Integrated Optics - Science topic

I've never worked with Si-SiO2 wafers, but when I've had to desiccate objects, warming in a hard vacuum (10^-4mbar and better) was the canonical method to 'bake out' the monolayer of water that is (unavoidably) present on metals that have seen ambient conditions.

It's worked for me more effectively than using short-wavelength UV to split the water.

These modulators do not work by the ferroelectric effect. They work by the electrooptic effect. The ferroelectric polarization messes up the electrooptic effect. Opposite domains result in opposite phase retardation effectively cancelling each other. In the regime of low field, the hysteresis of the polarization makes the electrooptic response nonlinear. Working with fields well above the coercive force ensures the domains are aligned so that the electrooptic effect is linear and predictable.

The crystals also have a piezoelectric response which also messes up the electrooptic effect. On acoustic time scales the physical movement of the crystal cancels some of the field reducing the electrooptic response. This makes the electrooptic response frequency dependent.

The point is you never get just the behavior you want. Nonlinear crystals have every nonlinear behavior. They have nonisotropic coefficients of thermal conductivity, complicated stress tensors, etc. all for the same reason: the crystal symmetry. If you want to exploit one effect, you usually have to work around the others. Thus, in this case, you want an electrooptic modulator, but you wind up having to think about ferroelectrics, piezoelectrics, walk-off from the nonisotropic dielectric tensor, etc., etc.

ASAP Software is the most recommended and quite powerful software for Optical interconnect.

http://www.breault.com/software/about-asap

Thanks for this helpful link!

up to my knowledge grephene should be combined to other materials to harvest photons and why not conduct them in integrated optics

When the nonlinear response is dominated by the Kerr effect (nonlinear refractive index change) the asymmetry is easy to understand in the limit of thin media. Essentially the Kerr effect induces a "nonlinear lens" in the material which changes the subsequent propagation of the beam. The example you show in your question would correspond to the case of a negative Kerr effect (n2<0). When the sample is placed before the focal plane of the physical lens, the combined effect of the lens originally used to focus the beam and the negative focal length "lens" induced in the sample is to translate the beam's focal plane further along the z-axis. Thus by the time the beam reaches the aperture it has not diverged as much and the transmission through the aperture is higher. Conversely when the sample is placed behind the focal plane of the physical lens, the induced negative lens leads to an increased divergence and consequently a lower transmission at the aperture. It is not just the phase shift, but difference in curvature of the wavefronts on either side of the focus that gives rise tot he assymmetry. You can find a more detailed explanation in the book "Fundamentals of Nonlinear Optics" by Powers and Haus, chapter 8.

Basically the optics is roughly the same as in the thermal lens case. The main difference is that for an electronic nonlinearity in the absence of absorption, the induced lens is an "instantaneous" response to the transverse profile of the incident beam, whereas in the thermal case it is the radial variation of the adsorbed energy convoluted with thermal diffusion that produced the effective lens induced by the incident beam. Of course the detailed shape of the Z-scan curve will be different, due to the difference in the transverse variation of the induced refractive index change in the two cases. In the "instantaneous" electronic case, this variation will follow the radial variation of the incident beam, in the thermal case the balance between the power deposited by the incident beam and diffusion of the heat within the sample will determine the spatial profile of the induced refractive index change.

I really appreciate your help with my project, Prof.Zbigniew Motyka and Prof.Maria Chiara Ubaldi.

fortunately i found the answer.

I am attached the answer below.

Yes, I need these kind suppliers a lot , thank with my great regards you dear colleague Alex Risos .

Fresnel coefficients indeed do not depend on layer thickness, but the total transmission is. Beyond extinction, you should also take into account frustration of evanescent waves (depending on Angle of incidence and ratio of refractive indexes). Summing dipoles is always right, but if the layer is very thin, you should take into account the dipoles also on its "other side". So microscopic picture gets rather involved there. 

Number one is the etch-mask you use.  If you go to high-res Deep-UV stepper, you'll typically get the lowest sidewall roughness.  Some tricks can be played with Contact (otherwise the highest-roughness lithography) eg. small address unit, photoresist reflowing.  

Additionally, the transfer of this mask into the InP plays a large role.  Here users found that the (rather complicated) process of Photoresist:Chrome:SiO2:InP ridge yielded the best loss, but then all of these etches need to be developed (Chrome in particular can be tricky).  The InP etch itself can have some smoothing capabilities, eg. addition of N2 or H2 as a process gas, with some tradeoff with verticality.

There are, as you've probably found, many papers on InP ridge waveguide sidewall roughness.  For example:

J. W. Bae, W. Zhao, J. H. Jang, I. Adesida, A. Lepore, M. Kwakernaak, and J. H. Abeles, “Characterization of sidewall roughness of InP/InGaAsP etched using inductively coupled plasma for low loss optical waveguide applications,” Journal of Vacuum Science & Technology B, vol. 21, no. 6, pp. 2888–2891, Nov. 2003.

C. Carlstrom, E. Van Der Drift, R. Van Der Heijden, R. Notzel, R. van Veldhoven, F. Karouta, H. Salemink, and A. Talneau, “Cl2/O2- and Cl2/N2-based inductively coupled plasma etching of photonic crystals in InP: sidewall passivation,” Indium Phosphide and Related Materials, 2005. International Conference on, pp. 315–318, 2005.

Note the inspection methods: SEM sometimes looks "smooth", only because the SEM's resolution (or charging) isn't good enough!  AFM is a great way to get RMS roughness, as long as you don't have an undercut (check your SEMs for this). AFM would be particularly applicable for checking the line-edge roughness of your masks (PR, SiO2 etc.)

One of the biggest differences between semiconductor waveguides  and dielectric waveguides is that the refractive index is typically much larger. This means that the scattering will be much larger for the same size defect or sidewall roughness.  This is significant in the design and processing. If you really want to make very good waveguides you spend a lot of time on sidewall and surface roughness. 

Other items less connected to your question but can be considered.

Semiconductors will also have loss dependent on their electrical conductivity.  As recent work with silicon modulators have show, if you can move the charge around this can be useful.

Semiconductors can also have higher nonlinearities which can be important when intensities become large. Raman and Brillouin scattering can be larger per unit length.

FInally direct semiconductors can have very high optical gains. 

If you have problems with COMSOL, start using RSOFT CAD software, way easier than COMSOL and better.

You may want to look in this recent publication in Optics Express and references therein:

Anisotropic model for the fabrication of annealed and reverse proton exchanged waveguides in congruent lithium niobate

[http://dx.doi.org/10.1364/OE.23.001748]

The bending loss strongly depends on the bending radius and the waveguide size. For a conventional silicon strip waveguide with a core size of 500 nm x 220 nm surrounded by glass and the bending radius of 1 micron, the bending loss is ~2 dB per bend at telecommunication wavelength (1550 nm). But the loss drops rapidly as you increase the bending radius. 

The insertion loss for this kind of waveguides is ~3 dB/cm. Another important figure of merit for dense photonic integration is cross talk. Light confinement is also important for nonlinear optics and quantum optics applications. 

It seems you could use a bundle of 100 micron plastic fibers  instead LPG.

Thanks

Suppose you have a solution for propagation constants of your particular waveguide and those for the uniformly filled waveguied. Then you can change (in several steps) the parameters of your waveguide filling toward parameters of  the uniformly filled waveguide for which modes are clearly defined to see the transformation of your dispersion curves into those of the uniformly filled waveguide.

i.e. if we have a waveguide filled with a layered dielectric we can gradually change the permittivities of layers toward the homogeneous filling controlling the transformation of the dispersion curves. 

Yes, but there is a coupling between the CW and CCW transmitted waves which causes the gyro to be locked at low rotation rates.

This makes change in frequency to be zero..

For analytical details u can refer the link provided..

Hello,

have no experience with waveguides. Formally, if you reach pump intensity and phase matching is OK, gain values should be similar to one obtained in the free-space setup. Should be no difference is there waveguide or not. Except damage threshold. I expect that waveguide is kind of defect or inclusion what may drastically decrease material damage threshold. Sometimes waveguides are much longer as bulk crystals with expectation that higher gain can be obtained in longer material with lower intensity, but often Raman starts and four-wave-mixing flows energy to frequency sidebands. As result pump intensity is clamped and (or) energy of the signal do not increase. And I believe you will not get the same... but almost.

For the multi-cascaded OPA you can do what you want. Formally there is no limit how much you can amplify. For example with gain 10^10 you can get ~1.6nJ from a single photon. nJ is huge energy but energy and time profile contrast will be extremely bad because of amplified "quantum noise" and Poissonian statistics of single photon seed source. Therefore main questions are how much your seed is larger compared to quantum noise (in the time frame of pump pulse and spatial and frequency bandwidth of OPA) and how you will perform cleaning of the resulting signal. For the best result you need to filter wavelength (frequency) to time limited by the pulse duration and do spatial filtering down to single mode.

Therefore in huge OPA systems with multi-mili-Joule output main question is where to get seed signal strong enough to be 6-8 orders stronger compared to quantum noise level. But not how to get proper gain. In the article you send noise is 20% what is really bad for most practical applications. In the article authors claim that contrast can be easily increased by use of "color" filter. However it is not demonstrated that "easily".

Eevn though in theory the RFOG should be capable of reaching the same sensitivity level as IFOG with much less optical fiber, the reality is still far away from that. I think Honeywell is trying to develop their first commercial RFOG though.

Slits  filter-out the transverse modes, therefore information that is stored spatially will be destroyed.  Nevertheless, the slit will not filter longitudinal modes, therefore temporal modulation will not be destroyed. Therefore my answer is: it  depends how your information is stored :) In your case, as I understand the setup, your modulation is temporal therefore the answer yes is correct.

Specialty tools are important for reproducible and yield. However, a diamond scribe can work well for many materials. GaAs is relatively easy to cleave. Silicon is tougher. Sapphire c-plane is hard, but you can sometimes get nice facets on a-plane. It depends on the material I usually don't rub the diamon scribe along the wafer back and forth. Instead I put down a chem wipe, and wrap a microscope slide with another chem wipe. I align the edge of the chip  and on the back surface and put a small scratch using the diamond scribe. Then I flip the chip over, and place it on either the edge of the microscope slide, or on several folded layers of chem-wipes, then using the rounded ball on the back of the diamond scribe apply a rolling pressure. Placing the edge of the chip on the folded paper seems to help. I have the mark suspended between folded paper and the chem-wipe, There are other tricks for laser diodes people used to bond the chip to a foil with wax , thin the wafer by polishing,  then under a microscope then make a series of tick marks. Then wrapping the foild around a cylinder like a pencil, the laser facets could be cleaved. A dicing saw was used to cut the other dimension. and have more non-reflective edges.

Cleaving optical fibers can also done with a scribe, but I recommend using a tool if possible. 

In all the above cases, if done carefully it is possible to get a very nice cleaved surface. Yield though can be a problem..

I have't really tried to quantify the surface roughness by AFM or other technique, but mirror like surfaces are a reasonable expectation if you have a crystal plane to work with, or in the case of glass fibers, you can apply tension and let the crack propagate.

What are the additional parameters

Closed-form solutions are possible in the domain of simplifying the situation, such as the assumption of an undepleted pump. In fact, the coupled equation framework assumes that the meaning of 'modes' make senses. However, if the physical situation is quite complicated where as the modes couldn't be clearly defined, the coupled mode theory would yield increasing error. Working in the Maxwell's equation framework allows handling of broader situations. However, again the mathematical construct is perhaps beyond any human's capability in the past history. Perhaps, you can try solve it?

As an example of compact hologram printer, see

Hi - I guess there is not much choice of materials when it comes to optical fibres, because they are generally silica or polymer. The former is not bad, because, as a glass, the Raman it gives is relatively broadband, so some people use background subtraction successfully. I have never tried a polymer fibre, but I would expect it to be a lot worse than silica. More interesting would be to use an air-core photonic crystal fibre, where the light is guided in air. You could in principle even turn off Raman scattering completely by filling the fibre with a monatomic gas. Take a look at this paper for an example application: ACS Nano, 2011, 5 (5), pp 3823–3830 DOI: 10.1021/nn200157z

best regards, Daniel