Feedback to this page: click here
This page is written by Berit Herstrøm @ DTU Nanolab (BGHE) if nothing else is stated

At the moment DTU Nanolab has 2 PECVDs that can deposit silicon oxide with or without dopants of boron, phosphorus. PECVD1 and PECVD2 has been decommissioned and now we have PECVD3 and PECVD4. PECVD3 are for silicon based processing allowing wafers with small amount of metal (<5% wafer coverage). PECVD4 is for clean wafers both for silicon based materials and for III-V materials. Look at the PECVD page to learn more about the PECVDs at Nanolab. All though PECVD4 and PECVD3 are very much alike you cannot count on that a recipe on one system will give exactly the same results on the other system.

Deposition of SiO2 with PECVD4

Quality Control (QC) for PECVD4 - oxide

QC Recipe:	QCOXYD2
SiH₄ flow	12 sccm
N₂O flow	1420 sccm
N₂ flow	392 sccm
Pressure	550 mTorr
RF-power	60 W

QC limits	PECVD4 - OXIDE
Deposition rate	70 - 85 nm/min
Non-uniformity	<3.0
Refractive index	1.474 - 1.486

Recipe	Dep. rate [nm/min]	RI	Unif. [%]	Stress [MPa]	Comments	SiH4 [sccm]	N2O [sccm]	N2 [sccm]	B2H6	PH3	Pressure [mTorr]	Power [W]	Time [mm:ss]	Tested
LF SiO	75-78 nm/min	1.480-1.483	± 2.1-2.7%	compressive 309 MPa	Click for more results	12	1420	392			550 mTorr	60LF"	1:15/13:00(stress)	February 2017 bghe
HF SiO	63-64 nm/min	1.476-1.477	± 0.3-0.5%	Compressive: 250.5 MPa	Click for more results	10	1420	392			900 mTorr	30HF	2:00/16:00(stress)	February 2017 bghe
waveguide	159.5 nm/min	1.462	± 0.8%	Compressive: 121.9 MPa		17	2000				300 mTorr	700 LF	40:00	February 2017 bghe
BPSG	259 nm/min	1.4593	± 1.7%	Compressive: 36.7 MPa	Measured after anneal in Clad1000	17	1600		135	40	900 mTorr	800 LF	10:00/56:00(stress)	February 2017 bghe
BPSG low stress	302 nm/min	1.4598	± 1.7%	Compressive: 1.4 MPa	Measured after anneal in Clad1000	17	1600		240	60	500 mTorr	800LF	10:00	February 2017 bghe

SEM images of the conformity of the layer on a structured surface

Recipes on PECVD3 for deposition of silicon oxides

Quality Control (QC) for PECVD3 - oxide

QC Recipe:	QCOXYD2
SiH₄ flow	12 sccm
N₂O flow	1420 sccm
N₂ flow	392 sccm
Pressure	700 mTorr
RF-power	150 W

QC limits	PECVD3 - OXIDE
Deposition rate	97 - 112 nm/min
Non-uniformity	<2.0
Refractive index	1.467 - 1.474

Recipes

Recipe name	SiH4 flow [sccm]	N₂O flow [sccm]	N2 flow [sccm]	B2H6 flow [sccm]	PH3 flow [sccm]	Pressure [mTorr]	Power [W]	Description
LFSiO2	12	1420	392	0	0	700	150	New QC implemented in April 2016
LFSiO	12	1420	392	0	0	550	60	Uniform silicon oxide
1PBSG	17	1600	0	135	40	500	800 LF	BPSG glass for waveguide cladding layer

LF=Low Frequency

Expected results

Recipe name	Deposition rate [nm/min]	RI	Uniformity [%]	Stress	Comment
LFSiO2	105nm/min	1.47	~1.5%	400-402 MPa compressive stress by Anders Simonsen @nbi.ku.dk April 2016	We have seen that this recipe does not deposit the first minute
LFSiO	~75	~1.48	<1	not measured	.
1PBSG	~228 nm/min	.	~17%	not measured	.

Recipes on PECVD3 for deposition of doped oxide

Recipes

Recipe name	SiH₄ flow [sccm]	N₂O flow [sccm]	N₂ flow [sccm]	GeH₄ flow [sccm] (scaled by 100)	B₂H₆ flow [sccm]	PH₃ flow [sccm]	Pressure [mTorr]	Power [W]	Description
Core-Ge (not possible anymore as we have no Germanium)	17	1600	300	300	0	0	400	600 LF	Process for germanium doped core layer developed by Haiyan Ou from DTU Photonics Annnealing: Anneal bond furnace, recipe "core1100"
Top-BPSG	17	1600	0	0	100	40	500	800 LF	Process for PBSG top clading layer developed by Haiyan Ou from DTU Photonics Annnealing/oxidation: Anneal bond furnace, recipe "clad1000"

Expected results

Recipe name	Deposition rate [nm/min]	Refractive index
Core-Ge	~188 nm/min	~1.46969
Top-BPSG	~248 nm/min	~1.458

Recipes on PECVD2 for deposition of silicon oxides EXPIRED!!!

Quality control recipe (recipe changed in June 2015)

Quality Controle (QC) for PECVD2

QC Recipe:	QCTOXIDE
SiH₄ flow	12 sccm
N₂O flow	710 sccm
N₂ flow	392 sccm
Pressure	700 mTorr
RF-power	150 W @380kHz
Deposition time	15 min

QC limits	PECVD2
Deposition rate of Silicon Oxide	58 nm/min - 78nm/min
Non-uniformity of the deposition rate: (max-min)/2AVG*	<2%
Refractive index of the Silicon Oxide	1.45 - 1.48
Non-uniformity of the refractive index: (max-min)/2AVG*	<0.2%

SiO2, High Rate EXPIRED!!!

1SiO2HR
N $_{2}$ O-flow	1600 sccm
SiH $_{4}$ -flow	17 sccm
RF-power (380 kHz)	380 W
Process Pressure	400 mTorr
Deposition rate	179 nm/min (before 2014)
index of refraction	1.461 (before 2014)

SiO2 Low Rate EXPIRED!!!

1OX_old		1SiO2 (as QC)
N $_{2}$ -flow	392 sccm	392 sccm
N $_{2}$ O-flow	1420 sccm (setting in software is 710 sccm)	1420 sccm (setting in software is 710 sccm)
SiH $_{4}$ -flow	12 sccm	12 sccm
RF-power	100 W	150 W
Process Pressure	550 mTorr	700 mTorr
Deposition rate	~100 nm/min	66 nm/min (2015-05-18 BGHE) 68-69 nm/min (2015-04-24 BGHE) 109 ± 2 nm/min [tested: 2014-03-18] Old shower head
index of refraction	1.47	1.463-1.464 (2015-04-24 BGHE) 1.465 [tested: 2014-03-18] Old shower head
Uniformity	<1 %	1% over the wafer (2015-04-24 BGHE) 3.2% over the wafer [tested: 2014-03-18] Old shower head
Comment	This recipe is not running stable. The load capacitor tunes to zero and that it cannot go further. The results in a small reflected power that is varying due to the bad matching. This lead to variations in results from time to time. You may use it if you like, but please expect some variation in deposition rate.	This recipe was developed because the old one didn't tune well. It may have a less good uniformity but should be more stable.

Deposition rate as a function of deposition time using 1Ox_old:

Deposition time [s]	Oxide thickness [nm]	Expected naturally grown oxide [nm]	Deposition rate [nm/min]
15	26.9	2	99.6
30	51.9	2	99.8
60	102.4	2	100.4
60	102.7	2	100.7
120	201.1	2	99.6

Thickness uniformity of test wafer using the test recipe

BPSG: RI vs. B/P EXPIRED!!!

B2H6 flow and PH3 flow was varied to map the RI of different B2H6/PH3 ratios.
Work done by BGHE@dtu in fall 2013

Recipe name	SiH4 flow [sccm]	N₂O flow [sccm]	N2 flow [sccm]	B2H6 flow [sccm]	PH3 flow [sccm]	Pressure [mTorr]	Power [W]	Description
1PBSG	17	1600	0	See below	See below	500	800LF	BPSG glass for waveguide cladding layer

LF=Low Frequency

Run number	PH3 flow	B2H6 flow	B2H6/PH3	RI	Thickness
A1	40 sccm	135 sccm	3.38	1.4582	3.28 µm
A2	55 sccm	115 sccm	2.09	1.4615	3.13 µm
A3	45 sccm	130 sccm	2.89	1.4590	3.13 µm
A4	35 sccm	140 sccm	4.0	1.4572	3.22 µm
A5	50 sccm	120 sccm	2.4	1.4602	3.14 µm

Recipes on PECVD1 for deposition of silicon oxides EXPIRED!!!

Recipes

Recipe name	SiH4 flow [sccm]	N₂O flow [sccm]	N2 flow [sccm]	B2H6 flow [sccm]	PH3 flow [sccm]	GeH4*100 flow [sccm]	Pressure [mTorr]	Power [W]	Description
1oxide/1ox_std/standard	17	1600	0	0	0	0	400	380LF	Process control recipe. Developed for waveguides
1PBSG	17	1600	0	135	40	0	500	800LF	Developed for waveguide top cladding by Haiyan Ou @DTU Photonics'.
BGE_PBSG	17	1600	0	240	60	0	500	800LF	Low stress PBSG
HO_core	17	1600	300	0	0	400	400	600LF	Developed by Haiyan Ou @fotonik@dtu
HO_top	17	1600	0	107	40	0	500	800LF	Developed by Haiyan Ou @fotonik@dtu

Expected results

Recipe name	Deposition rate [µm/min]	RI	Uniformity [%]	Comments
1oxide/1ox_std/standard	~0.193	1.46	2	The latest measured values can be seen in the process control sheet in LabManager
1PBSG	~0.3	1.458@633nm