CA2413650A1 - Optical filters - Google Patents
Optical filters Download PDFInfo
- Publication number
- CA2413650A1 CA2413650A1 CA002413650A CA2413650A CA2413650A1 CA 2413650 A1 CA2413650 A1 CA 2413650A1 CA 002413650 A CA002413650 A CA 002413650A CA 2413650 A CA2413650 A CA 2413650A CA 2413650 A1 CA2413650 A1 CA 2413650A1
- Authority
- CA
- Canada
- Prior art keywords
- filter
- thickness
- comb
- spacing
- cavity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
- 230000003287 optical effect Effects 0.000 title claims abstract description 46
- 230000004044 response Effects 0.000 claims abstract description 52
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 claims abstract description 35
- 229910052710 silicon Inorganic materials 0.000 claims abstract description 35
- 239000010703 silicon Substances 0.000 claims abstract description 35
- 238000000034 method Methods 0.000 claims abstract description 24
- 230000005540 biological transmission Effects 0.000 claims abstract description 13
- 238000004519 manufacturing process Methods 0.000 claims description 7
- 238000009792 diffusion process Methods 0.000 claims description 5
- 238000005259 measurement Methods 0.000 claims description 5
- 238000005498 polishing Methods 0.000 claims description 4
- 238000000151 deposition Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 2
- 238000000059 patterning Methods 0.000 claims description 2
- 235000012431 wafers Nutrition 0.000 abstract description 20
- 238000000926 separation method Methods 0.000 description 16
- 238000012545 processing Methods 0.000 description 6
- 239000000463 material Substances 0.000 description 5
- 239000000835 fiber Substances 0.000 description 4
- 125000006850 spacer group Chemical group 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 229910052691 Erbium Inorganic materials 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000008021 deposition Effects 0.000 description 1
- UYAHIZSMUZPPFV-UHFFFAOYSA-N erbium Chemical compound [Er] UYAHIZSMUZPPFV-UHFFFAOYSA-N 0.000 description 1
- 238000002839 fiber optic waveguide Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G02—OPTICS
- G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
- G02B5/00—Optical elements other than lenses
- G02B5/20—Filters
- G02B5/28—Interference filters
- G02B5/284—Interference filters of etalon type comprising a resonant cavity other than a thin solid film, e.g. gas, air, solid plates
-
- G—PHYSICS
- G02—OPTICS
- G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
- G02B5/00—Optical elements other than lenses
- G02B5/20—Filters
- G02B5/28—Interference filters
- G02B5/281—Interference filters designed for the infrared light
Abstract
A multi cavity comb filter for interleaving or de-interleaving WDM signals has a plurality of stacked optical cavities each having substantially the same thickness.
The multiple cavity arrangement provides a comb reflection response and a comb transmission response with broad peaks, so that the filter can be used for transmitting one group of channels and reflecting another group of channels at interleaved positions. The cavities are preferably formed from silicon wafers, so that existing techniques can be employed to obtain specific cavity thicknesses with sufficient accuracy and uniformity.
The multiple cavity arrangement provides a comb reflection response and a comb transmission response with broad peaks, so that the filter can be used for transmitting one group of channels and reflecting another group of channels at interleaved positions. The cavities are preferably formed from silicon wafers, so that existing techniques can be employed to obtain specific cavity thicknesses with sufficient accuracy and uniformity.
Description
OPTICAL FILTERS
Field of the invention The present invention relates to optical filters, and particularly to filters which can be used for separating interleaved channels from a W DM signal or for interleaving groups of channels.
Back4round of the invention Wavelength division multiplexing (WDM) is a technique widely used in optical communications systems to allow different wavelengths to be carried over a common fiber (or fiber optic waveguide). The most commonly used wavelength band for fiber optic transmission is centered at 1550nm, because of the low absorption and the commercial availability of erbium doped fiber amplifiers which are effective for this band.
Wavelength division multiplexing can separate this band into multiple channels, typically 32 or 64 multiple discrete channels, through a technique referred to as dense channel wavelength division multiplexing (DWDM). This is used to increase long-haul telecommunication capacity over existing fiber optic transmission lines.
Techniques and devices are required, however, for multiplexing and de-multiplexing the different discrete carrier wavelengths. Multiplexing and de-multiplexing will typically take place not only at the source and destination of the data, but also at intermediate routing locations, particularly if a network provides per-channel routing capability and per-channel add and drop functions. Thus, each routing device requires the multiplexing and de-multiplexing capability.
Various types of optical multiplexer are well known for the combination or separation of optical signals in wavelength division multiplexed systems.
Known _ ? _ devices for this purpose have employed, for example, diffraction gratings, prisms and various types of fixed or tuneable filters.
Regardless of the type of multiplexer and de-multiplexer to be employed, as the channel separation is reduced in order to increase the system capacity, the multiplexer filtering characteristics need to be improved to ensure separation of channels without cross talk between adjacent channels.
Channel interleaving is one technique which has been used to enable two groups of channels to be processed separately, with the channels in each group having greater separation. This reduces the required optical performance of the optical components used within the processing equipment for each group of channels.
This invention is particularly directed to an optical filter which enables two groups of interleaved channels to be separated or combined.
IS
Summary of the invention According to the invention, there is provided an optical filter comprising a plurality of stacked optical cavities each having substantially the same thickness, the filter having a first frequency response for transmission through the filter comprising a 2o first comb response and a second frequency response for reflection from the filter and comprising a second comb response, the peaks of the first comb response lying between the peaks of the second comb response.
The multiple cavity arrangement in this filter enables the peaks in the reflection 25 response and the peaks in the transmission response to be broadened (with respect to a single cavity response), so that the filter can be used for transmitting one group of channels and reflecting another group of channels at interleaved positions. The cavities are preferably formed from silicon wafers, so that existing techniques can be employed to obtain specific cavity thicknesses with sufficient 30 accuracy and uniformity.
_ '3 _ The thickness of the cavities is selected as a function of the channel spacing, and determines the separation between the peaks of the two comb responses. The spacing between cavities is selected as function of the wavelength band over which the filter is to be used.
The thickness of each cavity may therefore be selected such that the spacing between peaks of each comb response corresponds to double the channel spacing in a W DM optical communications system in which the filter is to be used.
The spacing between adjacent cavities may be equal to one quarter of a wavelength within a band of wavelengths of channels of a WDM optical communications system in which the filter is to be used.
Preferably, adjacent cavities are separated by an air spacing. The large refractive index difference between silicon and air enables a small number of cavities to achieve the required filter response. For example, only three cavities can achieve the required pass band profile for DWDM optical communications systems.
The thickness of the silicon wafer of each cavity may be between 180 and 210 microns, and this is appropriate for use with a channel separation of 0.8nm which is the standard 100GHz grid. More preferably in this example, the thickness of the silicon wafer of each cavity is between 194.5 and 195.0 microns.
A support layer may be provided between the cavities for defining the air spacing, the support layer comprising an oxide layer.
This layer may define a spacing between adjacent cavities of 0.36 - 0.41 microns. This range corresponds to one quarter of the wavelength of signals in the C-band.
The invention also provides a method of manufacturing an optical filter, comprising:
preparing a silicon wafer having a predetermined thickness;
depositing an oxide layer over the silicon wafer;
patterning the oxide layer to define a plurality of pattern portions;
dividing the wafer into sections, each section comprising a pattern portion;
stacking a plurality of the sections to define a multiple cavity filter, the wafer sections each defining a cavity and the pattern portions defining a spacing between adjacent cavities.
This method enables a single silicon wafer to be used to form all cavities of the filter, thereby ensuring uniform thickness and enabling precise thickness control using conventional silicon wafer processing techniques. For example, double sided polishing may be used, and with feedback control based on optical thickness measurement, for example infrared interferometric sensing.
The invention provides a low cost process which can be accurately controlled.
The spacing between cavities is preferably determined by an oxide layer, which can also be deposited and patterned to the required accuracy using conventional semiconductor device processing technology.
The pattern portions may each comprise a central opening around which one or more support portions of predetermined thickness are provided. The opening then defines the air cavity in the assembled device, and the support portion provides the required cavity separation.
The silicon wafer is preferably prepared to have a thickness between 180 and 210 microns. The oxide layer may be deposited with a thickness of 0.32 - 0.44 microns, preferably 0.37 - 0.39 microns.
_5_ The stacked sections are preferably diffusion bonded.
The invention is also directed to the use of an optical filter of the invention for interleaving or de-interleaving two groups of channels of a WDM optical signal.
The invention also provides an optical communications system using a filter of the invention for interleaving or de-interleaving two groups of channels of a WDM
optical signal.
Brief description of the drawings Examples of the invention will now be described in detail with reference to the accompanying drawings, in which;
Figure 1 shows a filter in accordance with the invention;
Figure 2 shows the method of manufacturing the filter of Figure 1;
Figure 3 shows the transmission and reflection response of the filter of Figure 1;
Figure 4 shows the relationship between one peak of the filter response and a standard pass band of a multiplexer filter;
Figure 5 shows the relationship between one trough of the filter response and a standard rejection band of a multiplexer filter;
Figure 6 is used to show the dependency of the filter response on the silicon thickness;
Figure 7 is used to show the dependency of the filter response on the spacer layer thickness;
Figure 8 shows the filter used for de-interleaving and interleaving; and Figure 9 shows one possible use of the filter of the invention to provide add/drop functions.
Detailed description Figure 1 shows a filter 10 in accordance with the invention, comprising three stacked optical cavities 12 of equal thickness. The cavities are spaced by spacers 14. Each cavity provides a Fabry-Perot cavity frequency response, determined by the so-called "Airy" function. By providing a number of cavities, the transmission and reflection transfer function is altered to provide flatter peaks, so that a comb filter response results. Thus, the filter has a comb transmission frequency response and a comb reflection frequency response, with the peaks of the transmission comb response lying between the peaks of the reflection comb response.
As a result, the filter can be used to divide a W DM signal 20 comprising frequency components A1,I~2,~3... ~n into a first group of channels 22 transmitted through the filter, and a second group of channels 24 reflected by the filter.
For example, the first group of channels 22 has the odd channels ~1, h3 etc, and the second group of channels 24 has the even channels ~2, h4 etc.
The input W DM signal is provided at an angle to the filter, so that the reflected channels can be collected.
The thickness of the cavities 12 is selected as a function of the channel spacing, and determines the separation between the peaks of the transmission and reflection comb responses. In particular, the free Spectral Range (FSR) for a single cavity Fabry Perot is given by:
FSR = X02 /(2nd) where ao is the centre wavelength of operation, n is the refractive index of the cavity and d is the physical length of the cavity.
This equation can be used to derive the cavity length for a required FSR, which corresponds to the separation of the peaks in the comb response. For example, for a frequency corresponding to wavelength 1530nm (one standardised frequency is 1530.48nm) and with channel separation 0.8nm (giving an FSR of 1.6nm and corresponding to a 100GHz grid), with a silicon cavity (refractive index 3.63), we obtain a physical cavity length of 202 microns.
In practice, a slightly smaller thickness is required due to the angle of incidence, which may be, for example 3 degrees. Thus, the required cavity thickness for C
band operation using a 100GHz grid with a silicon wafer works out to be approximately 195 microns, particularly between 194.5 and 195.0 microns.
The spacing between cavities is selected to be one quarter wavelength of the wavelength over which the filter is to be used. For use in the C- band (1524nm -1563nm), the cavity separation is selected to be one quarter wavelength of any wavelength in this range, for example based on 1528nm, giving a thickness of 382nm.
For using the filter as a de-interleaving filter, the thickness of each cavity is selected so that the spacing between the comb peaks corresponds to double the channel spacing of the WDM optical communications system.
The frequency response of the filter is a function of the refractive index of the material defining the cavity and the material in the spacing. In particular, the greater the refractive index difference, the fewer cavities are required to obtain the required frequency response.
In the preferred embodiment of the invention, each cavity is defined by a silicon wafer, and adjacent cavities are separated by an air spacing. The use of silicon wafers enables existing techniques to be employed to obtain specific cavity thicknesses with sufficiently accuracy and uniformity.
Figure 2 shows the method of manufacturing the filter. Figure 2A shows a silicon wafer 30 prepared to have the desired thickness, using conventional silicon processing techniques, for example double-sided polishing controlled using optical thickness measurement. This optical thickness measurement preferably comprises infrared thickness sensing.
In Figure 2B, an oxide layer, for example Si02, is deposited over the silicon wafer to the desired spacing and is then patterned to define the regions 32.
These regions have an opening 34 and a support 36 surrounding the opening 34.
The oxide layer is patterned using conventional photolithographic techniques.
The wafer is then divided into sections 40 as shown in Figure 2C, each section comprising one of the regions 32. They are then physically stacked as shown in l0 Figure 2D and diffusion bonded as shown in Figure 2E. This diffusion bonding will compact the oxide layer, and the original deposition thickness of the oxide layer will take into account the change in thickness resulting from the diffusion bonding process.
Figure 3 shows the transmission and reflection response of the filter when used for a 50GHz grid. Of course, this requires a different thickness of silicon cavity.
The filter of the invention can, however, be tuned to any desired comb separation by calculating the thicknesses in the manner set out above.
The upward arrows 50 in Figure 3 indicate the transmitted frequency bands which together define a first comb frequency response, and the downward arrows 52 indicate the reflected frequency bands which together define a second comb frequency response. In this example, the channels have a separation of 0.4nm for the 50GHz grid. The filter rejection is around -23dB, and each comb peak has a reasonably flat attenuation of less than 1 dB over a wavelength band of around 0.3nm.
Figure 4 shows two adjacent peaks in one of the comb responses in a filter designed for the 100GHz grid, namely with a channel separation of 0.8nm, and shows that the comb filter response exceeds an example of a known band pass profile 60 required by multiplexing filters. Figure 5 shows that the rejection -C~-between channels of the comb filter response (around 27dB for a filter designed for 0.8nm channel separation, as shown) also exceeds an example of a known rejection profile 62 required by multiplexing filters.
The thickness of the silicon wafer should be accurately controlled during manufacture. Figure 6 shows the dependency of the filter response on the silicon thickness, and shows two theoretical plots 70,72 for filters with silicon thickness varying by 3nm. In practice, it is sufficient to control the silicon wafer thickness to within around ~l0nm from the desired thickness.
The thickness of the oxide spacer layer should also be accurately controlled during manufacture. Figure 7 shows the dependency of the filter response on the cavity spacing, and shows three theoretical plots 80, 82, 84 for filters with the cavity spacing at the desired dimension, 30nm from the desired dimension and 50nm from the desired dimension. In practice, it is sufficient to control the silicon wafer thickness to within around ~50nm from the desired thickness.
These thickness controls can be achieved using known processing techniques used in the semiconductor device manufacture industry.
Figure 8 shows how the filter of the invention can be used within a demultiplexer and multiplexes function. A filter of the invention 90 acts as a de-interleaves and divides the W DM signal into the odd channels 92 and the even channels 94.
Each of these groups is passed to a conventional de-multiplexes 96, 98, for example arrayed waveguide de-multiplexers. The channels at the input to these de-multiplexers have double the spacing of the W DM channels, so that the required optical performance of the de-multiplexers is lower.
In order to interleave channels, the filter of the invention 100 is again employed, with the direction of signals through the filter reversed. The interleaves 100 combines the channels from the two multiplexers 102,104.
Figure 9 shows the use of the filter of the invention to provide add/drop capability in a node. The input 120 carries the WDM signal, and in this example, add/drop capability is only required for a sub-set of the channels, and this capability is provided for the group of channels 122.
After de-multiplexing at 124, the individual channels from the group 122 are each supplied to an optical switch 126. The switch enables a channel from the demultiplexer 124 to be dropped or else passed forward. Alternatively, the switch 126 can enable an external signal 128 to be added. Each switch comprises a 2x2 optical switch, and enables add or drop functions to be implemented for each channel.
The channels passing forwards (either added or passed through) are each passed through a variable optical attenuator (VOA) 130 and then combined in a multiplexes 132. The VOAs ensure the signals have optical powers appropriate for the system, for example dependent upon the receiver characteristics within the system.
In the example of Figure 9 add/drop capability is provided for one group of channels 122, and the other group of channels 134 is passed through the node and recombined with the output from the multiplexes 132 at a filter of the invention 136 arranged to provide an interleaving function.
The filter of the invention can be used in other optical processing devices, and essentially enables channels to be divided into groups so that channels with greater separation can be further processed.
The specific example of filter of the invention uses a silicon cavity, and this is the preferred material to enable the thickness to be processed accurately.
However, other materials may be used, and this will result in different thickness requirements as a result of the different refractive index. A material may be provided between the cavities instead of the air spacing described above.
Similarly, the filter of the invention can be used in a range of wavelength bands.
In the specific examples above, the filter is used in the C-band. However, technologies are being developed enabling the L-band and S-band to be used in optical communications systems, and the filter of the invention can be used for these wavelength bands. Again, this wilt alter the required layer thicknesses.
Thus, it will be understood that the specific layer thicknesses given are in respect of the one specific example of the use of the filter.
Various other modifications will be apparent to those skilled in the art.
Field of the invention The present invention relates to optical filters, and particularly to filters which can be used for separating interleaved channels from a W DM signal or for interleaving groups of channels.
Back4round of the invention Wavelength division multiplexing (WDM) is a technique widely used in optical communications systems to allow different wavelengths to be carried over a common fiber (or fiber optic waveguide). The most commonly used wavelength band for fiber optic transmission is centered at 1550nm, because of the low absorption and the commercial availability of erbium doped fiber amplifiers which are effective for this band.
Wavelength division multiplexing can separate this band into multiple channels, typically 32 or 64 multiple discrete channels, through a technique referred to as dense channel wavelength division multiplexing (DWDM). This is used to increase long-haul telecommunication capacity over existing fiber optic transmission lines.
Techniques and devices are required, however, for multiplexing and de-multiplexing the different discrete carrier wavelengths. Multiplexing and de-multiplexing will typically take place not only at the source and destination of the data, but also at intermediate routing locations, particularly if a network provides per-channel routing capability and per-channel add and drop functions. Thus, each routing device requires the multiplexing and de-multiplexing capability.
Various types of optical multiplexer are well known for the combination or separation of optical signals in wavelength division multiplexed systems.
Known _ ? _ devices for this purpose have employed, for example, diffraction gratings, prisms and various types of fixed or tuneable filters.
Regardless of the type of multiplexer and de-multiplexer to be employed, as the channel separation is reduced in order to increase the system capacity, the multiplexer filtering characteristics need to be improved to ensure separation of channels without cross talk between adjacent channels.
Channel interleaving is one technique which has been used to enable two groups of channels to be processed separately, with the channels in each group having greater separation. This reduces the required optical performance of the optical components used within the processing equipment for each group of channels.
This invention is particularly directed to an optical filter which enables two groups of interleaved channels to be separated or combined.
IS
Summary of the invention According to the invention, there is provided an optical filter comprising a plurality of stacked optical cavities each having substantially the same thickness, the filter having a first frequency response for transmission through the filter comprising a 2o first comb response and a second frequency response for reflection from the filter and comprising a second comb response, the peaks of the first comb response lying between the peaks of the second comb response.
The multiple cavity arrangement in this filter enables the peaks in the reflection 25 response and the peaks in the transmission response to be broadened (with respect to a single cavity response), so that the filter can be used for transmitting one group of channels and reflecting another group of channels at interleaved positions. The cavities are preferably formed from silicon wafers, so that existing techniques can be employed to obtain specific cavity thicknesses with sufficient 30 accuracy and uniformity.
_ '3 _ The thickness of the cavities is selected as a function of the channel spacing, and determines the separation between the peaks of the two comb responses. The spacing between cavities is selected as function of the wavelength band over which the filter is to be used.
The thickness of each cavity may therefore be selected such that the spacing between peaks of each comb response corresponds to double the channel spacing in a W DM optical communications system in which the filter is to be used.
The spacing between adjacent cavities may be equal to one quarter of a wavelength within a band of wavelengths of channels of a WDM optical communications system in which the filter is to be used.
Preferably, adjacent cavities are separated by an air spacing. The large refractive index difference between silicon and air enables a small number of cavities to achieve the required filter response. For example, only three cavities can achieve the required pass band profile for DWDM optical communications systems.
The thickness of the silicon wafer of each cavity may be between 180 and 210 microns, and this is appropriate for use with a channel separation of 0.8nm which is the standard 100GHz grid. More preferably in this example, the thickness of the silicon wafer of each cavity is between 194.5 and 195.0 microns.
A support layer may be provided between the cavities for defining the air spacing, the support layer comprising an oxide layer.
This layer may define a spacing between adjacent cavities of 0.36 - 0.41 microns. This range corresponds to one quarter of the wavelength of signals in the C-band.
The invention also provides a method of manufacturing an optical filter, comprising:
preparing a silicon wafer having a predetermined thickness;
depositing an oxide layer over the silicon wafer;
patterning the oxide layer to define a plurality of pattern portions;
dividing the wafer into sections, each section comprising a pattern portion;
stacking a plurality of the sections to define a multiple cavity filter, the wafer sections each defining a cavity and the pattern portions defining a spacing between adjacent cavities.
This method enables a single silicon wafer to be used to form all cavities of the filter, thereby ensuring uniform thickness and enabling precise thickness control using conventional silicon wafer processing techniques. For example, double sided polishing may be used, and with feedback control based on optical thickness measurement, for example infrared interferometric sensing.
The invention provides a low cost process which can be accurately controlled.
The spacing between cavities is preferably determined by an oxide layer, which can also be deposited and patterned to the required accuracy using conventional semiconductor device processing technology.
The pattern portions may each comprise a central opening around which one or more support portions of predetermined thickness are provided. The opening then defines the air cavity in the assembled device, and the support portion provides the required cavity separation.
The silicon wafer is preferably prepared to have a thickness between 180 and 210 microns. The oxide layer may be deposited with a thickness of 0.32 - 0.44 microns, preferably 0.37 - 0.39 microns.
_5_ The stacked sections are preferably diffusion bonded.
The invention is also directed to the use of an optical filter of the invention for interleaving or de-interleaving two groups of channels of a WDM optical signal.
The invention also provides an optical communications system using a filter of the invention for interleaving or de-interleaving two groups of channels of a WDM
optical signal.
Brief description of the drawings Examples of the invention will now be described in detail with reference to the accompanying drawings, in which;
Figure 1 shows a filter in accordance with the invention;
Figure 2 shows the method of manufacturing the filter of Figure 1;
Figure 3 shows the transmission and reflection response of the filter of Figure 1;
Figure 4 shows the relationship between one peak of the filter response and a standard pass band of a multiplexer filter;
Figure 5 shows the relationship between one trough of the filter response and a standard rejection band of a multiplexer filter;
Figure 6 is used to show the dependency of the filter response on the silicon thickness;
Figure 7 is used to show the dependency of the filter response on the spacer layer thickness;
Figure 8 shows the filter used for de-interleaving and interleaving; and Figure 9 shows one possible use of the filter of the invention to provide add/drop functions.
Detailed description Figure 1 shows a filter 10 in accordance with the invention, comprising three stacked optical cavities 12 of equal thickness. The cavities are spaced by spacers 14. Each cavity provides a Fabry-Perot cavity frequency response, determined by the so-called "Airy" function. By providing a number of cavities, the transmission and reflection transfer function is altered to provide flatter peaks, so that a comb filter response results. Thus, the filter has a comb transmission frequency response and a comb reflection frequency response, with the peaks of the transmission comb response lying between the peaks of the reflection comb response.
As a result, the filter can be used to divide a W DM signal 20 comprising frequency components A1,I~2,~3... ~n into a first group of channels 22 transmitted through the filter, and a second group of channels 24 reflected by the filter.
For example, the first group of channels 22 has the odd channels ~1, h3 etc, and the second group of channels 24 has the even channels ~2, h4 etc.
The input W DM signal is provided at an angle to the filter, so that the reflected channels can be collected.
The thickness of the cavities 12 is selected as a function of the channel spacing, and determines the separation between the peaks of the transmission and reflection comb responses. In particular, the free Spectral Range (FSR) for a single cavity Fabry Perot is given by:
FSR = X02 /(2nd) where ao is the centre wavelength of operation, n is the refractive index of the cavity and d is the physical length of the cavity.
This equation can be used to derive the cavity length for a required FSR, which corresponds to the separation of the peaks in the comb response. For example, for a frequency corresponding to wavelength 1530nm (one standardised frequency is 1530.48nm) and with channel separation 0.8nm (giving an FSR of 1.6nm and corresponding to a 100GHz grid), with a silicon cavity (refractive index 3.63), we obtain a physical cavity length of 202 microns.
In practice, a slightly smaller thickness is required due to the angle of incidence, which may be, for example 3 degrees. Thus, the required cavity thickness for C
band operation using a 100GHz grid with a silicon wafer works out to be approximately 195 microns, particularly between 194.5 and 195.0 microns.
The spacing between cavities is selected to be one quarter wavelength of the wavelength over which the filter is to be used. For use in the C- band (1524nm -1563nm), the cavity separation is selected to be one quarter wavelength of any wavelength in this range, for example based on 1528nm, giving a thickness of 382nm.
For using the filter as a de-interleaving filter, the thickness of each cavity is selected so that the spacing between the comb peaks corresponds to double the channel spacing of the WDM optical communications system.
The frequency response of the filter is a function of the refractive index of the material defining the cavity and the material in the spacing. In particular, the greater the refractive index difference, the fewer cavities are required to obtain the required frequency response.
In the preferred embodiment of the invention, each cavity is defined by a silicon wafer, and adjacent cavities are separated by an air spacing. The use of silicon wafers enables existing techniques to be employed to obtain specific cavity thicknesses with sufficiently accuracy and uniformity.
Figure 2 shows the method of manufacturing the filter. Figure 2A shows a silicon wafer 30 prepared to have the desired thickness, using conventional silicon processing techniques, for example double-sided polishing controlled using optical thickness measurement. This optical thickness measurement preferably comprises infrared thickness sensing.
In Figure 2B, an oxide layer, for example Si02, is deposited over the silicon wafer to the desired spacing and is then patterned to define the regions 32.
These regions have an opening 34 and a support 36 surrounding the opening 34.
The oxide layer is patterned using conventional photolithographic techniques.
The wafer is then divided into sections 40 as shown in Figure 2C, each section comprising one of the regions 32. They are then physically stacked as shown in l0 Figure 2D and diffusion bonded as shown in Figure 2E. This diffusion bonding will compact the oxide layer, and the original deposition thickness of the oxide layer will take into account the change in thickness resulting from the diffusion bonding process.
Figure 3 shows the transmission and reflection response of the filter when used for a 50GHz grid. Of course, this requires a different thickness of silicon cavity.
The filter of the invention can, however, be tuned to any desired comb separation by calculating the thicknesses in the manner set out above.
The upward arrows 50 in Figure 3 indicate the transmitted frequency bands which together define a first comb frequency response, and the downward arrows 52 indicate the reflected frequency bands which together define a second comb frequency response. In this example, the channels have a separation of 0.4nm for the 50GHz grid. The filter rejection is around -23dB, and each comb peak has a reasonably flat attenuation of less than 1 dB over a wavelength band of around 0.3nm.
Figure 4 shows two adjacent peaks in one of the comb responses in a filter designed for the 100GHz grid, namely with a channel separation of 0.8nm, and shows that the comb filter response exceeds an example of a known band pass profile 60 required by multiplexing filters. Figure 5 shows that the rejection -C~-between channels of the comb filter response (around 27dB for a filter designed for 0.8nm channel separation, as shown) also exceeds an example of a known rejection profile 62 required by multiplexing filters.
The thickness of the silicon wafer should be accurately controlled during manufacture. Figure 6 shows the dependency of the filter response on the silicon thickness, and shows two theoretical plots 70,72 for filters with silicon thickness varying by 3nm. In practice, it is sufficient to control the silicon wafer thickness to within around ~l0nm from the desired thickness.
The thickness of the oxide spacer layer should also be accurately controlled during manufacture. Figure 7 shows the dependency of the filter response on the cavity spacing, and shows three theoretical plots 80, 82, 84 for filters with the cavity spacing at the desired dimension, 30nm from the desired dimension and 50nm from the desired dimension. In practice, it is sufficient to control the silicon wafer thickness to within around ~50nm from the desired thickness.
These thickness controls can be achieved using known processing techniques used in the semiconductor device manufacture industry.
Figure 8 shows how the filter of the invention can be used within a demultiplexer and multiplexes function. A filter of the invention 90 acts as a de-interleaves and divides the W DM signal into the odd channels 92 and the even channels 94.
Each of these groups is passed to a conventional de-multiplexes 96, 98, for example arrayed waveguide de-multiplexers. The channels at the input to these de-multiplexers have double the spacing of the W DM channels, so that the required optical performance of the de-multiplexers is lower.
In order to interleave channels, the filter of the invention 100 is again employed, with the direction of signals through the filter reversed. The interleaves 100 combines the channels from the two multiplexers 102,104.
Figure 9 shows the use of the filter of the invention to provide add/drop capability in a node. The input 120 carries the WDM signal, and in this example, add/drop capability is only required for a sub-set of the channels, and this capability is provided for the group of channels 122.
After de-multiplexing at 124, the individual channels from the group 122 are each supplied to an optical switch 126. The switch enables a channel from the demultiplexer 124 to be dropped or else passed forward. Alternatively, the switch 126 can enable an external signal 128 to be added. Each switch comprises a 2x2 optical switch, and enables add or drop functions to be implemented for each channel.
The channels passing forwards (either added or passed through) are each passed through a variable optical attenuator (VOA) 130 and then combined in a multiplexes 132. The VOAs ensure the signals have optical powers appropriate for the system, for example dependent upon the receiver characteristics within the system.
In the example of Figure 9 add/drop capability is provided for one group of channels 122, and the other group of channels 134 is passed through the node and recombined with the output from the multiplexes 132 at a filter of the invention 136 arranged to provide an interleaving function.
The filter of the invention can be used in other optical processing devices, and essentially enables channels to be divided into groups so that channels with greater separation can be further processed.
The specific example of filter of the invention uses a silicon cavity, and this is the preferred material to enable the thickness to be processed accurately.
However, other materials may be used, and this will result in different thickness requirements as a result of the different refractive index. A material may be provided between the cavities instead of the air spacing described above.
Similarly, the filter of the invention can be used in a range of wavelength bands.
In the specific examples above, the filter is used in the C-band. However, technologies are being developed enabling the L-band and S-band to be used in optical communications systems, and the filter of the invention can be used for these wavelength bands. Again, this wilt alter the required layer thicknesses.
Thus, it will be understood that the specific layer thicknesses given are in respect of the one specific example of the use of the filter.
Various other modifications will be apparent to those skilled in the art.
Claims (27)
1. An optical filter comprising a plurality of stacked optical cavities each having substantially the same thickness, the filter having a first frequency response for transmission through the filter comprising a first comb response and a second frequency response for reflection from the filter and comprising a second comb response, the peaks of the first comb response lying between the peaks of the second comb response.
2. An optical filter as claimed in claim 1, wherein each cavity is formed from a silicon wafer.
3. A filter as claimed in claim 1, wherein the thickness of each cavity is selected such that the spacing between peaks of each comb response corresponds to double the channel spacing in a WDM optical communications system in which the filter is to be used.
4. A filter as claimed in claim 1, wherein the spacing between adjacent cavities is equal to one quarter of a wavelength within a band of wavelengths of channels of a WDM optical communications system in which the filter is to be used.
5. A filter as claimed in claim 1, wherein the thickness of each cavity is given by:
d = .lambda.02(2n.FSR) where .lambda.0 2 is a wavelength within a range of wavelengths for filtering by the filter, n is the refractive index of the cavity and FSR is desired spacing between adjacent peaks of the first and second comb responses.
d = .lambda.02(2n.FSR) where .lambda.0 2 is a wavelength within a range of wavelengths for filtering by the filter, n is the refractive index of the cavity and FSR is desired spacing between adjacent peaks of the first and second comb responses.
6. A filter as claimed in claim 1, wherein adjacent cavities are separated by an air spacing.
_13_
_13_
7. A filter as claimed in claim 6, wherein the thickness of the silicon wafer of each cavity is between 180 and 210 microns.
8. A filter as claimed in claim 7, wherein the thickness of the silicon wafer of each cavity is between 194 and 195.5 microns.
9. A filter as claimed in claim 6, wherein a support layer is provided between the cavities for defining the air spacing, the support layer comprising an oxide layer.
10. A filter as claimed in claim 6, wherein the spacing between adjacent cavities is 0.36 - 0.41 microns.
11. A filter as claimed in claim 10, wherein the spacing between adjacent cavities is 0.37 - 0.39 microns.
12. A filter as claimed in claim 1, comprising three cavities.
13. A filter as claimed in claim 1, comprising a de-interleaving filter for de-interleaving a group of evenly spaced channels into first and second groups, wherein the spacing between the peaks of the first and second comb responses is equal to double the channel spacing.
14. A method of manufacturing an optical filter, comprising:
preparing a silicon wafer having a predetermined thickness;
depositing an oxide layer over the silicon wafer;
patterning the oxide layer to define a plurality of pattern portions;
dividing the wafer into sections, each section comprising a pattern portion;
stacking a plurality of the sections to define a multiple cavity filter, the wafer sections each defining a cavity and the pattern portions defining a spacing between adjacent cavities.
preparing a silicon wafer having a predetermined thickness;
depositing an oxide layer over the silicon wafer;
patterning the oxide layer to define a plurality of pattern portions;
dividing the wafer into sections, each section comprising a pattern portion;
stacking a plurality of the sections to define a multiple cavity filter, the wafer sections each defining a cavity and the pattern portions defining a spacing between adjacent cavities.
15. A method as claimed in claim 14, further comprising diffusion bonding the stacked sections.
16. A method as claimed in claim 14, wherein the step of preparing a silicon wafer having a predetermined thickness comprises double sided polishing to the required thickness.
17. A method as claimed in claim 16, wherein the step of preparing a silicon wafer having a predetermined thickness further comprises controlling the double sided polishing in dependence on optical thickness measurement.
18. A method as claimed in claim 17, wherein the optical thickness measurement comprises infrared interferometric thickness sensing.
19. A method as claimed in claim 14, wherein the pattern portions each comprises a central opening around which one or more support portions of predetermined thickness are provided.
20. A method as claimed in claim 14, wherein the silicon wafer is prepared to have a thickness between 180 and 210 microns.
21. A method as claimed in claim 20, wherein the silicon wafer is prepared to have a thickness of between 194 and 195.5 microns.
22. A method as claimed in claim i4, wherein the oxide layer is deposited with a thickness of 0.36 - 0.41 microns.
23. A method as claimed in claim 22, wherein the oxide layer is deposited with a thickness of 0.37 - 0.39 microns.
24. A method as claimed in claim 14, wherein stacking a plurality of the sections comprises stacking three sections.
25. An optical filter comprising a plurality of stacked optical cavities each defined by a silicon wafer and having substantially the same thickness, wherein adjacent cavities are separated by an air spacing, and wherein the filter having a first frequency response for transmission through the filter comprising a first comb response and a second frequency response for reflection from the filter and comprising a second comb response, the peaks of the first comb response lying between the peaks of the second comb response.
26. Use of an optical filter as claimed in claim 1 for interleaving or de-interleaving two groups of channels of a WDM optical signal.
27. An optical communications system comprising a filter as claimed in claim 1 for interleaving or de-interleaving two groups of channels of a WDM optical signal.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US10/011,078 | 2001-12-07 | ||
| US10/011,078 US6710922B2 (en) | 2001-12-07 | 2001-12-07 | Optical filters |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA2413650A1 true CA2413650A1 (en) | 2003-06-07 |
Family
ID=21748786
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA002413650A Abandoned CA2413650A1 (en) | 2001-12-07 | 2002-12-05 | Optical filters |
Country Status (3)
| Country | Link |
|---|---|
| US (1) | US6710922B2 (en) |
| EP (1) | EP1318417A3 (en) |
| CA (1) | CA2413650A1 (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7881805B2 (en) * | 2002-02-04 | 2011-02-01 | Boston Scientific Neuromodulation Corporation | Method for optimizing search for spinal cord stimulation parameter settings |
| KR100891265B1 (en) * | 2002-06-27 | 2009-03-30 | 주식회사 케이티 | Planer lightwave circuit type interleave device for wavelength division and manufacturing method therefor |
| US7092162B1 (en) | 2002-11-02 | 2006-08-15 | Yung-Chieh Hsieh | Bandwidth enhancement device |
| AU2004256148B2 (en) * | 2003-07-14 | 2006-04-06 | Commonwealth Scientific And Industrial Research Organisation | An optical filter, an optical interleaver and associated methods of manufacture |
| AU2003903606A0 (en) * | 2003-07-14 | 2003-07-24 | Commonwealth Scientific And Industrial Research Organisation | An optical filter, an optical interleaver and associated methods of manufacture |
| US20080264553A1 (en) * | 2007-04-27 | 2008-10-30 | Hewlett-Packard Development Company Lp | Embossing |
| CN104459891A (en) * | 2014-12-26 | 2015-03-25 | 昂纳信息技术（深圳）有限公司 | Optical comb filter |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5233464A (en) * | 1991-03-20 | 1993-08-03 | Costich Verne R | Multilayer infrared filter |
| US5212584A (en) * | 1992-04-29 | 1993-05-18 | At&T Bell Laboratories | Tunable etalon filter |
| US6208444B1 (en) * | 1996-10-29 | 2001-03-27 | Chorum Technologies Inc. | Apparatus for wavelength demultiplexing using a multi-cavity etalon |
| CA2199996C (en) * | 1997-03-14 | 2002-08-13 | Cindy Xing Qiu | Methods to fabricate dense wavelength division multiplexers |
| WO1999047956A1 (en) * | 1998-03-19 | 1999-09-23 | Ciena Corporation | Fabry-perot optical filter and method of making the same |
| US6341040B1 (en) * | 1999-06-08 | 2002-01-22 | Jds Uniphase Corporation | Multi-plate comb filter and applications therefor |
| US6356689B1 (en) * | 2000-03-25 | 2002-03-12 | Lucent Technologies, Inc. | Article comprising an optical cavity |
| GB0029224D0 (en) * | 2000-11-30 | 2001-01-17 | Secr Defence | Optical filters |
| TW528891B (en) * | 2000-12-21 | 2003-04-21 | Ind Tech Res Inst | Polarization-independent ultra-narrow bandpass filter |
-
2001
- 2001-12-07 US US10/011,078 patent/US6710922B2/en not_active Expired - Fee Related
-
2002
- 2002-12-04 EP EP02258392A patent/EP1318417A3/en not_active Withdrawn
- 2002-12-05 CA CA002413650A patent/CA2413650A1/en not_active Abandoned
Also Published As
| Publication number | Publication date |
|---|---|
| EP1318417A2 (en) | 2003-06-11 |
| US6710922B2 (en) | 2004-03-23 |
| EP1318417A3 (en) | 2004-05-26 |
| US20030107811A1 (en) | 2003-06-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20020003643A1 (en) | Multi-channel wavelength slicing using an etalon-based interleaver for dense wavelength division multiplexing | |
| US20010031113A1 (en) | Miniature monolithic optical add-drop multiplexer | |
| CN107735707B (en) | Optical device and method for tuning an optical signal | |
| WO1999027673A9 (en) | A programmable add-drop module for use in an optical circuit | |
| EP1033841B1 (en) | Reconfigurable add/drop for optical fiber communication systems | |
| US6496616B2 (en) | Miniature monolithic optical demultiplexer | |
| US20050105902A1 (en) | Optical performance monitor using co-located switchable fiber bragg grating array | |
| US6710922B2 (en) | Optical filters | |
| US6650809B2 (en) | Multiple band optical multiplexer and demultiplexer | |
| EP1030480A2 (en) | Configurable optical circuit | |
| US20020176660A1 (en) | Optical wavelength multiplexer/demultiplexer and use method thereof | |
| US6832018B2 (en) | Dynamically reconfigurable optical add/drop multiplexer | |
| US6904203B2 (en) | Passband flattened demultiplexer employing segmented reflectors and other devices derived therefrom | |
| GB2378333A (en) | Optical wavelength de-interleaver with periodic 'square wave' reflection properties | |
| JP3511010B2 (en) | Optical filter, tunable add-drop continuation module and circuit arrangement for bundled cross-connect function | |
| KR20200034628A (en) | Multiplexer and Demultiplexer For Dense Wavelength Division Multiplex | |
| US20020118915A1 (en) | Optical multiplexer/demultiplexer device | |
| KR100709880B1 (en) | A tuneable optical filter | |
| CN1874195B (en) | Multiplexing / demultiplexing ROADM based on TFF | |
| US6768841B2 (en) | Integrated optical interleaver/de-interleaver | |
| EP1126293A2 (en) | Multiplexing/demultiplexing optical circuit | |
| WO2001086825A2 (en) | Miniature monolithic optical devices for dwdm | |
| CA2281044A1 (en) | Optical multiplexor/demultiplexor | |
| EP1447693A1 (en) | Flexible Passband Filter | |
| US6950571B1 (en) | Optical switchable component |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FZDE | Discontinued |