How to measure thickness of an ultra-thin metal layer? I have a sample of a metal (aluminum/oxide) layer with a gradual thickness ranging from monolayer to 100nm. This layer is deposited on a transparent substrate, like glass. How can I measure this thickness variation across the sample, qualitatively or quantitatively?
Film thickness ranges from 0 to several nms. Film size, i.e. from left to right side of the image below, is about 10mm.
I know that Atomic force microscopy(AFM) and X-ray photoelectron spectroscopy(XPS) are two possible options. But are there any other simpler characterization techniques? Like some optical techniques (light absorption variation, etc)?

 A: if your sample is partially coated and partially bare with the subtract, you use the old interferometry method very easily and precisely. See this reference. Otherwise, you may want to check out Ellipsometry and Profilometry for precise measurement. 
A: I faced exactly this problem in my PhD, though it was silver on glass not aluminium. All the really precise methods are time consuming, so I ended up using them to make a calibration curve for the optical absorption. Then I measured the thickness using a standard lab spectrometer. This wasn't stunningly accurate, but allowed me to process the dozen or so samples per day that I was using in a reasonable time. I was working with thicknesses from 10 to 50nm - if you go a lot thicker the light transmission becomes quite low.
The lateral resolution isn't great, so you may not find this useful if your samples change rapidly with distance along the sample.
You need a calibration curve because, at least for silver, the film microstructure changes with thickness. Having said that, the calibration curve was very close to the expected exponential dependence.
A: I have a project in which thin metal films up to 5000 Angstroms thickness are sputtered onto various substrates, and the tool that is regularly used to determine thickness is just a mechanical contact or stylus profilometer. That's the preferred method because it's fast and fairly accurate. We also have an optical profilometer which uses optical interference fringes to determine film thickness. In principle it's superior because it offers much higher resolution and sensitivity. In practice, it's much less convenient to use compared to a stylus profilometer, so we rarely use it.
