facet 14 is relatively low. The radiation emitted by diode 10 is focused, preferably by a system of converging lenses such as lenses 16 and 18 shown in Fig. 1, onto the input face 20 of an integrated unit 36. Unit 36 is preferably integrally formed on a single substrate, such as a crystal, as described in more detail below with refer- ence to Figs. 2A and 2B. In a preferred embodiment of the invention, the emitted radiation is focused only onto a small portion of face 20 near the upper surface 40 of integrated unit 36. Therefore, the input radia- tion is received and guided only through the uppermost layer of unit 36, hereinafter referred to as integrated wave-guide 21. It should be appreciated that in reality layer 21 is very thin, typically approximately 5 mi- crometers, and that layer 21 is shown disproportionally thick in Fig. 1 for illustrative purposes only. According to the present invention, the radiation entering integrated wave-guide 21 through input face 20 is led, first, through a frequency converting portion 22 and, then, through a distributed Bragg reflector portion 24. Frequency converting portion 22 converts part of the primary radiation to radiation at a higher frequency, preferably in the visible range, hereinafter referred to as the converted-frequency radiation. According to one, preferred, embodiment of the inven- tion, the converted frequency is equal to exactly twice the fundamental frequency and, in such case, the con- verted-frequency will be referred to as second-harmonic frequency. ¹⁸ In a preferred embodiment, the generated convert- ed-frequency radiation is not influenced by Bragg reflector portion 24 and, therefore, it freely exits wave-guide 21 through a transmissive output face 26. The unconverted part of the primary radiation, on the other hand, is reflected by Bragg reflector 24, through portion 22, back into diode laser 10. Preferred embodi- ments of integrated unit 36, including preferred struc- tures of integrated wave-guide 21, will be described in detail below with reference to Figs. 2A and 2B. When the arrangement described above is used, an external cavity is formed between back facet 12 of diode 10 and distributed Bragg reflector 24. In a preferred embodiment, distributed Bragg reflector 24 is designed to be highly reflective in response to a preselected frequency "fg" hereinafter referred to as the fundamental frequency. Preferably, fundamental frequency "fo" is chosen such that a resonant condi- tion, i.e. a QPM condition for frequency conversion, is met. The external cavity resonates at the chosen funda- mental frequency "f₀" and, if frequency "f₀" is proper- ly adapted to the frequency input requirement of fre- quency converter 22, a converted, stable, output in the visible range is provided through output face 26. More efficient results are achieved when a thermal controller 34, such as a controlled thermoelectric element or any other suitable thermal control means, is provided to control the temperature of integrated wave- guide 21. By controlling the temperature of wave-guide 21, perfect matching between the fundamental frequency and the converted frequency can be maintained. The radiation exiting wave-guide 21 through output face 26 mainly includes frequency-converted secondary radiation since nearly all the unconverted radiation is reflected back to diode 10. However, some residual infrared radiation is also included in the output of wave-guide 21. The output radiation is preferably collimated by a lens 28 and, then, filtered by an infrared filter 30 which filters out the residual infrared radiation. The efficiency of the proposed apparatus can be improved by coating a surface 23 of input facet 20 with an anti-reflective featuring high transmission in the infrared range. The anti-reflecting coating prevents destabilization in the operation of the laser diode due to infrared radiation reflected from input facet 20. The anti reflective coating is preferably a dichroic coating which reflects light in the converted frequency range back to waveguide 21, thereby preventing loss of secondary radiation through input facet 20. In a preferred embodiment of the invention, the intensity of the frequency-converted secondary radia- tion is monitored by an output radiation detector, associated with output face 26, which provides an electric feedback responsive to the intensity of the frequency-converted secondary radiation. The electric feedback from the radiation detector is preferably received by thermal controller 34, whereby the tem- perature in waveguide 21 is controlled in accordance with the feedback to provide optimal temperature tun- ing. In a further preferred embodiment of the inven- tion, the electric power supplied to laser diode 10 and the temperature of laser diode 10 are also controlled in accordance with the feedback from the output radia- tion detector, through an appropriate laser control unit, so as to provide a desired frequency-converted output intensity. In contrast to the apparatus suggested in U.S. Patent 5,185,752, the radiation emitted by diode 10 reaches distributed Bragg reflector (DBR) 24 only after passing through frequency converter 22. Due to the present arrangement, the entire frequency converting portion 22 is included in the external cavity which is formed between facet 12 and Bragg reflector 24, as described above. Consequently, all of the unconverted radiation which enters frequency converter 22 partici- pates in the conversion process. Thus, the power con- version efficiency (i.e. the fraction of primary radia- tion power converted into secondary emission at the converted frequency) of the present apparatus is con- siderably higher than that of prior art converters not employing such an external cavity. By placing DBR 24 after frequency converter 22, the conversion efficiency can be increased by a factor of up to 1+R², wherein "R", typically between 0.2 and 1.0, is the reflectivity of DBR 24. It is appreciated that both the over-all doubling efficiency and the Bragg reflectivity of the converted- frequency laser apparatus are substantially quadratic functions of a coupling constant K, defined as the ratio between the intensity of infrared radiation coupled into frequency converting portion 22 and the intensity of laser radiation incident at input facet 20. Thus, to enhance doubling efficiency and intensity stability, the apparatus should preferably be designed to have the largest possible coupling constant K. Many laser applications require an output beam having a substantially circular cross-section. However, existing infrared laser diodes normally generate a beam having a substantially elliptical cross-section. Thus, in order to obtain a high coupling constant K while producing a circular cross-sectioned beam, cylindrical optics are preferably provided between laser diode 10 and waveguide 21. For example, either or both of lenses 16 and 18 may be cylindrical converging lenses designed to reshape the generally elliptical beam generated by diode 10 to have the desired circular cross section at the input of waveguide 21. Alternatively, an anamorphic prism arrangement for reshaping the elliptical beam is placed between lenses 16 and 18. Alternatively, the elliptical infrared beam is adiabatically reshaped into a circular beam by appropriate, preferably anamorphic, construction of the input end of waveguide 21 or by attachment of an anamorphic microlens thereto. KTP waveguides feature a particularly high fre- quency-conversion efficiency for type one interac- tions, denoted i.e. TM -♦ TM, wherein both the fundamen- tal and second harmonic radiations are Z-axis polar- ized, i.e. polarized in a direction perpendicular to the surface of the KTP wafer. However, since the radia- tion of laser diodes, which generate in a TE mode, is normally polarized in a direction generally parallel to the surface on which the laser is mounted, the polari- zation of the laser radiation is preferably rotated. For example, laser diode 10 may be coupled to wave- guide 21 with a rotation of 90°. Alternatively, stand- ard coupling between diode 10 and waveguide 21 is used and a polarity converting element, such as a quarter- wave plate, is provided, for example, between lenses 16 and 18. It should be appreciated that, in an alternative embodiment of the invention, diode 10 may be coupled directly to the input of waveguide 21, preferably just a few micrometers apart from input face 20, thereby obviating the need for lenses 16 and 18. This erabodi- ment may reduce production costs. It is appreciated that, for best results, the temperature in wave-guide layer 21 must be kept sub- stantially constant, so as to maintain the quasi phase matching (QPM) condition between the fundamental wave and the frequency-converted harmonic wave. Therefore, integrated unit 36 is preferably mounted on a thermally conductive base element 38. Base element 38, which is preferably thermally associated with thermal controller 34, acts as a heat sink for heat produced in wave-guide layer 21. For high power lasers, wherein heat is rapid- ly produced in wave-guide 21, more efficient heat dissipation is required. In such cases, unit 36 is preferably mounted on base element 38 using upper surface 40, i.e. the surface under which wave-guide layer 21 is formed, and not as shown in Fig. 1. Reference is now made to Fig. 2A, which illus- trates integrated unit 36 in more detail. Unit 36 is preferably made of a single substrate, such as a KTP (i.e. KTiOPO ) crystal or any other suitable substrate known in the art. Integrated wave-guide 21, formed in the uppermost layer of unit 36 as described above, iε preferably periodically segmented as explained in detail below. As described above, with reference to Fig. 1, wave-guide 21 includes a frequency converting portion 22 (also referred to, herein, as secondary radiation generator) and a distributed Bragg reflector (DBR) portion 24. As can be seen in Fig. 2A, converter portion 22 is divided into segments L^ of equal length "L". Each segment ^ is subdivided into first and second sections L^+ and L^-, respectively. In a preferred embodiment of the invention, the refractive index of the L^+ sections is higher than that of the L^- sections and, more importantly, the crystal polarity of the L^+ sections is opposite that of the L^- sections. DBR portion 22 is divided into segments Q± of equal length Ω, and each segment is sub-divided into first and second sections Ω^+ and Ω^-, respectively. In a preferred embodiment of the invention, the refractive index of the Ω_A+ sections is higher than that of the Ω^- sections but the crystal polarities of the two kinds of sections are not neces- sarily reversed. It should be noted that for efficiently guiding light through waveguide layer 21, the average refrac- tive index in layer 21 is generally larger than the refractive index of the crystalline material on which the waveguide is formed. In accordance with a preferred embodiment of the invention, the reflector portion 24, as described above, meets the Bragg reflection condition, for a frequency "fo"_* when: (1) B(f₀) = πN/Ω wherein £ is the propagation constant for the frequency f_Q and N, an integer, is the reflection order. Frequency converter 22, as described above, meets the QPM condition, for doubling of a fundamental fre- quency "f", when: (2) β(2f)-2β(f) = 2τrM/L wherein M, which is preferably equal to 1, is the QPM order. It should be appreciated that, Since L and Ω are independently selected parameters, both the Bragg and QPM conditions can be complied with, simultaneously, for any wavelength in the transmission range of wave- guide 21. However, it is appreciated that, for effi- cient frequency doubling, the following must hold: (3) δf = |f-f₀| < e wherein e, typically e/f»10~⁴, is the line-width of a characteristic doubling curve. As mentioned above, thermal tuning can improve frequency matching and, therefore, when thermal tuning is used, the burden imposed by equation (3) can be reduced to: (4) δf < r wherein r is the change in frequency induced by chang- ing the temperature of the wave-guide within a given controlled range. Typically, r/f«5-10^"" . Thus, for an efficient apparatus, periods L and Ω must be properly chosen so that equation (4) holds. It is appreciated that existing wave-guide fabri- cation methods, such as using an electron beam mask, feature relatively low resolution in selecting periods "L" and "Ω". Generally, L and Ω must be selected from a set of discrete values, in steps of at least 0.1 μm. This dramatically reduces the leeway in choosing laser frequencies for which equation (4) holds. The present inventors have, therefore, developed a method of pro- ducing improved integrated wave-guides which overcome the above mentioned problem. A preferred embodiment of such an improved wave-guide and a preferred method for producing the improved wave-guide are described below. Reference is now made to Fig. 2B, which illus- trates an alternative, preferred, embodiment of inte- grated unit 36. As in the embodiment of Fig. 2A, an integrated wave-guide layer 51 includes a frequency doubling portion 52 and a DBR portion 54 which are both periodically segmented. But, unlike the embodiment of Fig. 2A, each periodic segment in the embodiment of Fig. 2B includes more than two sub-sections. Portion 52 is divided into superperiods of length L_k, wherein each segment is sub-divided into k sub-periods 41. Portion 54 is divided into superperiods of length Ωj, wherein each segment Ω_j is sub-divided into j sub-periods 45. Fig. 2B illustrates one example of such an arrangement, wherein k=3 and j=5, but it should be appreciated that any other suitable values of k and j may also be used, in accordance with specific requirements. Each of the k sub-periods 41, in each superperiod L_k, includes a first section 42 followed by a second section 44. Sections 42 have properties (i.e. refrac- tive index and electric polarity) similar to those of sections L^, described above with reference to Fig. 2A, and sections 44 have properties similar to those of sections L^ of Fig. 2A. Similarly, each of the j sub- periods, in each superperiod Ω_k, includes a first section 46 followed by a second section 48. Sections 46 have properties (i.e. refractive index and electric polarity) similar to those of sections Ω^+, described above with reference to Fig. 2A, and sections 48 have properties similar to those of sections Ω^- of Fig. 2A. It can be seen that for k=j=l, portions 52 and 54 of Fig. 2B are identical to portions 22 and 24, respec- tively, of Fig. 2A. When the super period construction of Fig. 2B is used, the Bragg condition should be restated as fol- lows:

and the QPM condition is restated as follows: (6) β(2f)-2β(f) = 2πk/L It should be appreciated that, when such super- structures are used, QPM frequency "f" is controlled by the average length of sub-periods 41, i.e. L_k/k. For a given resolution in selecting the length of sub-periods 41, typically 0.1 μra as described above, the effective resolution in selecting the average sub-period L_k/k is improved by a factor of k and, typically, equals 0.1/k μm. Thus, the limited resolution in selecting frequency "f" can be improved considerably. A similar improvement of resolution can be achieved in selecting Bragg fre- quency "fo", which is controlled by the average length of sub-periods 45, i.e. Ω_j/j. The average sub-period Ω_j/j may thus be selected with a resolution of 0.1/j μm. It is appreciated that due to additional, random, variance in the reproducibility of integrated wave- guide 51, the frequency difference δf (δf=|f-fø|) may fluctuate within a frequency range μ which is larger than r, as defined in equation (4) above. However, the present inventors have developed a production procedure which overcomes this problem. According to the new procedure, a series of N (N=2μ/r) successive integrated wave-guides, such as integrated wave-guide 51, are formed on single chip of a suitable substrate (preferred substrates are dis- closed in the examples given below) . The wave-guides in the series are prearranged to have gradually increas- ing QPM frequencies "f", i.e. each integrated wave- guide in the series is designed to have a target QPM frequency "f" higher than that of the preceding wave- guide and lower than that of the subsequent wave-guide. The design QPM target frequencies are typically sepa- rated by intervals of α=r/2. In contrast, the target Bragg frequencies "fo" of the different wave-guides in the series are equal. The different wave-guides in the series thus formed are then tested for compliance with equations (3) and (4). It should be appreciated that at least some of the resultant integrated wave-guides are bound to comply with the condition set forth by equation (4) or, even, with the more rigid condition of equation (3). The substrate bearing the series of integrated wave-guides may be cut at appropriate locations, be- tween adjacent wave-guides 51, thus forming a series of separate integrated units 36. In accordance with a preferred embodiment of the invention, any of the resultant wave-guides complying, in fact, with either of the above mentioned conditions may be used in fre- quency-converted laser apparatus as described above. The remaining integrated units, failing to comply with both of conditions (3) and (4), are not used and may be disregarded. Alternatively, the series of integrated wave- guides can be designed to have gradually increasing DBR frequencies "f₀" and equal QPM frequencies f". Again, some of the resultant integrated wave-guides are ex- pected to comply with conditions set by equations (3) or (4). To ensure efficient frequency locking, the present invention provides a method for repressing undesired Bragg reflections from frequency doubling portion 52 (Fig. 2B) by imposing controlled randomization in superperiods L_k, thereby "breaking" potentially reflec- tive patterns in doubling portion 52. Such controlled randomization can be achieved, for example, by slightly modifying the length of some of segments 41, in ac- cordance with a predetermined scheme, while ensuring that, over all, equation (6) still holds. For example, it has been found that by slightly modifying the length of one of segments 41 in each superperiod L_k, Bragg reflections from portion 52 can be almost eliminated. According to the present invention, laser frequen- cy stability may be improved by maximizing the reflec- tivity, R_e, of Bragg reflector portion 24 (Fig. 2A) . It can be shown that when Bragg reflector portion 24 or 54 is sufficiently long, the following equation holds: (7) R_e * sin² (2πΩ_i ⁺/ , wherein Ω_± is the length of a segment in Bragg reflector 24 and i is the radia- tion wavelength in the medium of which Bragg reflector 24 is formed. It will be appreciated that when

= n, where- in n is an integer, R_e is substantially equal to zero, due to destructive interference between the two ends of each segment Ω_i ⁺. It has been found that to avoid frequent occurrence of such destructive interference, segment length Q^⁺ must be controlled with an accuracy of approximately ι/4, which is typically on the order of 0.1 microns. However, due to the above mentioned limitations of lithographic techniques and ion exchange processes, such control is very difficult. Thus, the present inventors have devised a practical solution to this problem which obviates the need for such high accuracies. According to a preferred embodiment of the present invention, two waveguides, having substantially identi- cal frequency doubling portions 22 but different Bragg reflector portions 24, are formed. The two waveguides are designed such that the Bragg reflector segments in one waveguide are longer by a preselected difference, δΩ^, than the Bragg reflector segments in the other waveguide. It has been found that while the absolute value of segment lengths Ω^ is difficult to control, the difference δQ^ between the segments of the two waveguides can be controlled with the desired accuracy of approximately 0.1 micrometers. It will be appreciat- ed that since the difference between the Bragg reflec- tor segments of the two waveguides is approximately ι/4, at least one of the waveguides will have a reflec- tivity, R_e, of at least 50 percent. Specific examples of actual integrated wave-guide units, as well as methods of producing such units, are presented in the following paragraphs. These examples are provided for explanatory purposes only and should not, in any way, be understood as limiting the scope of the present invention. Other designs, substrates and fabrication methods, some of which may yield equal, or even better, results when used with the present inven- tion, are also within the scope of the present inven- tion. For example, although unit 36 is preferably formed of a KTP crystal, other crystals of the same crystal family, such as KTA, RTP and CsTP, may be equally suitable. More generally, any crystals of the family XYOZ0₄, wherein X is K, Cs, Rb, Na, etc., Y is Ti, etc. and Z is P, As, etc., or any combinations thereof, may be used. Alternatively, other crystalline materials may be used, for example the crystals LiNb0₃, KNb0₃ and LiTa0₃.and A series of integrated wave-guide units was formed on a single chip of a KTP, i.e. KTiOP0₄. All of the Bragg reflector portions in the series were designed to reflect a frequency "fø" corresponding to a wavelength of 853 nm. The series included nine integrated wave- guides having nine different frequency doubling por- tions. Each of the nine doubler portions was construct- ed in accordance with a superstructure of periodic segments, including 12 sub-periods, such as the L_ks of Fig.2B (wherein k=12) . The difference in super-period length (L_k in Fig. 2B) between successive doubler portions was 0.1 μm. The shortest super-period, which was 47.8 μra long, included ten sub-periods 41 of 4 μm and two sub-periods 41 of 3.9 μm. Each sub-period 41 included two adjacent sections such as sections 42 and 44 of Fig. 2B. The difference of 0.1 μm, between the super-period lengths L_k of successive doubler portions 52, was obtained by a modification of 0.1 μm in one of their respective sub-periods 41. Each of the nine DBR portions was constructed in accordance with a periodic structure of segments, such as the Ω_j^s of Fig. 2A or the Ω S of Fig. 2B (wherein j=l). All of the DBR portions were designed to have the same period, 3.7 μm, including a 2.3 μm section, such as Q±+ in Fig. 2A, and a 1.4 μm section, such as sec- tion Q - in Fig. 2A. The series of wave-guides according to the above design was produced using a production method as de- scribed below. A Z-cut KTP crystalline chip was coated with a layer of photoresist. The photoresist coat was selectively exposed to light, through a patterned mask made by an electron beam, in accordance with the de- sired pattern of the series of wave-guides described above. The exposed layer was then developed in order to remove the exposed portions, thereby creating a patterned layer of photoresist. The photoresist layer was then overcoated with a thin layer of Ti which was, then, selectively removed off the underlying photore- sist coated regions by a lift-off process, as is well known in the art. Thus, a patterned layer of Ti was created on the surface of the chip in the areas not masked with the photoresist. Finally, a solution containing RbN0₃ and Ba(N0₃)₂ was applied to the patterned surface and, by cation replacement at the portions of the uppermost layer of the KTP chips not covered with titanium, a series of integrated wave-guides, as described above, was formed. The titanium was removed by etching. Each of the wave- guides, in the series thus formed, was approximately 10 mm long, of which the frequency doubling portion occu- pied approximately 6 mm and the DBR portion occupied approximately 4 mm. Some of the 9 wave-guides thus produced have complied with equation (4) above. As another example, a series of integrated wave- guide units was designed in accordance with a fundamen- tal infrared wavelength of 830 nm. As in the previous example, this series also included 9 integrated wave- guide units formed on a KTP chip. All of the Bragg reflector portions of this series were designed to have the same period length, of 2.7 μm, adapted to reflect a wavelength of 830 nm (the fundamental wavelength) . As in the previous example, the superperiods L_k of the frequency doubling portions included 12 sub-periods (i.e. k=12) . The first wave-guide in the series had a superpe- riod length of 43.2 μm divided into 12 sub-periods 41 of 3.6 μm length each. The difference in super-period length L_k between successive doubler portions was, again, 0.1 μm, obtained by a modification of 0.1 μm in one of their respective sub-periods 41. The series was produced in accordance with the method described in the previous example. While a preferred method for fabricating the waveguides of the present invention is described above, other methods and techniques as known in the art may be equally suitable for the present invention. The integrated waveguide units of the present invention provide a reliable and relatively inexpensive means for producing visible laser radiation embodied in a particularly compact device. The integration of a doubler portion and a Bragg reflector portion into one unit, in the production stage, considerably reduces the sensitivity to ambient conditions, such as temperature, and obviates the need for post-production matching of the two portions. Furthermore, the optical wave-guiding continuum formed by the integration of the two portions avoids loss of energy, which commonly occurs at the interface of two optical elements. Such an integrated waveguide is virtually unaf- fected by mechanical vibrations in the laser system. A frequency-converted laser apparatus using an integrated unit, in accordance with the present invention, will generally be less sensitive to small variations in the distance between the laser diode and the waveguide, compared to prior art systems which do not use a Dis- tributed Bragg reflector. This is because control of the locked fundamental frequency "f₀" is achieved by adjusting the peak of reflection of the DBR portion, which is independent of the distance between the diode and the waveguide. Reference is now made to Fig. 3 which schematical- ly illustrates a system for controlling a preferred ion exchange process used for the formation of a frequency doubling optical waveguide. The system includes an ampule 60, preferably of silica, surrounded by a heat conducting tube 62, preferably of nickel, which is preferably surrounded, in turn, by a cylindrical fur- nace 92. The temperature in furnace 92, which is pref- erably measured by a thermocouple 114 and displayed on a display 124, is preferably regulated by a tempera- ture regulator 130, adjusted to maintain a desired temperature. A first tubular section 116 is sealingly mounted on ampule 60, preferably via a first sealing member 90, and connected to one side of a gate valve 106, prefera- bly via a second sealing member 88. A second tubular section 108 is mounted to the other side of gate valve 106, preferably via a third sealing member 86, and to a guiding tube 82, preferably via a flange 84. The top of guiding tube 82 is preferably sealed from the external environment by a vacuum fitting 80. A transfer rod 78, associated with an external motor 76, sealingly extends through tube 82, section 108, valve 106 and, when valve 106 is open, to section 116. The bottom end of rod 78 is preferably adapted for supporting a sample holder 64, which preferably includes a platinum basket. Ac- cording to this arrangement, controlled activation of motor 76 results in controlled movement of sample holder 64 in the substantially sealed interior volume defined by ampule 60, first and second sections 116 and 108, gate valve 106 and tube 82. As described below, ampule 60 preferably contains a sufficient amount of molten salt solution 66, such that a sample (not shown) on sample holder 64 can be completely immersed in solution 66. The substantially sealed volume, defined by ampule 60, first and second sections 116 and 108, gate valve 106 and tube 82, will be hereinafter referred to as the sealed environment. In a preferred embodiment of the invention, second section 108 is connected to a vacuum pump 74 which pumps air out of the interior of ampule 60 when a vacuum valve 102 is open. Second section 108 is prefer- ably also connected, via a gas input valve 100 and a gas mixer 110, to a plurality of control valves, pref- erably including an N₂ control valve 94, an 0₂ control valve 96 and a control valve 98 for additional gases, for example H₂0 vapors. To measure the humidity in ampule 60, first section 116 is preferably connected, via a hydrometer 68 and an output gas valve 104, to a ventilated bubbler 72. A preferred procedure for controlling the concen- tration of oxides in molten salt solution 66 will now be described. As mentioned above, an efficient frequen- cy doubler is generally obtained only when the oxide concentration in the ion exchanging solution, [0 ], lies within a narrow range, c_a≤[0^2""]≤c_b, wherein c_a and c_b depend on the type of KTP crystal used. When [0²~ ]<c_a, the resultant waveguides are expected to yield a very poor frequency conversion efficiency. When [0 ] >c→-₀, the optical transmission efficiency of infrared radiation is expected to be very poor. Since the interior of ampule 60 is sealed from the external environment, uncontrollable effects of the external environment on the concentration of salts in solution 66 are avoided. According to the present invention, solution 66 is kept in a controlled environment, preferably an inert gas environment, as described below, thereby minimizing undesirable inter- action between solution 66 and the external environ- ment. Molten salt solution 66 is preferably prepared as follows. A predetermined mixture of pure salts, for example 60 grams of a salt mixture containing RbN0₃ and Ba(N0₃)₂, wherein the concentration of Ba(N0₃)₂ in the mixture is 3 to 10 atomic percent, which is preferably stored in a dry environment prior to use, are heated by a heater at a temperature of approximately 200^βC, in an ambient atmosphere, removing undesirable moisture from the mixture. Then the temperature is increased gradual- ly until, to a temperature of approximately 320"C, the mixture starts to melt. After the mixture has complete- ly melted, it is cooled until it solidifies. The solid mixture is then placed in ampule 60 and reheated by furnace 92 to approximately 330^βC until it is again melted. Although the salts mentioned above are preferred, other salts may be equally suitable for the process of the present invention. For example, the rubidium (Rb) may be replaced by cesium (Cs) or thallium (Tl) and the barium (Ba) may be replaced by calcium (Ca) , strontium (Sr) or lead (Pb) . Dry nitrogen from (N₂), preferably having a rela- tive humidity of less than 10~⁴ atmospheres, is then introduced through N₂ control valve 94, mixer 110, input valve 100, second connector 108, gate valve 106 and first connector 116, into ampule 60 to provide a substantially inert environment therein. The molten salt mixture is preferably kept in the dry nitrogen environment for approximately 24 hours to ensure horao- genenization of the melt. If the original salt mixture includes pure and dry salts, as is preferably the case, the oxide concentration during this stage of the proc- ess is lower than the required minimum, i.e. [0^2"~]<c_a. In a preferred embodiment of the present inven- tion, the concentration of oxides, [0²~] in solution 66 is controlled by adding predetermined amounts of do- nors, i.e. additives which increase the concentration of oxides in the solution, and/or acceptors, i.e. additives which reduces the concentration of oxides in the solution, as necessary for maintaining the concen- tration of oxides within the range c_a≤[0²~]≤C_jD. Typical donors include, for example, oxides such as Na₂0, BaO and the like, peroxides such as Na₂0₂ and the like, hydroxides such as NaOH and the like and carbonates such as Na₂C0₃ and salts of S0 ²~, Cr0₄ ²~ and P0₄ ^J~. Typical acceptors are NH₄N0₃ and salts of P0₃~, S₂0₇ ²~ or Cr₂0₇ ~. In accordance with a preferred embodiment of the invention, a doping solution containing the preselected amounts of donors and/or acceptors, in accordance with the salts used in the process, is added to molten salt solution 66 in ampule 60. Since, initially, the oxide concentration of molten salt solution 66 is preferably lower than the required minimum, c_a, the doping solution will normally contain donors only. The appropriate amount of donor is determined empirically. Since the correct amount of oxide donor is generally very small, better results are achieved when a diluted doping solution is used. For example, assuming that the initial concentration of oxides in the molten salt solution is approximately 10 ppm or less, a doping solution containing approximate- ly 40 grams of RbN0₃ and approximately 20 milligrams of Na₂0 has been found suitable for KTP crystals. To prepare this doping solution, the RbN0₃is first melted and, then, the Na₂0₂ is added to the melt. The melted mixture is then cooled rapidly. A predetermined amount of the prepared doping solution, depending on the type of KTP used, is then added to ion exchanging molten salt solution 66, and the doped mixture is kept in the inert environment of ampule 60 for approximately 1 - 3 hours, to allow horaogenization of the doped mixture. For the RbN0₃ and Ba(N0₃)₂ molten salt solution described above, 2 - 4 grams of this doping solution are required in order to obtain an oxide concentration of approximately 30 - 60 ppm, which is a preferred concentration range for certain types of KTP crystals. The oxide concentration control method described in the preceding paragraphs uses a nitrate solution, doped with a predetermined amount of oxygen donor, which is maintained in an inert environment. Since the oxide concentration may change with time, subsequent corrections may be made by adding predetermined amounts of oxide donors or acceptors to the solution. However, since these corrections are generally made on a trial and error basis, the present invention also provides a method for dynamically controlling the oxide concentra- tion by providing a mixed environment including active and inert gases, rather than an exclusively inert environment as described above. The relative pressures of the inert and active gasses are preferably con- trolled via control valves 94, 96 and 98. As described above, valve 94 controls the supply of nitrogen, valve 96 controls the supply of oxygen and valve 98 controls the supply of additional gasses such as water vapors. As described in G. Chariot et. al., "Chemical Reactions in Solvents", Pergamon Press, London (1969), the concentration of oxides is determined by the ratio between the concentration of active gasses in the environment and the concentration of dopant in the solution. The active gases in the environment interact with the dopant in the solution until a steady state is reached, thereby preventing subsequent changes in the concentration of oxides in the solution. In actual experiments, H₂0 vapors have been successfully used with solutions containing OH^" type dopants and C0₂ can be used with solutions containing C0₃ ²~ type dopants. In one embodiment of the invention, no dopants are added to the molten salt solution and, according to this embodiment, oxide concentration control is achieved by using a mixture of several active gasses. A typical gas mixture for such oxide concentration con- trol includes N0₂ and 0₂. Appropriate gas pressures are provided using control valves 94, 96 and 98. It should be appreciated that the concentration of oxides in molten salt solution 66 can be monitored directly by suitable electrochemical potential measure- raents, as known in the art. However, since proper execution of such measurements is generally complicated and inefficient for the purposes of the present inven- tion, monitoring of the oxide concentration in solution 66 is preferably achieved indirectly, using post-fabri- cation testing, as described in detail below. A preferred technique for introducing a titanium masked KTP crystal wafer to molten salt solution 66 will now be described. First, the KTP crystal is placed in sample holder 64, which may be a platinum basket. Then, after assuring that gate valve 106 is air-tightly closed, flange 84 is opened and sample holder 64 is mounted to the bottom of rod 78. Then, flange 84 is sealed and vacuum pump 74 evacuates second section 108 and tube 82, via valve 102. Then, valves 100 and 104 are opened to resume the flow of nitrogen from N₂ source 96 into the sealed environment and out, through hydrometer 68, into bubbler 72. Then, gate valve 106 is opened and sample holder 64 is lowered, manually or by use of an appropriate driving unit, into the interior of ampule 60, preferably to a level higher than the level of molten salt in the ampule, so as to enable temperature matching between the KTP crystal and solu- tion 66. Sample holder 64 is then further lowered until the KTP crystal is completely immersed in solution 66, where the crystal is kept for an ion exchange period of approximately 30 minutes. After the ion exchange process has been completed, sample holder 64 is raised above gate valve 106 and gate valve 106 is, then, closed. After the KTP crystal cools, flange 84 is opened and sample holder 64 is detached from rod 78. To complete the process, the KTP crystal is then cleaned from residual salts, preferably using hot deionized water. Reference is now made to Figs. 4A, 5A and 6A, which schematically illustrate X-ray diffraction pat- terns of waveguides formed on crystal surfaces. Fig. 5A illustrates the diffraction pattern of an efficient frequency doubler, while Figs. 4A and 6A illustrate the diffraction patterns of two respective types of ineffi- cient frequency doublers. As described above, the present inventors have found a method of evaluating the performance of the resultant waveguide using X-ray diffraction pattern analysis. According to the present method, the back surface of the processed KTP crystal is scanned with an X-ray having a preselected wave- length and the resultant diffraction pattern is com- pared with a reference pattern which corresponds to a functionable waveguide, for example the pattern shown in Fig. 5A. In a preferred embodiment of the present inven- tion, a double crystal X-ray diffractoraeter in a rock- ing curve measurement mode is used. In Figs. 4A, 5A and 6A, the graphs denoted "N" correspond to measurements along the Y-axis of the KTP surface while the graphs denoted "R" correspond to measurements along the X-axis of the KTP surface. The horizontal axis in the graphs of Figs. 4A, 5A and 6A corresponds to the X-ray dif- fraction angle, wherein each centimeter is equivalent to approximately 100 arc seconds, while the vertical axis corresponds to the X-ray intensity. The X-ray diffraction pattern is indicative of the depth distribution of salt components, such as rubidium (Rb) , in the processed surface of the KTP crystal. Based on the salt depth distribution of the KTP sur- face, it is possible to determine whether the ion exchange process described above, with reference to Fig. 3, has been successful, i.e. whether an operable waveguide has been formed. Thus, the preferred X-ray diffraction analysis methods described below are merely examples of techniques for evaluating the ion exchange profile of the KTP surface. It should be appreciated that other techniques, some of which are mentioned herein, may be equally suitable for evaluating the ion exchange profile. Fig. 4A illustrates typical double crystal rocking curve (DCRC) graphs of a KTP sample which had been prepared using a solution having a deficiency of oxides, i.e. having [0²~]<c_a. The DCRC graph is charac- terized by a distinct asymmetry, wherein the slope on the right of each peak is much steeper than the slope on the left of the peak, which is substantially asyrap- totic. Fig. 4B illustrates the Rb atomic profile of the KTP surface, as a function of depth, determined based on an appropriate mathematical transformation of the DCRC of Fig. 4A. Fig. 4C illustrates the logarithm of the profile of Fig. 4B. The solid graph in Fig. 4C is based on actual measurements, while the dashed line is the best linear fit to the results on the logarithmic scale. Comparing Figs. 4B and 4C, it can be seen that in this case the Rb profile is substantially exponen- tial, i.e. it almost matches the straight, dashed, line in Fig. 4C. As mentioned above, a waveguide formed on this type of KTP surface will feature high infrared transmission but substantially no frequency conver- sion. Fig. 5A illustrates typical double crystal rocking curve (DCRC) graphs of a KTP sample which had been prepared using a solution having the correct concentra- tion of oxide, i.e. having c_a<[0^2""]<c_b. In this case, the graphs are characterized by a main peak and a distinctive "shoulder" to the left of the main peak. The Rb profile of this sample, shown in Fig. 5B, is substantially linear and, therefore the logarithm of the Rb profile, shown in Fig. 5C, is distinctly not linear. Fig. 6A illustrates typical double crystal rocking curve (DCRC) graphs of a KTP sample which had been prepared using a solution having an excessive oxides concentration, i.e. having [0²~]>C_jD. In this case, the graphs are characterized by very narrow, almost syramet- ric, peaks. This indicates a very low Rb concentration in the layer. It will be appreciated that the Rb pro- file of this sample, shown in Fig. 6B, and its loga- rithm, shown in Fig. 6C, are clearly distinguishable from those of Figs. 4B, 4C, 5B and 5C. As mentioned above, there are other methods for determining the Rb profile of a sample. For example, the profile may be analyzed directly by Secondary Ion Mass spectrometry (SIMS), Auger Electron Spectroscopy (AES) and Electron Microprobe Analysis or, indirectly, by optical measurements. However, X-ray diffraction analysis is preferred since it is relatively uncompli- cated to perform, it yields distinctive results and it is non-destructive. It should be appreciated that X-ray diffraction pattern analysis as described above can be utilized to determine the proper amount of donors and/or acceptors which should be added to the ion exchange solution before immersing a masked crystal therein. In a pre- ferred embodiment of the present invention, the amount of donors and/or acceptors in the doping solution used, for a given crystal, is varied according to X-ray analyses as described above, until the proper oxide concentration for the given crystal is determined. Experiments have shown that the efficiency of frequency-doublers formed in accordance with the pre- ferred embodiments described above, measured in norraal- ized units, was in excess of 800% W^-1cra , which is higher than any existing frequency doubler and any non- linear waveguide. Furthermore, when such frequency doublers were coupled to infrared diode lasers, a blue light output of approximately 3 milliwatts was success- fully generated. It should be appreciated that the present inven- tion is not limited to what has been thus far described with reference to preferred embodiments of the inven- tion. Rather, the scope of the present invention is limited only by the following claims:

Claims

C L A I M S

1. A frequency-converted laser apparatus comprising: a laser source which generates primary radiation; a Bragg reflector which reflects at least part of the primary radiation, within a frequency range cen- tered at a preselected fundamental frequency, back along an optical path to the laser source, thereby to form an external laser cavity including the laser source resonant at the preselected fundamental frequen- cy; and a frequency converter located within the external laser cavity, on the optical path between the laser source and the Bragg reflector, and operative to con- vert part of the primary radiation into frequency- converted secondary radiation.

2. A frequency-converted laser apparatus according to claim 1 wherein the laser source comprises a reflective back facet and wherein the external laser cavity ex- tends from the reflective back facet to the Bragg reflector.

3. A frequency-converted laser apparatus according to claim 1 or claim 2 wherein both the frequency converter and the Bragg reflector are formed on a crystalline material.

4. Apparatus according to claim 3 wherein the crys- talline material comprises a crystal of the family XYOZ0₄, wherein X comprises at least one of the ele- ments K, Cs, Rb and Na, Y comprises Ti and Z comprises at least one of the elements P and As.

5. Apparatus according to claim 1 or claim 2 wherein the primary radiation emitted by the laser source is in the infrared range.

6. Apparatus according to claim 1 or claim 2 wherein the laser source is a laser diode.

7. Apparatus according to claim 1 or claim 2 wherein the frequency-converted secondary radiation is in the visible range.

8. Apparatus according to claim 1 or claim 2 wherein the frequency of the secondary radiation is higher than the fundamental frequency.

9. Apparatus according to claim 8 wherein the fre- quency of the secondary radiation is a second harmonic frequency equal to twice the fundamental frequency.

10. Apparatus according to claim 1 or claim 2 wherein the frequency-converter comprises a plurality of peri- odic segments including sections of a first kind, having a first index of refraction, and sections of a second kind, having a second index of refraction, and wherein the first kind of sections have a crystal polarity opposite the crystal polarity of the second kind of sections.

11. Apparatus according to claim 10 wherein each periodic segment is a superperiod L_k comprising k sub- periods and wherein each sub-period comprises one section of the first kind followed by one section of the second kind.

12. Apparatus according to claim 1 or claim 2 wherein the Bragg reflector comprises a plurality of periodic reflecting segments including reflecting sections of a first kind, having a first index of refraction, and reflecting sections of a second kind, having a second index of refraction.

13. Apparatus according to claim 12 wherein each periodic reflecting segment is a superperiod Ω_j com- prising j reflecting sub-periods and wherein each reflecting sub-period comprises one reflecting section of the first kind followed by one reflecting section of the second kind.

14. Apparatus according to claim 1 or claim 2 wherein the frequency converter comprises an input face for receiving the primary radiation, and further comprising at least one lens, located between the primary oscilla- tor and the frequency-converting portion, for focusing the primary radiation onto the input face of the fre- quency converter.

15. An integrated wave-guide unit comprising: a frequency converting portion adapted to receive primary radiation and, upon receipt of the primary radiation, to convert part of the primary radiation into frequency-converted secondary radiation; and \ a Bragg reflector portion adapted to receive radiation from the frequency-converting portion and to reflect part of the primary radiation, within a limited frequency range centered at a preselected fundamental frequency, back through the frequency converting portion.

16. An integrated wave-guide unit according to claim 15 wherein the frequency converting portion and the Bragg reflector portion are both formed on a single chip of crystalline material.

17. An integrated wave-guide unit according to claim 16 wherein the single chip of crystalline material comprises a crystal of the family XY0Z0₄, wherein X comprises at least one of the elements K, Cs, Rb and Na, Y comprises Ti and Z comprises at least one of the elements P and As.

18. An integrated unit according to any of claims 15 - 17 wherein the primary radiation is in the infrared range.

19. An integrated unit according to any of claims 15 - 17 wherein the primary radiation is the output of a laser diode.

20. A unit according to any of claims 15 - 17 wherein the frequency-converted secondary radiation is in the visible range.

21. A unit according to any of claims 15 - 17 wherein the frequency of the secondary radiation is higher than the fundamental frequency.

22. A unit according to claim 21 wherein the frequency of the secondary radiation is a second harmonic fre- quency equal to twice the fundamental frequency.

23. A unit according to any of claims 15 - 17 wherein the frequency converting portion comprises a plurality of periodic segments including sections of a first kind, having a first index of refraction, and sections of a second kind, having a second index of refraction, and wherein the first kind of sections have an electric polarity opposite the electric polarity of the second kind of sections.

24. A unit according to claim 23 wherein each periodic segment is a superperiod L_k comprising k sub-periods and wherein each sub-period comprises one section of the first kind followed by one section of the second kind.

25. A unit according to any of claims 15 - 17 wherein the Bragg reflector portion comprises a plurality of periodic reflecting segments including reflecting sections of a first kind, having a first index of refraction, and reflecting sections of a second kind, having a second index of refraction.

26. A unit according to claim 25 wherein each periodic reflecting segment is a superperiod Ωj comprising j reflecting sub-periods and wherein each reflecting sub- period comprises one reflecting section of the first kind followed by one reflecting section of the second kind.

27. Apparatus according to claim 25 wherein the lengths of the periodic reflecting segments are select- ed so as to provide optimal Bragg reflection by the Bragg reflector at the primary radiation.

28. A frequency-converted laser apparatus comprising: a laser source, having a reflective back facet and a front facet, operative to emit primary radiation through the front facet; and an integrated wave-guide unit according to any of claims 15 - 17 having an input face, optically associ- ated with the front facet of the laser source, for receiving the primary radiation, whereby an external laser cavity, resonating at the fundamental frequency, is formed between the re- flective back facet of primary oscillator and the Bragg reflector portion of the integrated wave-guide.

29. Apparatus according to claim 28 and further com- prising at least one lens, located between the front facet of the primary oscillator and the input face of the integrated wave-guide unit, for focusing the pri- mary radiation onto the input face of the integrated wave-guide.

30. Apparatus according to claim 28 wherein the input face of the integrated waveguide unit is directly coupled to the front facet of the laser source.

31. Apparatus according to claim 28 and further com- prising at least one optical element which shapes the frequency-converted secondary radiation to produce a substantially circular output beam.

32. Apparatus according to claim 31 wherein the at least one optical element comprises at least one cylin- drical lens.

33. Apparatus according to claim 31 wherein the at least one optical element comprises at least one anara- orphic prism.

34. Apparatus according to claim 28 and further cora- prising a thermal controller, associated with a surface of the integrated waveguide unit, which controls the temperature of the integrated waveguide unit.

35. Apparatus according to claim 34 and further com- prising an output radiation detector which provides a feedback responsive to the intensity of the frequency- converted secondary radiation, wherein the thermal controller receives the feedback from the output radia- tion detector and controls the temperature of the integrated waveguide unit in accordance with the feed- back.

36. Apparatus according to claims 28 and further comprising: an output radiation detector which provides a feedback responsive to the intensity of the frequency- converted secondary radiation; and a laser control unit which receives the feedback from the output radiation detector and controls the electric power provided to the laser source and/or the temperature of the laser source in accordance with the feedback.

37. A method of forming on the surface of a crystal a frequency converting optical waveguide including a plurality of periodic segments comprising sections of a first kind, having a first electric polarity, and sections of a second kind, having a second electric polarity, comprising: masking a plurality of discrete regions on the surface of the crystal according to a preselected masking pattern; preparing an ion exchanging molten salt solution; controlling the concentration of oxides in the molten salt solution; immersing the masked surface of the crystal in the molten salt solution for a preselected period of time; and removing the masking material from the surface of the crystal.

38. A method according to claim 37 wherein the ion exchanging molten salt solution is maintained in a substantially inert environment during immersing of the masked surface in the solution.

39. A method according to claim 37 wherein the ion exchanging molten salt solution is maintained in an environment containing both inert and active gases during immersing of the masked surface in the solution.

40. A method according to any of claims 37 - 39 where- in the molten salt solution comprises at least one of the ions Rb⁺, Ba²⁺, Cs⁺, Sr²⁺, Na⁺, Ca²⁺, Tl⁺ and Pb²⁺.

41. A method according to any of claims 37 - 39 where- in controlling the concentration of oxides in the solution comprises maintaining the concentration of oxides within a preselected concentration range.

42. A method according to claims 41 wherein raaintain- ing the concentration of oxides in the solution cora- prises maintaining an oxides concentration of between 10 ppm and 60 ppm.

43. A method according to any of claims 37 - 39 where- in controlling the concentration of oxides in the solution comprises adding an appropriate amount of doping solution to the molten salt solution.

44. A method according to claim 43 wherein the doping solution comprises at least one oxide donor.

45. A method according to claim 44 wherein the at least one oxide donor comprises at least one donor of at least one of the groups OH^", C0₃ ^2", HC0₃ ^"", 0₂~, 0²~, S0₄ ²~, Cr0₄ ^2" and P0₄ ^3".

46. A method according to claim 43 wherein the doping solution comprises at least one oxide acceptor.

47. A method according to claim 46 wherein the at least one oxide acceptor comprises at least one accep- tor of at least one of the groups NH₄ ⁺, P0₃~, S₂0₇ ²~ and Cr₂0₇ ^2".

48. A method according to any of claims 37 - 39 where- in controlling the concentration of oxides comprises controlling the concentration of chemically active gasses in the environment surrounding the molten salt solution.

49. A method for evaluating whether a frequency converting optical waveguide formed on a surface of a crystal will efficiently convert an input frequency to an output frequency, comprising: determining an ion exchange depth profile of an area of a surface of the crystal; and comparing the ion exchange profile of said area to a predetermined ion exchange profile indicative of efficient frequency conversion.

50. A method according to claim 49 wherein determin- ing the ion exchange depth profile of said area com- prises determining the depth distribution of at least one ion in said area and wherein comparing the ion exchange profile comprises comparing the distribution of the at least one ion in said area to a predetermined distribution of the at least one ion.

51. A method for evaluating whether a frequency converting optical waveguide formed on a surface of a crystal will efficiently convert an input frequency to an output frequency, comprising: irradiating an area of a surface of the crystal with X-ray radiation; measuring the magnitude of X-ray radiation dif- fracted by said area across a preselected range of angles to obtain a diffraction pattern of said area; and comparing the diffraction pattern of said area to a predetermined X-ray diffraction pattern indicative of efficient frequency conversion.

52. A method according to claim 51 and further cora- prising calculating an ion exchange depth profile of said area based on the diffraction pattern of said area, wherein comparing the diffraction pattern of said area comprises comparing the calculated ion exchange profile of said area to a predetermined ion exchange profile.

53. A method according to claims 49 or claim 51 where- in said area is on a surface of the crystal other than the surface on which the waveguide is formed.

54. A method according to claim 53 wherein said area is on a back surface of the crystal opposite the sur- face on which the waveguide is formed.

55. A method according to claim 49 wherein determining the ion exchange depth profile of the surface comprises determining the ion exchange depth profile by at least one of Secondary Ion Mass Spectrometry (SIMS), Auger Electron Spectroscopy (AES) and Electron Microprobe Analysis.

56. A method according to claim 49 wherein determining the ion exchange depth profile of the surface comprises determining the ion exchange depth profile based on optical measurements.

57. A method according to claim 50 wherein determining the depth distribution of the at least one ion cora- prises determining the distribution of the at least one ion by at least one of Secondary Ion Mass Spectrometry (SIMS), Auger Electron Spectroscopy (AES) and Electron Microprobe Analysis.

58. A method according to claim 50 wherein determining the depth distribution of the at least one ion cora- prises determining the distribution of the at least one ion based on optical measurements.

59. A method according to any of claims 50, 57 and 58 wherein the at least one ion comprises at least one of the salt components rubidium (Rb) , cesium (Cs) and Thallium (Tl) .

60. A method according to claim 49 or claim 51 where- in the crystal is a KTP crystal.

61. Apparatus according to claim 4 wherein the XYOZ0₄ crystal comprises a KTP crystal.

62. An integrated wave-guide unit according to claim 17 wherein the XY0Z0 crystal comprises a KTP crystal.