A number of characteristics make chromium a popular
contrast medium for high-quality mask fabrication:
- excellent adherence to glass and silica substrates;
- reflectivity stability in air;
- good abrasion resistance;
- excellent contrast.
After coating a new "blank" chromium mask with resist, and after exposure and development, the etching of the chromium film through the "windows" opened in the resist layer can be made by a wet or plasma process. In these notes, only wet etching is discussed.
Two main parameters must be kept under control :
- the uniformity of the etching speed on the whole patterned surface;
- "undercut" control, that is the amount of line broadening due to the action of the etching both below the resist layer, in horizontal direction.
The undercut control is closely related to the chromium film thickness and the etching speed. After a complete process characterization, an accurate control of both concentration, its temperature and etching time usually leads to well reproducible results.
Several chromium wet etching "recipes" are available in the literature :
a - 1 volume of glycerol + 1 volume of HCl;
b - 2 volumes of iron chloride 2.25-2.75 M + 1 volume of HCl, at 80 °C;
c - 9 volumes of saturated solution of cerium sulphate + 1 volume of nitric or hydrochloric acid;
d - 1 volume of a solution of X g of NaOH in 2X ml of water + 3 volumes of a solution of X g of K3 Fe(CN)6 - potassium hexacianoferrate - - in 3X ml of water (Kodak recipe for masks, no depassivation needed);
e - 164.5 g of cerium and ammonium sulphate, Ce(SO4)2 2(NH)4SO4 2H2O, + 43 ml of concentrated perchloridric acid (HClO4) + water till 1 liter (Kodak recipe for masks);
f - 454 g of AlCl3 6H2O + 135 g ZnCl2 + 30 ml H3PO4 + 400 ml H2O.
However, when possible, the use of a commercial etching solution with proprietary composition is recommended.
Among the a to f solutions proposed above, the "a" HCl process is described here. This has been chosen for its readily availability, low cost and acid formulation (this will not affect most positive resists, which can be etched only in basic solutions). The acid, taken from a commercial 37% solution, must be diluted first, and mixed with a moderating fluid. Dilution and mixing slow down the etching process, and make it less sensitive to local surface conditions. HCl can be diluted in deionized water up to a concentration not higher then 10 - 20%. A commercial diluted solution can be used as well. The moderating agent is glycerol, which does not take part in the etching reaction. It must be accurately mixed 1 : 1 with the diluted acid solution.
Note that soon after preparing the etching bath, this can develop heat and rise its temperature. In order to achieve reproducible result, the operator should wait until room temperature is reached. In any case, the bath temperature should be measured (a glass termometer or other inert sensor should be used, for avoiding bath contamination).
A problem which arises when using HCl is its inability to remove the thin native oxide layer which always covers the chromium film. The oxide makes up a "passivation" layer that prevents the etching reaction from starting. In order to remove such layer, an electrical current must flow for a short time between the chromium film and the etching solution. This can be easily accomplished by placing in the solution a sheet or mask (at a uniform distance from the mask) made of a metal more electropositive than chromium, and into electrical contact with it. This makes up a battery that generates a small current capable of removing the oxide layer, through the reducing action of the hydrogen atoms which are created near the chromium "electrode". Aluminium or zinc are well suited to this purpose.
An aluminium wire or foil wrapped around a teflon or plastic sheet can be used as electrode. The electrical contact should be made outside the solution, in a mask corner where the resist has been previously removed (a q-tip wet with acetone can be used).
The contact must last for a few seconds (5") and then opened. At this point the mask remains in the solution, while the aluminium electrode can be taken out of the bath. In the whole operation, maximum attention must be paid to avoid the formation of scratches on the mask, by improper handling of the mask or other electrode.
After removal of the aluminium electrode, the mask must be kept immersed for the whole etching time. During this time the mask should be occasionally shaked or hit to make sure that small bubbles that may appear on its surfaces get detached. Otherwise they could locally affect the etching speed.
A typical 100 nm chromium film is fully etched in 60 seconds. A longer time may originate a visible undercut. This can be evaluated by inspecting the mask under a microscope, before stripping the resist film: the edges of the resist mask are surrounded by a thin dark line, which is the chromium edge.
With a good approximation, a chromium undercut proceeds with the same speed of film etching. This means that a 100 nm thick film is associated to 100 nm undercut on both sides of the etched feature. Hence, a window with a width of 1 mm on the mask becomes a 1.2 mm wide etched feature on the chromium film. If this cannot be tolerated, the effect of the undercut can be taken into account when designing the mask.