Feedback to this page: click here

This page is written by Evgeniy Shkondin @DTU Nanolab if nothing else is stated.
All images and photos on this page belongs to DTU Nanolab.
The fabrication and characterization described below were conducted in 2022 by Evgeniy Shkondin, DTU Nanolab.

This page presents the results of Scandium Nitride deposition using RF Reactive Sputtering in Sputter-System Metal-Nitride (PC3) commonly known as "Cluster Lesker". The deposition target is 4" Sc (0.250" Indium bonded with indium to 3mm Cu plate). Source #1 was used. Materials (ScN) from this source can either be deposited with RF or p-DC sputtering although only RF mode has been tested so far.

The main focus of the study was the deposition rate, optical functions, and the general performance of deposition from a bigger 4” target. The scandium nitride sputtering from the target of this size allows achieving a higher deposition rate and service to many potential applications where several hundreds of nm are needed to be put uniformly on a wafer of 150 mm or similar. Additionally, the flexibility of this source (z-shift and tilting) allows tailoring the deposition output.

The most common application of scandium is its use as a dopant or addition to aluminum in the deposition of AlScN thin films. The presence of scandium enhances the preffered crystallographic orientation of aluminum nitride (AlN) and significantly improves the overall piezoelectric performance of the thin film. With the Cluster Lesker tool (PC3), scandium can be loaded into Source 1, while aluminum can be loaded into Source 2. Co-sputtering can then be performed from both sources, allowing for precise adjustment of individual powers, gas flows, and other parameters. This flexibility enables the tailoring of thin film stoichiometry and other properties to meet specific requirements.

4” magnetron in PC3 has more functionalities compared to the ordinary guns in cluster Lesker. It can(to some degree) moves in Z-direction. This will affect the deposition rate, uniformity and most likely play an important role in reactive sputtering. Besides that, it can be easily tilted if necessary. In this study, two z-high positions have been compared (lowest-“home” and highest –“extended”). The effect of power has also been evaluated.

Note! It seems that Scandium is prone to oxidation, better to keep the prepared samples in a inert environment.

RF Sputtering of Scandium Nitride in Sputter-System Metal-Nitride(PC3)

The RF ScN process recipe in a Sputter-System Metal-Nitride(PC3) is following:

* Recipe Name:  MD PC3_Src1 - RF_Downstream with Reactive N2

PC Gun Z-shift Position: Home (00.00 mm) and Extended (95.20 mm)
Tilt: No (Target in horizontal plane)
Deposition mode: Downstream
Rotation speed: 10 rpm
Pressure: 3 mTorr
Power: 100-200 W (It is recommended to use 200W for getting high deposition rate and using "Extended" gun Z-shift position).
Ar flow: 40 sccm
N₂ flow: 18 sccm
Deposition time: 1000 s
Deposition temperature: 400 °C
Pre-sputtering time : 300s (better to give longer time to sputter away the native oxide, especially on a recently installed target)

Samples: 6" Si without native oxide (HF-treated).

The photography of Sc target mounted in 4” magnetron (PC3 Src1).
PC Gun Z-shift Position: Home (00.00 mm).
PC Gun Z-shift Position: Extended (95.20 mm).

Deposition Rate

Scandium nitride thin film deposition rate (PC3 Src1) deposited at 400 °C.
Scandium nitride deposition rate as a function of power at two different PCZ-shift Gun positions.

Scandium nitride deposition conditions

PC Gun Z-shift Position: Extended (95.20 mm).
Power (W)	Pressure (mTorr)	Time (s)	Total Thickness (nm)	Deposition rate (nm/s)
200	3	500	37.7	0.0754
150	3	750	25.0	0.0334
100	3	1000	14.1	0.0141

PC Gun Z-shift Position: Home (00.00 mm).
Power (W)	Pressure (mTorr)	Time (s)	Total Thickness (nm)	Deposition rate (nm/s)
200	3	500	16.6	0.0333
150	3	750	10.0	0.0133
100	3	1000	5.8	0.0058

X-ray Reflectivity (XRR)

X-ray Reflectivity analysis of ScN samples have been performed to investigate the thicknesses, roughness, and density profiles.

The scans has been obtained using Rigaku XRD SmartLab equipment. The voltage and current settings for the Cu X-ray tube were standard 40kV and 30mA. The incident optics contained a IPS (incident parallel slit) adaptor with 5 ° Soller slit. Other slits: IS=0.05mm RS1=0.05mm and RS2=0.075mm. Step size: 0.01 and measurement time - 5s for each point. Fitting procesure was performed using commercial GlobalFit software assuming the model based on Si substrate followed by Sc-Si transition layer for better fit, followed by the main ScN film with thin oxides and moisture surfaces. The results are summarized in a tables below.

X-ray reflectivity scans of ScN thin films (PC3 Src1).
ScN thin film XRR. Power: 200W, Deposition time: 1000s, Pressure 3 mTorr. PC Gun Z-shift Position: Extended (95.20 mm).
ScN thin film XRR. Power: 150W, Deposition time: 1000s, Pressure 3 mTorr. PC Gun Z-shift Position: Extended (95.20 mm).
ScN thin film XRR. Power: 100W, Deposition time: 1000s, Pressure 3 mTorr. PC Gun Z-shift Position: Extended (95.20 mm).

X-ray reflectivity scans of ScN thin films (PC3 Src1).
ScN thin film XRR. Power: 200W, Deposition time: 1000s, Pressure 3 mTorr. PC Gun Z-shift Position: Home (00.00 mm).
ScN thin film XRR. Power: 150W, Deposition time: 1000s, Pressure 3 mTorr. PC Gun Z-shift Position: Home (00.00 mm).
ScN thin film XRR. Power: 100W, Deposition time: 1000s, Pressure 3 mTorr. PC Gun Z-shift Position: Home (00.00 mm).

Extracted values

XRR results for ScN thin films deposited at 3 mTorr and different powers. PC Gun Z-shift Position: Extended (95.20 mm)

	Top layer (moisture)			ScN_xO_y			Main ScN layer			Sc-Si intermediate layer			Si substrate
Deposition name	Moisture thickness (nm)	Moisture density (g/cm³)	Moisture roughness (nm)	ScN_xO_y thickness (nm)	ScN_xO_y density (g/cm³)	ScN_xO_y roughness (nm)	ScN thickness (nm)	ScN density (g/cm³)	ScN roughness (nm)	Sc-Si intermediate layer thickness (nm)	Sc-Si intermediate layer density (g/cm³)	Sc-Si intermediate layer roughness (nm)	Si thickness (nm)	Si density (g/cm³)	Si roughness (nm)
200W, 3mTorr, 1000s	1.33	0.60	0.15	0.43	2.73	0.72	37.28	4.82	1.87	0.75	2.87	0.92	$\infty$	2.33	0.03
150W, 3mTorr, 1000s	0.63	0.67	0.05	0.85	1.61	1.24	24.17	4.92	0.78	1.20	3.20	0.76	$\infty$	2.33	0.04
100W, 3mTorr, 1000s	1.69	0.19	0.00	1.63	2.99	0.51	12.43	4.83	0.00	2.06	1.97	1.27	$\infty$	2.33	0.00

XRR results for ScN thin films deposited at 3 mTorr and different powers. PC Gun Z-shift Position: Home (00.00 mm)

	Top layer (moisture)			ScN_xO_y			Main ScN layer			Sc-Si intermediate layer			Si substrate
Deposition name	Moisture thickness (nm)	Moisture density (g/cm³)	Moisture roughness (nm)	ScN_xO_y thickness (nm)	ScN_xO_y density (g/cm³)	ScN_xO_y roughness (nm)	ScN thickness (nm)	ScN density (g/cm³)	ScN roughness (nm)	Sc-Si intermediate layer thickness (nm)	Sc-Si intermediate layer density (g/cm³)	Sc-Si intermediate layer roughness (nm)	Si thickness (nm)	Si density (g/cm³)	Si roughness (nm)
200W, 3mTorr, 1000s	3.03	0.08	0.00	0.60	4.66	1.28	16.04	4.87	0.00	2.22	2.12	1.08	$\infty$	2.33	0.00
150W, 3mTorr, 1000s	1.29	1.17	0.00	0.53	2.71	1.28	9.44	4.42	0.08	1.25	2.94	0.48	$\infty$	2.33	0.03
100W, 3mTorr, 1000s	0.64	1.62	0.00	1.06	2.23	1.91	4.79	4.63	0.61	1.16	3.04	0.62	$\infty$	2.33	0.05

Fitting parameters
Sample	Moisture		ScN_xO_y		ScN		Sc-Si intermediate layer		Si substrate		Fitting parameters
Sample	Delta	Beta	Delta	Beta	Delta	Beta	Delta	Beta	Delta	Beta	Intensity	Background	Fitting area ( $2\Theta$ )	R	$\mathrm {X} ^{2}$
PC Gun Z-shift Position: Extended (95.20 mm) 200W, 3mTorr, 500s	1.9484E-6	4.5244E-8	8.9012E-6	2.0670E-7	1.3473E-5	1.2304E-6	9.3367E-6	2.1681E-7	7.5860E-6	1.7616E-7	1.30347E-000	3.58520E-007	0.4055 - 4.0000	0.01563	0.00513
PC Gun Z-shift Position: Extended (95.20 mm) 150W, 3mTorr, 750s	2.1902E-6	5.0859E-8	5.1589E-6	1.1980E-7	1.3736E-5	1.2545E-6	1.0416E-5	2.4187E-7	7.5860E-6	1.7616E-7	1.08491E-000	9.64995E-007	0.4055 - 4.0000	0.01452	0.00413
PC Gun Z-shift Position: Extended (95.20 mm) 100W, 3mTorr, 1000s	6.3024E-7	1.4635E-8	9.7505E-6	2.2642E-7	1.3484E-5	1.2314E-6	6.4131E-6	1.4892E-7	7.5860E-6	1.7616E-7	1.17155E-000	2.25376E-006	0.4055 - 4.0000	0.00744	0.00101
PC Gun Z-shift Position: Home (00.00 mm) 200W, 3mTorr, 500s	2.5771E-7	5.9844E-9	1.5175E-5	3.5239E-7	1.3596E-5	1.2417E-6	6.8990E-6	1.6020E-7	7.5860E-6	1.7616E-7	1.00181E-000	3.89203E-011	0.4055 - 4.0000	0.01830	0.00741
PC Gun Z-shift Position: Home (00.00 mm) 150W, 3mTorr, 750s	3.8018E-6	8.8284E-8	8.8171E-6	2.0475E-7	1.2333E-5	1.1264E-6	9.5589E-6	2.2197E-7	7.5860E-6	1.7616E-7	1.37654E-000	5.17162E-009	0.4055 - 4.0000	0.00825	0.00129
PC Gun Z-shift Position: Home (00.00 mm) 100W, 3mTorr, 1000s	5.2890E-6	1.2282E-7	7.1123E-6	1.6516E-7	1.2924E-5	1.1803E-6	1.0009E-5	2.3243E-7	7.5860E-6	1.7616E-7	1.19799E-000	1.15388E-006	0.4055 - 4.0000	0.00466	0.00039

X-ray Photoelectron Spectroscopy

X-ray Photo Electron Spectroscopy (XPS) was employed to investigate the chemical composition of deposited scandium nitride (ScN) thin films using the XPS K-Alpha instrument. The analysis was conducted in a 'depth profile' mode, where the film was incrementally etched, and the composition was measured layer by layer. Multiple scans were performed, including survey scans and scans for specific elements like O1s, N1s, Sc2p, Sc3s, and C1s, to gather comprehensive chemical information.

Interpreting the stoichiometry of the ScN film is challenging due to the overlap between the N1s peak and the Sc2p peak. Additionally, while scandium has various interesting electronic states that can be measured and utilized in calculations (such as Sc3s), there is only one electronic state for nitrogen, N1s, which requires careful deconvolution from the Sc2p peak.

The ScN sample used for XPS analysis was deposited at the PC Gun Z-shift Extended position (95.20mm) under the following conditions: 400°C, Ar/N 40/18 sccm, 200W power, 3mTorr pressure, and a deposition time of 1000 seconds.

Estimated stoichiometry is following: N=37.73%, Sc=48.11%, O=13.89%.

The Sc target used in the experiment was new, and due to its susceptibility to oxidation, there is a possibility of a pronounced level of Scandium oxynitride formation. We believe that more extensive usage of the target will remove the oxidized top layer. Additionally, it is necessary to perform preconditioning (pre-sputtering) of the target before using it in applications. This pre-sputtering process will help minimize the oxygen level in the resulting film.

X-ray Photoelectron Spectroscopy of ScN thin films (PC3 Src1). High resolution scans fitted with Lorentz-Gaussian functions.
Survey spectrum of ScN thin film. The films has been etched for 160s wiht Ar⁺ ions at low current and 1000eV energy.
Sc 3s signal with fitted functions. Useful for stoichiometry calculation, due to the profile simplicity. The film has been etched for 160s wiht Ar⁺ ions at low current and 1000eV energy before the scan.
O 1s signal with fitted functions. The film has been etched for 160s wiht Ar⁺ ions at low current and 1000eV energy before the scan.
Sc 2p and N 1s (in yellow) signals with fitted functions. Less useful for stoichiometry calculation, due to the profile complexity. N1s is dangerously close to Sc 2p, and need to be deconvolved. The film has been etched for 160s wiht Ar⁺ ions at low current and 1000eV energy before the scan.
N 1s (in yellow) signal with fitted functions, where Sc 2p is also present (grey fitted area). Difficult for stoichiometry calculation, due to the profile complexity. N1s is dangerously close to Sc 2p, and need to be deconvolved. However, there are no other electronic states of nitrogen that can be used in XPS, so the deconvolution approach is the only option. The film has been etched for 160s with Ar⁺ ions at low current and 1000eV energy before the scan.

X-ray Photoelectron Spectroscopy of ScN thin films (PC3 Src1). High resolution scans during the Ar ion etch (depth-profiles).
Sc 2p andN 1s signals. 30 levels in total. 20s of etch with Ar⁺ ions at low current and 1000eV energy between the measurements.
Sc 3s signal. 30 levels in total. 20s of etch with Ar⁺ ions at low current and 1000eV energy between the measurements.
O 1s signal. 30 levels in total. 20s of etch with Ar⁺ ions at low current and 1000eV energy between the measurements.
C 1s signal. 30 levels in total. 20s of etch with Ar⁺ ions at low current and 1000eV energy between the measurements.

X-ray Photoelectron Spectroscopy of ScN thin films (PC3 Src1). 2-Dimensional depth profiles.
2D profile of Sc 2p and N 1s signals. 30 levels in total. 20s of etch with Ar⁺ ions at low current and 1000eV energy between the measurements.
2D profile of Sc 3s signal. 30 levels in total. 20s of etch with Ar⁺ ions at low current and 1000eV energy between the measurements.
2D profile of O 1s signal. 30 levels in total. 20s of etch with Ar⁺ ions at low current and 1000eV energy between the measurements.
2D profile of C 1s signal. 30 levels in total. 20s of etch with Ar⁺ ions at low current and 1000eV energy between the measurements.

Scanning Electron Microscopy

The images below present Scanning Electron Microscopy (SEM) images of the ScN layers from the top and tilted view.

Scanning Electron Microscopy.
ScN layer deposited at 200 W RF power, 3mTorr, and 1000s. Gun Z-shift position is "Extended". Top view.
ScN layer deposited at 200 W RF power, 3mTorr, and 1000s. Gun Z-shift position is "Extended". Tilted view.
ScN layer deposited at 200 W RF power, 3mTorr, and 1000s. Gun Z-shift position is "Extended". Tilted view