JP2000081527A - Production of proton-exchanged waveguide - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a multifunctional integrated optical chip having a high electro-optical coefficient and low loss without having microcracks or inner stress which produces microcracks by repeating proton exchange treatment and annealing process till a specified value is achieved. SOLUTION: A mask 22 is disposed on a lithium niobate wafer 20 to form a waveguide pattern. A fused acid 24 which is benzoic acid is used to exchange proton (H+ hydrogen ion) and lithium ion (Li+). The proton diffusion depth in the region 26 where proton is exchanged is measured, and the process to dip the wafer 20 in the fused acid 24 is repeated till the first (proton exchange) target depth is obtained. Then the mask 22 is removed and the wafer 20 is thermally annealed. The proton diffusion depth in the annealed proton-exchange region, namely, in the waveguide region 26 is measured, and thermal annealing is repeated till the second (annealing) target proton diffusion depth is obtd.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a multifunctional integrated optical chip (MIOC) having an optimally high electro-optic coefficient and a method for producing a proton-exchanged waveguide, preferably having a rhombohedral structure.

[0002]

Typical method for manufacturing such chips BACKGROUND technical Multifunction integrated optical chips are made particularly in lithium niobate (LiNbO ₃₎ is at a temperature in the range of about 0.99 ° C. to about 250 ° C., about 10 Subjecting the chip to proton exchange with an acid such as pure benzoic acid for a period of time ranging from minutes to about 50 minutes or more. Proton exchange process at high temperature,
Or, when performed for an extended period of time, or both, the lithium niobate substrate tends to create internal stresses that cause microcracking and / or waveguide cross hatching in the substrate region exposed to the proton exchange step. As a result, chips manufactured by methods that include this step often exhibit refractive index instability, high propagation losses, and lower than desired electro-optic coefficients.

[0003] Cheng et al., US Pat. No. 5,193,7.
No. 19 and Cheng et al., US Pat. No. 5,442,7.
No. 19 discloses the manufacture of lithium niobate waveguides and related materials using a single proton exchange step followed by a single thermal anneal step. If the proton exchange step is performed in a concentrated acid (eg, benzoic acid) melt or a dilute melt of an acid containing a small amount (not more than about 2.5%) of the corresponding lithium salt, the β-phase in the crystalline phase change known as to form a layer of protonated lithium niobate _{(H x Li 1-x NbO} 3). This phase (comprising the sub-phases β _{1 to} β ₄ ) is characterized by high losses and poor electro-optical properties. The thermal annealing step, which causes the protons to diffuse deeper into the substrate, produces a waveguide with low losses and good electro-optical properties. At the same time, several crystalline phase transitions occur. The goal of this method is to obtain a waveguide with a crystal phase that provides a depth, width index gradient and low loss and good electro-optical properties.
These desirable operating characteristics are generally in the α phase,
And to a lesser extent in the kappa ₁ phase.

Also, a dilute acid melt containing from about 2.5% to about 5% of the corresponding lithium salt can be used to form a layer of substantially alpha-phase protonated lithium niobate. This type of proton exchange step eliminates the need for thermal annealing.

[0005]

The above method is characterized in that it is difficult to simultaneously and repeatedly obtain optimum values of desired parameters in a manufacturing environment having a single proton exchange step and a single thermal annealing step. . Several factors contribute to this problem.

First, in both the proton exchange and thermal annealing steps, precise temperature control is required to define the final waveguide proton content, proton diffusion depth and surface refractive index change (Δn). You. Such control would result in very long proton exchange and annealing times,
Deep waveguides, such as those for long wavelengths, become increasingly difficult.
As a result, a range of proton diffusion depths and Δn values can be obtained for each wafer, primarily under the constraints of process parameter control.

In addition, even within a single rough, it is known that there is wafer-to-wafer variability in lithium niobate supply, which changes the proton diffusion rate. This further widens the range of waveguide quantities and dimensional parameters during the manufacturing process.

Furthermore, the annealing step actually consists of two simultaneous processes. There are several crystal phase transitions with internal diffusion of protons and back diffusion of lithium ions. The annealing time required for a certain proton amount at a certain depth depends on the composition of the phase of the original proton exchange layer. If more than one phase is formed in the proton exchange process, and if its relative amount changes between wafers, a variation in the final waveguide parameters will occur. Any initial variation will result in an unacceptable range of final waveguide parameters in a well controlled manner.

[0009]

SUMMARY OF THE INVENTION The present invention, in a preferred embodiment, provides a proton-exchanged multifunctional integrated optical chip based on the desired κ and α phases of the rhombohedral H _x Li _1-x NbO ₃ structure. MIOC). These chips have substantially diffused protons, are substantially free of microcracking and the internal stresses that cause microcracking, have optimally high electro-optic coefficients and low losses. The method is particularly well adapted for the production of such chips from proton exchanged lithium niobate.

The process of the present invention is an improvement over the process disclosed and claimed in Chang et al., US Pat. No. 5,193,136, which is incorporated herein by reference and whose disclosure is incorporated by reference. Therefore, it is incorporated in the present application.
As disclosed in Chang et al., US Pat. No. 5,193,136, conventional method techniques include (a) providing a wafer made of a material such as lithium niobate;
(B) applying a removable mask or mask pattern to at least one surface of the wafer to form one or more proton exchange patterns of desired size and shape on the wafer surface; (c) removing the masked wafer. The temperature and time are sufficient to cause substantial proton exchange at the unmasked wafer surface and below it,
Treating with a proton exchange acid such as benzoic acid for a time insufficient to create microcracking and internal stresses leading to microcracking in the wafer; (d) removing the mask and mask pattern from the wafer; and (e) A) in an oxygen-containing environment, at a temperature and for a time sufficient to allow hydrogen ions at and near the surface of the wafer to diffuse substantially below the surface and sufficient to restore and optimize the electro-optic properties of the wafer; Time and temperature
And thermally annealing the wafer.

[0011] The present invention provides for "tuning" of the proton exchange and thermal anneal steps to provide accurate wafer-to-wafer in a manufacturing environment where the limiting factors are the accuracy of control of process parameters and the uniformity of wafer material from wafer to wafer. It is based on the discovery that waveguides with highly controllable and reproducible dimensions and refractive index parameters can be manufactured. This "tune"
Also, compensation for differences in the rate of proton diffusion during the proton exchange and thermal anneal steps can be accomplished by deliberately stopping each of these steps prior to the entire intended period, and removing one or more waveguides on each wafer. The analysis technique is based on the idea that it can be achieved by processing each wafer in the final stage of the process stage for the time indicated by the analysis evaluation result. The final stage of each process is typically much shorter than the entire intended period, allowing for more precise control of targeted dimensional and refractive index values.

The analysis technique defines the depth (d), width and surface refractive index change (Δn) of the waveguide. Δn value is
Provides information about the current crystalline phase of the process step. Tuning techniques also require knowledge of the diffusion constant at the end of each "tuning" process step. In the manufacturing process, the diffusion constant is usually determined empirically.

Therefore, the production method of the present invention can be summarized as comprising the following series of steps. (A) preparing a wafer made of a material such as lithium niobate;
(B) applying a removable mask or mask pattern to at least one surface of the wafer to form one or more proton exchange patterns of a desired size and shape on the wafer surface; Treating the unmasked portion with a proton exchange procedure for a time less than a predetermined longest proton exchange time; (d) analyzing the masked wafer, wherein certain parameters of the unmasked portion are first predetermined; Determining whether a value has been achieved, (e) repeating the proton exchange step and analyzing the masked wafer until a first predetermined value is achieved, and (f) removing the mask or mask pattern from the wafer. (G) thermally annealing the wafer in an oxygen-containing environment for a time shorter than the longest annealing time; Analyzing the unprocessed wafer to determine whether certain parameters of the wafer have achieved a second predetermined value; and (i) analyzing the unmasked wafer and the unmasked wafer until the second predetermined value is achieved. This is a step of repeating the step of performing.

If a dilute melt proton exchange process is used that can directly form the κ and α phases with low loss and good electro-optical properties, the last three steps [annealing step (g), unmasked wafer The analyzing step (h) and the repeating step (i)] may be omitted. this is,
This is usually the case when benzoic acid, where lithium benzoate is present at a concentration greater than about 2.5%, is a melt.

In a preferred embodiment of the method, a particular parameter is the proton diffusion depth. The analysis technique used for masked wafer analysis preferably includes prism coupling, and the unmasked wafer analysis
Preferably, it includes prism coupling and phase contrast ("Nomarski") microscopy. These analytical techniques are well known in the art, and their utility in the present invention lies in their use to determine the degree of progress in the desired diffusion method for the particular waveguide set being evaluated. Although not required, another valuable technique is infrared spectroscopy, in which the position of the O—H stretching band can tell the resulting crystal phase. Other non-destructive analytical techniques may prove useful.

To use the method of the present invention, the wafer is:
It is necessary to have at least one test area or waveguide of sufficient size to utilize prism coupling. This gives the depth and Δn parameters of the region,
These parameters can be empirically or theoretically related for a channel waveguide to be patterned on a wafer. Phase contrast or Nomarski microscopes can depict annealed and proton exchanged waveguides. With this means, the width of the waveguide can be measured using crosshairs, verniers, CCD imaging or measurements from photographs. Features of the waveguide surface such as cross hatching can also be described, and if there is sufficient latitude in the refractive index and waveguide dimensions, it can be removed by further annealing.

In a preferred embodiment, the wafer to be converted to MIOC is covered on one or more surfaces with a mask or masking pattern, such as a mask made of a metal such as aluminum, gold or chromium.
These masks preferably include a pattern of a size and shape sufficient to form the desired pattern of electro-optic channels in the wafer.

After masking, the lithium niobate wafer is treated with an acid that causes proton exchange between the hydrogen ions in the acid and the mobile cations in the wafer, such as lithium ions, so that the niobium proton exchange method is used. Performed on lithium oxide wafers. The acid treatment is preferably performed at an elevated temperature in the range of about 100 ° C. to about 300 ° C., for a time ranging from a few minutes to 30 hours or more, depending on the concentration of the acid melt (above about 250 ° C.). At temperature, a closed system would be required). The temperature and time of the proton exchange is generally sufficient to cause substantial proton exchange in the desired portion of the wafer, but stresses that result in microcracking or the formation of microcracks in the areas of the wafer that are exposed to the proton exchange Short enough to avoid the formation of

The proton exchange step is controlled as a result of the masked wafer analysis step and, depending on the wavelength at which the waveguide is to be operated, does not exceed about 5 microns below the wafer surface, preferably A first predetermined value of a proton diffusion depth not exceeding about 1 micron can be achieved. For this reason, the longest proton exchange time is preferably limited not to exceed about 30 hours for a 1% dilute melt of lithium benzoate in benzoic acid.

After the proton exchange has been completed (ie, if the first predetermined value, the proton diffusion depth, has been achieved),
The mask or masking pattern is removed. Thereafter, the wafer is heat-treated, that is, thermally annealed at an appropriate temperature and time in an air environment to diffuse hydrogen ions formed at and near the surface of the wafer substantially below the wafer surface. Due to the heat treatment, when the wafer is made of lithium niobate, the phase of the proton-exchanged portion of the wafer is also converted from the initially formed β phase to a κ phase or an α phase having low loss and good electro-optical properties. I do.

In the annealing step, the proton concentration on the wafer surface is reduced from a value in the range of about 44% to 65% to about 0% to 35%.
And preferably at a time and at a temperature sufficient to reduce it to a value of less than about 20%. The annealing step is controlled as a result of the unmasked wafer analysis step, and the waveguide can achieve a second predetermined value of about 2 to 5 microns of proton diffusion depth depending on the wavelength at which the waveguide is designed. The second predetermined value is defined as a depth at which the proton concentration becomes 1 / e times the surface proton concentration. In a broader sense, the annealing step is continued until the electro-optic coefficient of the wafer prior to proton exchange is substantially restored, and light propagation losses caused by proton exchange are reduced or substantially eliminated.

In a preferred embodiment, the anneal is performed by increasing the anneal temperature to room temperature, eg, from about 20 ° C. to a desired maximum temperature, preferably in the range of about 300 ° C. to about 400 ° C., for a time of up to about 30 minutes. It is carried out over time, but preferably as quickly as possible. The anneal causes the desired diffusion of protons into the wafer, and the time sufficient to restore the electro-optic coefficient of the wafer to its optimal value, i.e., the value prior to exposure to proton exchange and annealing, up to a maximum. Continue with temperature.

[0023]

Referring to the drawings, FIG. 1, which is a block diagram, illustrates a manufacturing process or method according to a preferred embodiment of the present invention. in this way,
A wafer 20 (shown in cross-section in FIGS. 2A and 2B), preferably made of lithium niobate or lithium tantalate, is provided (hereinafter a method using a lithium niobate wafer is described. The method using a wafer is substantially the same). After being cleaned in step 1, the wafer 20 proceeds to masking step 2.
Masking places a mask or mask pattern 22 (FIGS. 2A and 2B) of aluminum, gold or chromium, dimensionally controlled by photolithographic techniques, on a lithium niobate wafer 20, typically in the form of, for example, a Y-junction. Forming a waveguide pattern that is:

From the masking step 2, the masked lithium niobate wafer 20 proceeds to a proton exchange step 3. In this step, as shown in FIGS. 2A and 2B, a pure acid (typically benzoic acid) or a dilute acid (typically
The acid melt 24, which is either benzoic acid (including lithium benzoate), affects the entire surface of the masked wafer 20. However, since the mask 22 blocks all the masked portions of the wafer 20, the protons (H ⁺
The exchange between hydrogen ions) and lithium ions (Li ⁺ ) is limited to the unmasked portion, and one of the portions is shown in FIG.
In A and 2B, it is indicated by numeral 26. The goal of proton exchange step 3 is to achieve a first target proton diffusion depth d _P1 below the surface of the wafer 20, which is preferably about 1 micron. In the unmasked portion 26, protons enter the wafer 20 via path 28 and lithium ions leave the wafer 20 via path 30. After this step, the proton exchanged wafer 20 includes a proton exchanged region 32 as shown in FIG. This region has a depth d and a proton concentration proportional to the product of the refractive index difference Δn between the proton exchanged region 32 (measured at its surface) and the surrounding wafer 20 (ie, d × Δn). Or have the quantity. As will be seen below, the proton-exchanged region 32 becomes a waveguide region in the finished product after the thermal annealing step.

Proton Exchange Tuning In the proton exchange tuning process step, a proton exchange process designed based on knowledge of the diffusion constant of a certain acid composition 24 may be used to generate a first or intermediate target proton diffusion depth d _P1. Need to end at or just before the time needed to get This time is defined as the longest proton exchange time. In a well-controlled manufacturing method, there are usually variations in depth and Δn due to several factors. One of the factors is the variation in surface composition, cleanliness or roughness between wafers, which results from wafer polishing, cleaning or other processing prior to proton exchange. Another factor is the thickness of the mask, which affects the diffusion rate in unmasked, narrow waveguide channels. Other factors that are active in the long, dilute melt proton exchange process are due to external diffusion of lithium ions from the wafer and fractionation in an unsealed reactor (such as lithium benzoate / benzoic acid from a benzoic acid melt). Changes in the melt composition.

The next step in proton exchange tuning is masked wafer analysis step 4 (FIG. 1), which measures the proton diffusion depth and Δn value of the proton exchanged region 32 by prism coupling. If these values, corresponding to the step distribution at this point, are within the limits specified by the desired method, the wafer may be sent to the next step. That is,
For example, if the measured proton diffusion depth d is the first (proton exchange) target depth d _P1 plus or minus a predetermined allowable range, the process proceeds to the thermal annealing step 5. If the amount of protons in the proton exchanged region 32 is insufficient, the wafer 20
During the additional proton exchange "tune" time, determined by the melt diffusion constant and the desired amount of additional protons.
It can be immersed again in the acid melt 24. This step is additively, additional proton exchange time delta _tP may be determined from a desired depth by the following _{equation, (1) Δt P = 0.25} (d P1 2 -d P0 2) / D P1 In the above equation, Δt _P is the additional proton exchange time, d
_P1 is the first target proton diffusion depth, d _P0 is the initial proton diffusion depth and D _P1 is the expected diffusion constant for the repeated (tuned) proton exchange step. If necessary, the first target depth d _P1 with the required accuracy can be achieved using several continuous tuning steps.

In this tuning procedure, the proton exchange region is first
Formed in pure acid melt, and then in dilute melt
It can also be used when tuned,
Very long proton exchange times are required during demolition,
This is advantageous when a very deep waveguide is desired. Pure cheap
From the evaluation of the Δn value when the benzoic acid melt was used,
The waveguide obtained after tuning in the product process has a single-phase
Wachi β₁), But the first pure melting
Two original proton-exchanged regions from the product process
Phase combination (β₁And β_Two)
I understood that. For this reason, the two tuning steps
Would need to be returned, because β _TwoLink to phase
The interstitial protons are replaced during the process by optically active
Because it moves to the site and increases the measured Δn.
You. Initial interstitial proton concentration must be determined independently
If so, the length of a single tuning step cannot be accurately calculated.
The length of the tuning step is such that all protons are optically active
You can calculate accurately for the first time. Annealing tuning

When the target proton diffusion depth d _P1 is obtained,
The mask 22 is removed and the proton-exchanged wafer 20 is removed.
Goes to the thermal annealing step 5. In this step, the unmasked wafer 20 is heated from about 20.degree.
Thermal annealing at progressively increasing temperatures to a desired maximum temperature in the range of 00 ° C. to about 400 ° C. for a time of up to about 30 minutes, but preferably as fast as possible to achieve actual annealing time and temperature accuracy. Is done.

The annealing tuning process step requires that the first planned annealing step end at or shortly before the second (annealed) target proton diffusion depth d _A1 is obtained. . The proton concentration after thermal annealing follows a Gaussian or near Gaussian distribution curve from the maximum value on the wafer surface. The second target proton diffusion depth d _A1 can be defined as a well-defined point on the concentration curve. As a practical matter, the annealing target proton diffusion depth d _A1 is
It is preferably defined as the depth at which the proton concentration becomes 1 / e times the proton concentration on the wafer surface, or the depth at which 1 / e ² times the surface concentration. Preferably, an annealing target depth d _{A1 in} the range of about 2 to 5 microns is desired, depending on the wavelength at which the waveguide is designed.

Korishko et al. , "T
he Phase Diagram of H _x Li _1-x N
bO ₃ Optical Waveguides ", S
PIE, Vol. 2997, pp. 139-154. 188-200 (1
As described in US Pat. No. 997), it has been found that at the beginning of the annealing step, the optical parameters do not obey the square root diffusion law, presumably due to transitions between the various phases. Thus, annealing tuning, as the end point of the first annealing process approaches the final desired waveguide phase (kappa ₁ or alpha), it becomes effective.

At this point, there are usually variations in proton depth and Δn due to several factors. One of the factors is the initial proton exchange depth (first target depth d
_P1 ) and Δn. Another factor is the non-uniformity of the substrate composition between the wafers. Another factor is the non-uniformity of temperature control between wafers in the annealing process itself, especially when the annealing time is long. This may be due to fluctuations in the oven temperature itself or positional non-uniformity of the temperature within the oven. In the manufacturing process, the additional cost of improving the thermal uniformity of the annealing process to the required accuracy must be weighed against the cost of the extra thermal annealing process.

The next step in annealing tuning is an unmasked wafer analysis step 6 (FIG. 1) that requires measurement of the proton diffusion depth and .DELTA.n of the annealed proton exchange or waveguide region 32 by prism coupling. ). In addition, a phase contrast or Nomarski microscope is used here to measure the width of these regions. All waveguide regions or selected regions on the wafer 20 can be measured. At this point, it is possible to evaluate the surface stress on the waveguide region, but this is observed as a cross-hatched pattern that appears on some waveguide regions, and the high-level thin film remaining on the surface It depends on the layer. This pattern usually disappears when further annealing is performed.
Record the width of the waveguide region and adjust the additional anneal time (calculated to obtain a second (anneal) target depth d _A1 and Δn) so as not to exceed the waveguide design during the tuning process I do.

After the initial annealing time, if the measured proton diffusion depth d is equal to the target annealing depth d _A1 plus or minus a predetermined tolerance, the process proceeds to the final manufacturing process described below. If the measured depth is less than the anneal target depth, the wafer 20 can be treated with a thermal anneal for an additional “tuning” time Δt _A calculated by the following equation similar to equation (1): (2) Δt _A = 0.25 (d _A1 ² −d _A0 ² ) / D _{A1 In the} above equation, Δt _A is an additional thermal annealing time, d _A0
Is the proton diffusion depth after the first thermal anneal, d _A1 is the second (anneal) target depth and D _A1 is the expected diffusion constant for the repeated (tuned) anneal step. If necessary, the second annealing target depth d _A1 can be achieved with the required accuracy using several intermediate and continuous tuning steps.

If the diffusion constants D _A1 and D _{A0 of the process} are known and no phase change occurs, the additional annealing time can be calculated using equation (2) above.

As described above, the annealing step produces a waveguide having a Gaussian or near-Gaussian-type concentration distribution in the depth direction. Therefore, in order to obtain a desired 1 / e or 1 / e ² depth value, The prism combination data must be fitted to a suitable distribution. Several continuous tuning steps are used, depending on the accuracy of the endpoint. Additional information, such as information obtained from microscopic observation of the waveguide, can be used to determine whether additional annealing time is required or unnecessary. The increase in waveguide width with annealing time varies from wafer to wafer, but a good measure of relative speed can be calculated from values measured after the initial annealing time.

Referring again to FIG. 1, once the thermal anneal tuning is completed, the wafer 20 proceeds to a photolithography step 7, in which electrodes are formed on the wafer, if necessary, using photolithography. You.
The wafer 20 then proceeds to a dicing, polishing and packaging step 8 in which the wafer 20 is cut into products of desired size and shape, and then to a finishing step 9.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a method for manufacturing a proton-exchanged integrated optical chip according to a preferred embodiment of the present invention.

FIG. 2A shows the proton exchange step of the method of the present invention.

FIG. 2B shows the proton exchange step of the method of the present invention.

FIG. 3 is a cross-sectional view of an embodiment of a product manufactured by the method of the present invention.

Continued on the front page (72) Inventor Henry Abink United States. 91361 Westlake Village, Fallview Road 1884 (72) Inventor Rory Lane Gump United States. 91335 California, Reseda, Number 6, Sherman Way 19610 (72) Inventor Frances Gary Augeri United States. 90069 California, West Hollywood, Laraby Street 928

Claims (29)

[Claims] 1. A method for manufacturing an integrated optical chip, comprising: (a) preparing a wafer made of a material selected from the group consisting of lithium niobate and lithium tantalate; and (b) masking the wafer surface. Forming one or more proton exchange patterns on the masked wafer surface, each having a desired size and shape, comprising unmasked portions of the masked wafer surface; and (c) masking the masked surface. In the parts that are not
Applying a proton exchange treatment for a time shorter than a predetermined longest proton exchange time; (d) analyzing the masked wafer and determining whether certain parameters of the proton exchanged portion have attained a first predetermined value; (E) repeating the proton exchange treatment and analyzing the masked surface until a first predetermined value is achieved; (f) removing the mask from the wafer; (g) longest Thermally annealing the wafer for less than an annealing time; (h) analyzing the unmasked wafer to determine if a particular parameter has achieved a second predetermined value; and (i) a second Repeating the steps of annealing and analyzing the unmasked wafer until a predetermined value is achieved. 2. The method according to claim 1, wherein the specific parameter is a proton diffusion depth below a wafer surface, a first predetermined value of the specific parameter is a first target depth of proton diffusion, The method of claim 1, wherein the predetermined value is a second target depth of proton diffusion. 3. The method of claim 2, wherein said analyzing the masked wafer and analyzing the unmasked wafer comprises using prism coupling. 4. The method of claim 2, wherein the step of analyzing the unmasked wafer comprises using phase contrast microscopy. 5. The method of claim 3, wherein analyzing the unmasked wafer comprises using phase contrast microscopy. 6. The method of claim 1, wherein at least one of said analyzing steps comprises using infrared spectroscopy. 7. The method of claim 1, wherein said proton exchange step comprises treating said wafer in a melt comprising benzoic acid. 8. The step of repeating the above proton exchange treatment is performed for a time Δt _P determined by the following equation: Δt _P = 0.25 (d _P1 ² −d _P0 ² ) / D _P1 , Δt _P is the additional proton exchange time, d
3. The method of claim 2, wherein _P1 is a first target proton diffusion depth, _dP0 is an initial proton diffusion depth and _Dp1 is an expected diffusion constant for repeated proton exchange treatments. 9. The step of repeating the thermal annealing step is performed for a time Δt _A determined by the following equation: Δt _A = 0.25 (d _A1 ² −d _A0 ² ) / D _A1 , Δt _A is the additional thermal annealing time, d _A0
3. The method of claim 2, wherein is the proton diffusion depth after the first thermal anneal step, d _A1 is the second target proton diffusion depth and D _A1 is the expected diffusion constant for the repeated anneal steps. . 10. The method of claim 1, wherein after the anneal step, the proton concentration of the portion subjected to the proton exchange treatment as a function of the depth below the wafer surface is close to a Gaussian distribution from the maximum concentration value near the wafer surface. 3. The method of claim 2, wherein the target depth is measured as the depth at which the proton concentration is about 1 / e times the maximum concentration value. 11. The method of claim 10, wherein the second target depth ranges from about 2 to about 5 microns. 12. A method for manufacturing an integrated optical chip, comprising: (a) preparing a wafer made of a material selected from the group consisting of lithium niobate and lithium tantalate; and (b) masking the wafer surface. Forming one or more proton exchange patterns on the masked wafer surface, each having a desired size and shape, comprising an unmasked portion of the masked wafer surface; and (c) subjecting the wafer to benzoic acid. Subjecting the unmasked portion of the masked surface to a proton exchange treatment by treating with a melt comprising for a time less than a predetermined longest proton exchange time; (d) masking using a prism bond Analyzing the wafer to determine whether the proton diffusion depth of the proton exchanged portion has reached a first predetermined value; C) repeating the proton exchange procedure and analyzing the masked surface until a first target depth value is achieved; f) removing the mask from the wafer; g) a time shorter than the longest anneal time; Thermally annealing the wafer; (h) analyzing the unmasked wafer using prism bonding to determine whether the proton diffusion depth of the proton exchanged portion has reached a second target depth value. Determining and (i) repeating the annealing and analyzing the unmasked wafer until a second target depth value is achieved. 13. The method of claim 1, wherein analyzing the unmasked wafer comprises using phase contrast microscopy.
3. The method according to 2. 14. The method of claim 12, wherein at least one of said analyzing steps comprises using infrared spectroscopy. 15. step of repeating the proton exchange treatment of the can, is time implementation of Delta] t _P which is determined by the following equation, in ^{_{Δt P = 0.25 (d P1 2}} -d P0 2) / D P1 above formula , Δt _P is the additional proton exchange time, d
13. The method of claim 12, wherein _P1 is a first target proton diffusion depth, _dP0 is an initial proton diffusion depth, and _Dp1 is an expected diffusion constant for repeated proton exchange treatments. 16. The step of repeating the thermal annealing step is performed for a time Δt _A determined by the following equation: Δt _A = 0.25 (d _A1 ² −d _A0 ² ) / D _A1 , Δt _A is the additional thermal annealing time, d _A0
13. The method of claim 12, wherein is the proton diffusion depth after the first thermal annealing step, d _A1 is the second target proton diffusion depth and D _A1 is the expected diffusion constant for the repeated annealing step. . 17. The method according to claim 16, wherein after the annealing step, the proton concentration of the portion subjected to the proton exchange treatment as a function of the depth below the wafer surface is close to a Gaussian distribution from a maximum concentration value near the wafer surface, and the second 13. The method of claim 12, wherein the target depth is measured as the depth at which the proton concentration is about 1 / e times the maximum concentration value. 18. The method of claim 17, wherein the second target depth is in a range from about 2 to about 5 microns. 19. A method for performing proton exchange on a selected area of a wafer using a melt of benzoic acid, wherein the wafer is formed from a material selected from the group consisting of lithium niobate and lithium tantalate. It is made,
The method comprising: (a) subjecting a selected area of the wafer to an initial proton exchange treatment by treating the wafer in the melt for less than a predetermined longest proton exchange time; (b) A) analyzing said selected area to determine whether certain parameters of the selected area have achieved a predetermined value; and (c) proton exchange treatment and analysis steps until said predetermined value is achieved. Repeating the steps of: 20. The method according to claim 19, wherein the specific parameter is a proton diffusion depth below the wafer surface, and the predetermined value of the specific parameter is a target depth of proton diffusion. 21. The method according to claim 20, wherein said analyzing step comprises using prism coupling. 22. A process of repeating a proton exchange treatment of the can, is time implementation of Delta] t _P which is determined by the following equation, in ^{_{Δt P = 0.25 (d P1 2}} -d P0 2) / D P1 above formula , Δt _P is the additional proton exchange time, d
_21. The method of claim 20, wherein _P1 is a first target proton diffusion depth, _dP0 is an initial proton diffusion depth and _Dp1 is an expected diffusion constant for repeated proton exchange treatments. 23. A method for thermally annealing a proton-exchanged region of a wafer made of a material selected from the group consisting of lithium niobate and lithium tantalate, comprising: (a) initially shorter than a longest annealing time; Time, thermally annealing the wafer; (b) analyzing the proton exchanged region to determine whether a particular parameter has achieved a predetermined value; and (c) until the predetermined value is achieved. Repeating the annealing step and the analyzing step. 24. The method of claim 23, wherein the specific parameter is a depth of proton diffusion below the wafer surface, and wherein the predetermined value of the specific parameter is a target depth of proton diffusion. 25. The method of claim 23, wherein said analyzing step comprises using prism coupling. 26. The method of claim 23, wherein said analyzing step comprises using phase contrast microscopy. 27. The step of repeating the thermal annealing step is performed for a time Δt _A determined by the following equation: Δt _A = 0.25 (d _A1 ² −d _A0 ² ) / D _A1 , Δt _A is the additional thermal annealing time, d _A0
25. The method of claim 24, wherein is the proton diffusion depth after the first thermal anneal step, d _A1 is the second target proton diffusion depth and D _A1 is the expected diffusion constant for the repeated anneal steps. . 28. The method according to claim 28, wherein after the annealing step, the proton concentration of the portion subjected to the proton exchange treatment as a function of the depth below the wafer surface is close to a Gaussian distribution from a maximum concentration value near the wafer surface, 25. The method of claim 24, wherein the target depth is measured as a depth at which the proton concentration is about 1 / e times the maximum concentration value. 29. The method of claim 24, wherein the second target depth ranges from about 2 to about 5 microns.