Yes, depending on the film, it is possible to measure by AFM wi no problems! First of all the film has bo be less than 5 um, which is the limit of the z AFM scanner movement. Imagine you want to measure a gold thin film deposited on a silicon substrate. You fist clean you silicon substrate, make an ink mark using an OHP pen, crossing the sample. You deposit the gold film. Then you use isopropyl alcohol and a swab to remove the ink mark kibndly (and the film together). In this way you create a very nice step (believe me!) and you can measure by AFM, doing the applicable corrections.

Once I tried to measure thickness of my thin films by AFM. Those films were prepared by e-beam technique on glass substrate. I put on a mask on the portion of the glass and remove it after deposition. I tried to measure the height of the step that was created by the mask. The result was not very good, because the surface of the glass was not smooth and has ups and down itself.

In your case, if the films are thick, the AFM tip may break with this method. It is better to use optical measurement technique. Also if you have precise scale you can measure sample's weight before and after coating and measure the weight of the film. Then by knowing the density you can calculate the thickness.

Better to use optical or laser profilometer.....

Yes, depending on the film, it is possible to measure by AFM wi no problems! First of all the film has bo be less than 5 um, which is the limit of the z AFM scanner movement. Imagine you want to measure a gold thin film deposited on a silicon substrate. You fist clean you silicon substrate, make an ink mark using an OHP pen, crossing the sample. You deposit the gold film. Then you use isopropyl alcohol and a swab to remove the ink mark kibndly (and the film together). In this way you create a very nice step (believe me!) and you can measure by AFM, doing the applicable corrections.

It depends on the thickness. For a very thin film, you can use the mask method described by Sara Sharifpur above.

For a thicker film, a mechanical or optical profilometer is much better.

Make a vertical section of the glass and mount it vertically and take the measurement. Be care full with the tip. I am not sure of how far this may work.

In alternatives point of view, SEM is much simpler and easier than AFM. The only limitation is that the film should be conductive enough. If your film is metallic/semi-metallic, most of the time this idea will work. But for polymer coated thin films, the chances are relatively too low due to conductivity issues.

I am afraid that width of the removed material deposited on the line done with the marker will be much larger than available size of the microscope scan . My feeling is that an AFM profile is more reliable when it covers the scratch an unscratched area on both sides. Thus, another possibility is to make a narrow scratch (20-30 microns wide) on the film surface. You have to do it gently to remove only the film but you cannot scratch the substrate surface (should be smooth and flat in the section line). Then you can determine the film thickness from the AFM profile. If adhesion is too strong and you cannot remove the deposited layer (e. g. Co, Mo deposited on glass) use an Au or Cu seed layer (buffer). Then you will do a scratch easily, but another measurement of the buffer thickness has to be done This is an disadvantage of this approach (you need to subtract a total thickness of the bilayer from the buffer thickness). We have been applying successfully this method for a long time and it works well.

YES it is possible By taking a lateral image "AFM" of the sample

Dear Mr. Wawro

Why do you say '"an AFM profile is more reliable when it covers the scratch an unscratched area on both sides." ?

To Sara Sharifpur: It depends on a microscope type and setting parameters both on-line and off-line, of course . If the recorded image is automatically flattened then one needs to proceed again such image to measure a scratch depth. Such procedure may disturb a real value to be measured. Recording of unscratched areas on both sides of the removed material deforms a recorded image to a lesser extent. rgds, A.

@ Andrzej Wawro: Thank you for your answer.

How about if I measure the height and roughness before and after film deposition, Can I extract the film thickness from the difference in the height from the two measurement? Instead of creating steps in the film.

Make a scratch in the middle of the film . Then from Z- height  scan you may be able to measure the thickness of the film.  

Good luck!

https://www.researchgate.net/post/Would_anyone_have_a_clue_as_to_the_origin_of_unexpected_spots_in_the_real_time_2D_XRR_signal_from_thin_film_samples_on_substrates_Yoneda_Spots

https://www.researchgate.net/post/What_is_the_best_way_to_interpret_REAL_TIME_3D_X_Y_o-2TH_o_ph_ch_reciprocal_space_relative_intensity_data_from_HRXRD_rocking_curve_analyses?_tpcectx=profile_questions

https://www.flickr.com/photos/85210325@N04/19855148646/

https://www.flickr.com/photos/85210325@N04/19886365571/in/photostream/

https://www.youtube.com/watch?v=wkusJqpLPMo

https://www.youtube.com/watch?v=wTFOXdFcR_4

https://www.youtube.com/my_videos_annotate?v=ks4jvCSrPxw&video_referrer=watch

Related answered question on ResearchGate by following https://www.researchgate.net/post/Could_you_anyone_to_explain_me_the_AFM_technique_to_measure_the_thickness_coating

In addition, a feed on ResearchGate about measuring and characterizing layer thickness and roughness in accordance with ISO standards:

https://www.researchgate.net/post/which_is_the_best_method_to_study_the_thickness_and_roughness_of_my_FET_device_layers

@Akinwuni Amusan Sir, It is not possible. After deposition, a new reference will be set. You can proceed according to the technique suggested by Fernanda S Teixeira ma'am. 

If your film thickness is more than 5um, try using a linear profiler. It is like a lower resolution AFM that scans in a line. Scratch/etch off your material at certain areas, and use the linear profiler to check the thickness with reference to the substrate.

@Manish Gupta Thanks for the information