Mobility and Sheet Charge Measurements of Gallium Arsenide,
Gallium Nitride and other Semiconducting Materials

LEI 1600 instrument provides non-destructive, non-contact measurements of mobility and sheet charge in Gallium Nitride [GaN] and other semi conducting materials. As the technology roadmap leads the compound semiconductor forward, Gallium Nitride grown or deposited on different semi conducting and insulating substrates is finding new device applications. The epitaxially grown p-GaN, for example, may provide lower production costs due to higher growth rates. The mobility in these and other structures is of utmost importance.

LEI 1600 Mobility System is capable of making non-destructive mobility and sheet charge measurements on full wafers from 2" to 6" in diameter.  It can also measure square samples as small as 1.4"x 1.4". The measurement spot size covers an area of approximately 0.5 square inches.  Wafers can be manually positioned at various locations on the wafer holder to give an indication of the lateral variation in mobility and sheet charge for each layer.

Microwave energy and magnetic fields interact with multi-layer wafer structures to provide measurements of mobility, conductivity, and sheet charge for the cap and 2DEG layers in production wafers.  Sheet resistance can be measured using a magnetic field strength value of 0 Gauss.

The LEI 1600 Mobility System has numerous benefits when compared to conventional DC Hall measurements. 

These include:

1. The ability to evaluate production wafers. After measurement, wafers can be returned to the process line for further processing.

2. Elimination of the need for thin cap Hall wafers.

3. The option of measuring small samples.

4. No wafer preparation.  The LEI 1600 requires no special shapes or contacts.  For testing on GaN, it can be an asset.  GaN is hard to cleave, and contacts are usually non-ohmic.

5. Individual mobility and sheet charge measurements for the cap and two-dimensional electron gas (2DEG) layers.

6. Room temperature measurements. The LEI Model 1600 uses multiple magnetic field strength values from 0 ~ 10 KGauss.

7. Measured (not calculated) Hall output.

8. Zero-field sheet resistance values that correlate with eddy current measurements.

Theory of Operation and Method of Use

The Model 1600 uses microwave technology operating at 10 GHz to make multi-layer mobility measurements for production cap and 2DEG layers.  These layers are routinely used in HEMT and PHEMT device structures.

The Model  1600 uses a  microwave source and  three  microwave detectors for Hall-, forward-, and reflected power.  Signal levels for the three power detectors are determined with respect to a magnetic field value. The data analysis is based on a theoretical model, which examines the interaction of the microwave power and magnetic fields with the sample structure. The values of mobility, conductivity, and sheet charge for a single layer or multiple layers are derived from the statistical best fit to the theoretical model.

Comparison of LEI 1600 measurements with contact DC Hall mobility

Mobility measurements made with LEI 1600 agree closely with DC Hall mobility data. As an example, the following table provides a comparison on thin cap and thick cap GaAs pHEMT structures:

	LEI 1600 Nondestructive Testing		DC Testing done at ARL (Beck and Svensson)
	Mobility (cm2/V-s)	Carrier Density (cm-2)	Mobility (cm2/V-s)	Carrier Density (cm-2)
Thin Cap Wafer	6187	2.89E+12	5862	3.01E+12
Thick Cap Wafer
2DEG Layer	7084	1.95E+12	7418	1.81E+12
Cap Layer	1489	1.98E+13	1557	1.89E+13

Below is another example of a GaAs HEMT wafer that was measured on the LEI 1600:

	LEI 1600 Nondestructive System		Accent HL550 Hall System		BioRad 5200 Hall System
Test Site	m (cm2/Vs)	Ns x1012 (cm-2)	m (cm2/Vs)	Ns x1012 (cm-2)	m (cm2/Vs)	Ns x1012 (cm-2)
1	5.32E+03	3.16	4.50E+03	-3.21	Bad Contacts
A	5.37E+03	3.13	5.75E+03	-3.16	5.84E+03	-3.28
B	5.34E+03	3.19	5.31E+03	-3.21	4.87E+03	-3.33
C	5.25E+03	3.21	5.58E+03	-3.26	5.43E+03	-3.39
D	5.40E+03	3.14	5.63E+03	-3.26	5.47E+03	-3.38
Average	5.34E+03	3.17	5.35E+03	-3.22	5.40E+03	-3.35
Std. Dev.	56.84	0.03	503.82	0.04	401.68	0.05
Min	5.25E+03	3.13	4.50E+03	-3.26	4.87E+03	-3.39
Max	5.40E+03	3.21	5.75E+03	-3.16	5.84E+03	-3.28

Note: Hall sample symmetry and contact quality may have contributed to some of the measured mobility and Ns variation.

Estimated Savings (LEI 1600 vs. conventional Hall Measurements)

The substrate cost of a 6-inch compound semiconductor wafer tested for conventional Hall mobility is approximately $150.00.  In addition, the empty reactor runs required for each thin-cap Hall sample represent significant lost revenue.  For a nominal 30 Hall wafers per month, the lost revenue would exceed $1 million annually (assumes multiple reactors and a 52-week working year.  This figure could, of course, be different if the wafer cost is different, or the number of wafers tested per week is more/less than 30.)

Mobility measurements using the LEI 1600 provide substantial, measurable savings over conventional Hall methods.

Assuming:

• pHEMT Wafers
• Nine wafers per reactor load
• One single-wafer Hall run every two days
• Average wafer selling price = $750

$16,875 x 52 weeks/year = $877,500 in lost revenue

The Model 1600 will pay for itself in < 3 months

LEI Model 1600 Instrument Page