HF is clearly capable to etch SiO2. What I'm not sure about: whether HF etches Si as well. But it may be worth some research into this.

Dear Sedao,

I think it all falls back to the question what do consider as "damage-less" and what materials you are using (film material, thickness, silicon).

We studied the removal (ablation) of 75-90 nm thick SiNx layers from doped and textured silicon (PV material) by 30-fs Ti:Sa laser line processing. It appeared to be possible to remove the layer with minor damage of the silicon emitter unterneath, see

J. Bonse, G. Mann, J. Krüger, M. Marcinkowski, M. Eberstein,

"Femtosecond laser-induced removal of silicon nitride layers from doped and textured silicon wafers used in photovoltaics",

Thin Solid Films 542 (2013) 420–425

Article Femtosecond laser-induced removal of silicon nitride layers ...

But I would not call that damage-less. I also remember a discussion in the literature raising the point you mentioned. But for very short pulses multi-photon absorption occuring at the surface of the dielectric cover layer may help to reduce the effect of absorption in the silicon. On the other hand, if you have a textured silicon wafer underneath, near-field enhacement at the sharp maxima of the topography may be less beneficial.

Hope it helps.

Best regards,

Jörn

Dear Sedao,

The anti reflective coating using SIO2 has a thickness of lamda/4 where lambda is the wavelength of the incident radiation. The wavelength of light in the material = lambda/sq root of epsilonr , Assuming lambda= 550 nm, sq root of epsilonr= 2, then the thickness will be in approximately = 550/8= 70 nm. Then you mean the thickness is from 80:100 nm.

The second comment is you said that you plan to use IR femto second laser for ablation. in the literature , 532nm laser is used for ablation of anti reflective coating of silicon, see the attached link. This is understandable as this wavelength will path to silicon where the laser beam must be absorbed to locally melt the silicon at the surface and gets the anti reflective coating film separated from the silicon surface after rapid solidification. So the wavelength must be also as small as possible to restrict its absorption at the surface area of the silicon. Accordingly IR laser will be less suitable than the given wavelength above.

HF dos not etch silicon, it etches sio2 and some other metals..

https://www.spiedigitallibrary.org/proceedings/Download?fullDOI=10.1117/12.2269681&SSO=1

Best wishes

Hi Sedao,

there will always be some damage, but depending on your sample and process sequence, the impact of the laser process on the performance can be low and even comparable to mask and etch techniques. From our experience, an NIR pulse is not a good choice if deep damage needs to be avoided, but the shortest we tried was 180fs. As Mr Zekry already pointed out, the absorption length is too high. We use 355nm for ablation on the emitter side of the cell to keep the damage shallow. Because the SiO2 layer is still transparent (as long as nonlinear absorption can be neglected), silicon ablation will take place at the interface and I would expect a thin layer of silicon to become amorphized with detremental impact on sheet resistance (see Fig. 6 in Article Reduction of picosecond laser ablation threshold and damage ...

Kind regards, Fabian

Thanks Fabian for sharing your valuable experience!

By the way, very interesting approach used in your paper there. A side question: does the ns pre-pulse alone at that fluence alter the brightness of your "SiNx on Silicon" sample? I image a slightly-below-ablation-threshold ns pulse would still cause silicon to melt and re-solidify. If I am not mistaken there, these transitions induced by ns laser pulse do not cause significant silicon microstructure change, do they?