Specific Process Knowledge/Etch/ICP Metal Etcher/silicon oxide

Feedback to this page: click here

It is possible to etch SiO2 in the ICP metal etcher but it is not designed for it and the results are not fantastic. It is a challenge to get a good selectivity to resist (typically in the range of 1:1 or worse) and it is probably not possible to get a profile angle of 90 degrees. More likely about 75-85 degrees. Different chemistries can be applied either based on CF4 or C4F8. If seems that C4F8 can give the best selectivity to resist (best case I have seem was 1:11 but it depends a lot on the process parameters)). If low coil power is needed CF4 chemistry is used because C4F8 needs a higher power to generate a plasma. /bghe 2016-04-25

Slow etch of SiO2 with resist as masking material - using a 6" carrier wafer with recess

This recipe can be used for slow etching of SiO2 with resist as masking material. Here are some test results presented.

Parameter	Resist mask
Coil Power [W]	200
Platen Power [W]	25
Platen temperature [oC]	0
CF4 flow [sccm]	20
H2 flow [sccm]	10
Pressure [mTorr]	3

Results	Test on wafer with 50% load (Travka 50), by BGHE @nanolab	100% load on 100mm wafers with Barc and KRF (no mask)
Etch rate of thermal oxide	44.1 nm/min (50% etch load) (01-02-2014)
Selectivity to resist [:1]	~0.9 (SiO2:resist)	~1.25:1 (Barc:KRF)
Etch rate in silicon	bghe@Nanolab 20190117 33.8 nm/min (middle of wafer with 80% load) bghe@Nanolab 20190117 34.3 nm/min (edge of wafer with 80% load)
Wafer uniformity (100mm)	±1.6% (01-02-2014)
Profile [o]	Take a look at the images but be aware that the resist profile was not good to begin with.
Wafer uniformity map (click on the image to view a larger image)	Contour plot of the etch rate over the wafer, 9 points measured
SEM profile images	Resist profile before etch Resist profiler after etch 01-02-2014. Resist profiler after etch 01-02-2014. No line width reduction observed.
Etch rate in barc		50 nm/min (2014-09-09)
Etch rate in KRF resist	40 nm/min (2014-09-09)

SiO2 etch using DUV mask

Two chemistry regimes has been explored: One using CF4 and one using C4F8

• CF4: bad selectivity to the resist mask.
• C4F8: Better selectivity to the resist mask can be achieved

• Tests done by Peixiong
• Tests done by Berit
• Test by Zhibo Li @Danchip dec. 2016 - based on the work of Peixiong and Berit: File:Zhibo Li SiO2 ICP etch (dose205).docx

Parameter	Recipe on ICP metal: A SiO2 etch with C4F8 with resist mask
Coil Power [W]	1000
Platen Power [W]	200
Platen temperature [oC]	0
C4F8 flow [sccm]	10
H2 flow [sccm]	28
Pressure [mTorr]	2.5

Results	Test
Etch rate of thermal oxide	145-172 nm/min by bghe@danchip (2015-06-02) 145 nm/min by Martin Lind Ommen (fall 2016)
Selectivity to resist [:1]	4-5:1 (SiO2:resist) by bghe@danchip (2015-06-02)
Cr etch rate	1.6 nm/min (1:90 to SiO2) by Martin Lind Ommen (fall 2016)
Profile [o]	86-87 dg by bghe@danchip (2015-06-02)
Wafer uniformity map (click on the image to view a larger image)
SEM profile images	by bghe@nanolab (2015-06-02)
Etch rate in barc
Etch rate in KRF resist	34 nm/min by bghe@nanolab (2015-06-02)
Comments	Sample: s007592 by bghe@nanolab (2015-06-02) See Martin Lind Ommen's results with hard masks in Process2share: [1] There were problems with polymer on the surface after etching.

SiO2 etch nLOF

Parameter	Resist mask
Coil Power [W]	800
Platen Power [W]	100
Platen temperature [oC]	0
CF4 flow [sccm]	30
H2 flow [sccm]	10
Pressure [mTorr]	4

Results	Test on wafer with 50% load (Travka 50), by BGHE @nanolab
Etch rate of thermal oxide	>110 nm/min (50% etch load) (09-03-2015)
Selectivity to resist [:1]	<0.7:1 (SiO2:resist)
Wafer uniformity (100mm)	Not known
Profile [o]	Not known
Wafer uniformity map (click on the image to view a larger image)	Not known
SEM profile images	NONE
Etch rate in nLOF resist	1.6µm was removed after 10min
Comment	After 10min etch the resist was gone and the etch depth as 1.145µm in the oxide

SiO2 etch with e-beam resist

Process flow

Si, APOX 1152nm by filmtek Si zep520A 560nm by dektakXT jbx9500: 60na ap7 MF for all the exposure same step size: 20nm px1283mk: alignment mark for finfet dose 280uc 3x3 at x pitch 10mm y pitch 10 mm in wafer center px1283lablejan1542014t1 250uc at 40mm x y pxline400p1000jan142014dt2 y= -40 -45 -50 -55mm dose 200 240 280 320uc N50 20c 2min IPA, N2 gental blow dry 18:10 Jan152014 Bruker Dektakxt Zep 560.51nmk Metal ICP, 19:00 Jan152014, pxSiO2try9, -10C, 5min Filmtek:large Apox area 5mmx3mm without zep SiO2 remains 494.53nm SiO2 etched 1152-495=657nm SiO2 etch rate: 131 nm/min sem zeiss, 1:50am Jan162014 still as over 200 nm zep remains on the wafer for line400p1000, need high dose as 320uc. 280uc is not enough to go through 560 nm thick zep520A

Parameter	Resist mask
Coil Power [W]	800
Platen Power [W]	150
Platen temperature [oC]	-10
C4F8 flow [sccm]	8
H2 flow [sccm]	30
Pressure [mTorr]	2.5

Results	Test on 6" wafer, by Peixiong Shi@danchip
Etch rate of thermal oxide	131 nm/min (15-01-2014)
Selectivity to resist [:1]	~1.8:1 (SiO2:resist)
Profile [o]	Not measured
Wafer uniformity map (click on the image to view a larger image)	Not known
SEM images	Top view of resist after etch 400nm lines - large sidewal roughness Tilt 30 deg.