MIDAS98 Instructions


Main menu The program starts with a small introductory page which identifies which version you have and what it can do. Click on the banner or press any key to proceed to the main screen.

The MIDAS98 main form is presented and should look as shown opposite (if it doesn't look the same, the MIDAS program may be having confilcts with your monitor settings - this can cause some problems later further into the program).

The form has a header identifying it as MIDAS98w - presently, only the beta version is available.

Important points There are some important features of the main form which should be noted:

Of the two buttons on the right of the form, Run and Exit, only Exit is initially active since the program can not perform any processing until a file and actions are selected. Similarly, the menu options under File and Help are the only ones active initially.

Opening a file

File selector The first item that the program needs to operate is an AFM file which it will manipulate. The file formats presently supported by the MIDAS program can be seen by pulling down the list under the "File" menu. At present, only Digital Instrument's Nanoscope version 2 and 4.x are supported. The program will also not work for "Nano 4" files which are created by reading older versions.

As prompted, the user should "Open the AFM file to be used". This can be done in one of three ways - first, the filename can be directly input into the Processed File text box. Unless the user knows the path and name of the file exactly, this method is not recommended. It is far more convenient to use the File system which can be accessed by either pressing the active "File" button next to the text box or by selecting the "Open File" option under the "File" menu. Either of these brings up the form shown opposite, from which the user can select the file to be used.

File found Once the AFM file is selected, the MIDAS program will attempt to identify which format the file is and also determine if the file contains any "height" data. If the file cannot be identified, the user is informed by a beep and an "Unknown" message in the Manufacturer text box. If successfull, the file details will be shown in the text boxes in the second row. A form will show up giving the header from that file so that the user can check the details it might contain (such as comments, dates, etc..). Another form will appear which contains the device independant bitmap of the height image along with the height scale.

Process to be done

Here the user must decide what is to be done to the input file:


Deconvolution is the process whereby the AFM data is manipulated to account for the geometries of the tip or substrate, thereby respectively improving the quality of the images or giving an in situ means of inspecting the tip.


Convolution is used to form an image which is similar to the one obtainable by the AFM. The program needs that the input file contain an ideal topography of the sample which is to be convoluted with one of the ideal tip geometries given below.

Examples of these ideal samples are the DELTA, SIMGRID and 2SPHERE files included in the Digital version 2&3 files.


Here the user must decide which tip/sample/substrate they wish to use:


The sphere is both an ideal tip geometry and a substrate. It can therefore be used in deconvolution or convolution.

Unless they are terribly distorted, most AFM tips can be assumed to have a spherical shape at the apex. One can therefore represent the tip as an ideal sphere when using deconvolution to reduce the effect of the tip on the images obtained. The size of the tip/sphere can sometimes be obtained directly from the AFM images. If the object which was scanned has a spherical geometry of diameter "d" (which can be obtained from the height as measured by the AFM), the radius of the tip "R" can be calculated from the measured width "w" as w=sqrt(2Rd).

Otherwise, the radius of the tip can be estimated from electron microscopy. Typical results are around 25 to 30 nm for regular tips, 10 to 15 for special tips.

Using a spherical tip to deconvolute an image can vastly improve the quality of the images by reducing the distorted width. Use caution when using this shape since it will only have noticable effect on images up to around 50 nm in height.

As a substrate, the sphere can be used to obtain the tip geometry from images of samples such as polystyrene or colloidal gold spheres. See Li and Lindsay (Rev. Sci. Instrum. 62 (11), 1991, pgs 2630-2633) and Xu and Arnsdorf (J. Microscopy 173, Pt 3 1994 pgs 199-210) for notes on the AFM of polystyrene and gold spheres, respectively. The size of the sample will influence the amount of tip which is obtained.


Oblong structures are for deconvoluting AFM data to obtain improved images.

The program asks for the diameter of the sphere (which represents the ends of the oblong structure) and for the aspect ratio. The oblong structure takes the sphere and cuts it in half, then stretching these semispheres out to give the desired aspect ratio. Thus, an aspect ratio of 1:1 will give a true sphere.

It has been my experience that the most commonly distorted tip which is still somewhat usable has the oblong shape.


Holes represent circular depressions in a substrate which can be used to deconvolute images to obtain the tip geometry.

The program asks for the spacing (or pitch) of the holes and the diameter. The program does not require the depth of the holes.

The algorithm was created to work on the circular arrays designed to check the x-y calibration. These arrays were originally distributed by Digital Instruments but are no longer available. Papers suggesting the use of nuclear tracks in polycarbonate filters (J. Tentschert et al., click here to download their really BIG postscript preprint) or special polymer blends (T.O. Glasbey et al., Surface Science 318 (1994) L1219-L1224) might be useful as substitutes for the arrays.

The circular depressions have been shown to be of limited use and are chiefly good for conical tips only.


The conical tip can represent both an ideal tip and a substrate, and can therefore be used in deconvolution or convolution.

The program asks only for the aspect ratio of the tip. An aspect ratio of 2.0 (2:1) will give a tip angle of 90 degrees.

No substrates using true conical tips is available (high aspect material grown on a surface has a rounded tip and might best be represented by a sphere).

Deconvoluting to obtain better images is of little or no use.


The pyramidal tip can represent an ideal tip and substrate, and can therefore be used in deconvolution and convolution.

The program ask only for the aspect ratio of the tip. An aspect ratio of 2.0 (2:1) will give a pyramid with opposite faces at 90 degrees.

The selection was offered to answer the question of whether AFM tips can be used to examine AFM tips. Broken tips (intentional or not) on a surface can be imaged as shown here and deconvoluted to give a limited view of the imaging tip.

Deconvoluting to obtain better images is of little or no use.


Grates represent square depressions in a substrate which can be used for deconvoluting images to obtain the tip geometry.

The program asks for the spacing (or pitch) of the holes and their diameter.

The algorithm is similar to that for Holes. It has been applied to the grids used in transmission electron microscopy, but these diameters are tens of microns and so are poor for seeing the two micron tips of the AFM. Other grids are available (possibly) from VLSI Standards Inc. (contact Len Anderson 800-228-8574 ?).

Round Post

Round Posts are cylindrical projections from a substrate which can be used for deconvoluting images to obtain tip geometries. In theory, they might also represent very blunt tips.

The program asks for the diameter of the post: The height is not needed.

Images of these arrays are given in the literature, but no sources have as yet been found. These would be easier to use than Square Posts since there is no orientational dependance (the round posts "look" the same no matter how they are positioned relative to the tip).

Square Post

Square Posts are rectangular projections from a substrate which can be used for deconvoluting images to obtain the tip geometries.

The program asks for the length (or width) of the post, with the assumption that the post is square in cross section (length=width). The height of the post is not needed. It is also assumed that the post was scanned such that it appears as a square and not a triangle.

These rectangular posts are available from Topometrix (I think) for use as x-y calibration. The size of these is roughly four (4) microns which may be too large for good resolution of the tip apex.

Previously found tip

The tip geometry can then be used to improve other images made using the same tip.

This feature is not available on MIDAS since it becomes a serious ethical question when one uses a highly distorted tip to further process images. The question of how much one can scale these to other sizes and still be representative of the true geometry is also yet to be debated.


At this point the program begins crunching numbers. The user can get some idea as to how long the program will take by looking at the progress bar in the prompt's text box.

To end the program prematurely, press the Exit button to abort the run and MIDAS program, or press the Ctrl-Alt-Del buttons simultaneously to get into the Windows Manager and abort the program from there.

Save form

Save File

The program has finished and will ask if you want to save the file (as shown) and do another calculation.

To Main Page Hand


Last revised on 03-06-1999 by Peter Markiewicz