The reflection of microwaves by glass is explained very simply by the difference in dielectric constant, and its square root, the refractive index, between air and glass.

Reflection by a layer of glass is described by the same maths as optical reflection from thin films or a Fabry–Pérot interferometer (see wikipedia).

The impedance is the ratio of the electric to the magnetic field. The change in dielectric constant means that the electric field is different in the glass to in the air, so the ratios are different, and glass has a different impedance to air, causing a reflection. Alternatively, a step change in the parallel electric field is not permitted so some signal has to be reflected so that the total electric field parallel to the boundary is the same on both sides, at the same time as power is conserved. These descriptions and any other are all equivalent and are results of the requirements of Maxwell's equations.

Malcolm White Thank you

I think that Malcolm and Aparna both are right but not completely. The truth is that the microwave signal will be reflected and refracted due to the difference in the dielectric constant. Usually, glass has some finite thickness and there will be multiple reflections from the first and second surfaces of the glass. Some energy will go back and force due to reflections from these borders on multiple occasions. It is easy to calculate all the resulting waves with consideration of all air and glass parameters. Such analysis is often made for building big windows that have a special glass covered with layers.

Sam Belkin

optical reflection from thin films or a Fabry–Pérot interferometer involves the kind of multiple reflection that you describe. Refraction is also included in all these calculations and descriptions

https://en.wikipedia.org/wiki/Fabry%E2%80%93P%C3%A9rot_interferometer#/media/File:Fabry_Perot_Diagram1.svg

Muhammad Panachikkool only asked about reflection, so my answer was about reflection, not about refraction. Refraction is part of the same requirement to match the fields on opposite sides of a boundary between two materials, although it can be described in other ways.

When I'm doing some experiments with a microwave synthesizer, there will be a spark due to the reflection of the microwave above 750 W power. I suspect this is due to the reflection of the beaker, not from the samples in an aqueous solution because there is also a spark for water (without a sample).

It depends on the cross-section of the beam you are using. If you have 750W in a beam 1 cm diameter the power density is about 7MW/m2 and the electric field will be close to 50 kV/m (sqrt(7000000x377)). 377 ohms is the impedance of free space. This electric field is still a bit low for breakdown in air which is about 2000 kV/m.

The curved beaker may be reflecting some of the microwaves to a focus, which can increase the power density, but the diameter of the focus is (very) unlikely to be smaller than half a wavelength. The electric fields of the forward and reflected waves will add (in places) resulting in higher fields.

Sharp edges and dust may result in sparks for lower average fields. The local electric fields around sharp edges and points are much higher than the average field.

There will almost certainly be a bigger reflection from an air-glass-water boundary than from air-glass-air.

Malcolm White, I understand your point. But in nature, there is no artificial division of science into different parts and disciplines. Reflection is an energy exchange process and it always shows some level of refraction. It is the result of the interaction between the EM field and the change in media properties. Our analyses show us how the resulting energy exchange divides the total energy into different components and allow the calculation of these levels.

What microwave frequency/wavelength are you using?

Describe the design of the experiment.

There is one point not very adressed in this discussion yet; Normal glass is easy we get reflection and transmission depeending on the angle of incidence and polarisation of the incident wave.Hoever these days we often have glass with IR reflective coating which is basically a very thing optically transparent film and it does have often a certain resistivity perunit quare.This very thin resistive layer (but with Rsoften in the order of 10 Ohm) can lead to serious shielding in the microwave range (cell phone signals) and there are tehcniques how to make it still work as IR reflector but having passbands for cell phone bands