4.3.1 Development of a LabView Program for Semiconductor Hall ...

38 Annual Report 2007 - Solid-State Electronics Department
4.3.1 Development of a LabView Program for Semiconductor Hall
Measurements
Student
Supervisor
Technical Assistant:
R. Shabanah
I. Nannen
U. Doerk
Introduction
Semiconductor transport measurements are routinely done by van-der-Pauw/Hall measurements. In
this work a new software platform based on LabView programme has been developed providing a
user friendly interface and extended data storage utilities. It replaces the existing platform based on
a HP Basic programme. Special emphasis was given in the optimization of the measurement data
aquisition. This way the range of measurement conditions offering precise transport data has been
successfully extended.
The Program
The transport of charged carriers in a rsemiconductor crystal is affected as by an magnetic field
r r
described by the of Lorenz force: FL
= q⋅ ( v×
B)
. If the magnetic field is perpendicular to
direction of current flow, the Lorenz force results into a accumulation of charge in a semiconductor
sample which gives rise to the Hall voltage given here for electron transport:
VH
1
= ⋅B⋅I
q⋅n⋅ d . (1)
The challenge of Hall measurement is a precise set up of the magnetic field B, the measurement
current I, and the precise measurement of the Hall voltage V H in the presence of parasitic conditions
like a remaining magnetic field, and a non-symmetric sample. For the control and read-out of the
measurement data the LabView (Laboratory Virtual Instrumentation Engineering Workbench) has
been adopted. LabView is a graphic programming language. The programming is carried out
according to the data flow model. LabView programs are called virtual instruments or simply VIs.
VIs consists of two components:
- the front panel, that contains the Graphical User Interface
- the block diagram of the graphic program code.
Fig. 1 shows as an example the programming of the read out of the Hall voltage from a digital
multimeter (Keithley 199). Every 0.5 ms measurement data are recorded. In a sub-VI the standard
deviation and the mean value of four subsequent data are determined. The data are accepted if the
standard deviation is less than 10% of the average value. An error message window will appear if
the data are not valid. In the same manner the programming is done for the magnetic field
adjustment and the various current source adjustments according to the van der Pauw technique.
This program makes it possible to store results of measurement on the computer as HTML file
which facilitates the exchange of the results over E-Mail and network. In order to proof the ohmic

Annual Report 2007 - Solid-State Electronics Department 39
conductivity of all contacts and the degree of symmetry of the sample, an integrated I-V test using
the HP parameter analyser HP 4145B has been implemented.
Fig. 1
Block diagram of the read-out of measurement data using Lab View
Results of measurement
We report here on the precision of the Hall measurement in a wide range of measurement
conditions. For this purpose, two InP-based samples with very different transport data have been
selected: a highly p-type doped InGaAs layer as part of a pin diode and a high mobility twodimensional
electron gas of a heterostructure field-effect transistor structure (InAlAs/InGaAs/InP).
Tab.1 shows the results of the initial 4-point measurement without a magnetic field of the van-der-
Pauw technique for the determination of the sheet resistance. The measurement current I AB is
applied to the contacts AB of the sample and the voltage drop V CD is measured at terminals CD.
IAB [µA]
HP-basic
VCD [mV]
LabView VCD[mV] ∆VCD [%]
5 -0.26 -0.263 -1.2
10 -0.516 -0.518 -0.4
20 -1.02 -1.03 -1.0
40 -2.05 -2.05 0.0
100 -4.10 -4.10 0.0
200 -5.12 -5.13 -0.2
400 -20.5 -20.5 0.0
1,000 -52.0 -52.2 -0.4
2,000 -104 -105 -1.0
4,000 -116 -117 -0.9
Tab.1
Comparision of digital multimeter read-out data of the LabView and the HP basic
programme for the sheet resistance determination of p-doped sample without magnetic
field.

40 Annual Report 2007 - Solid-State Electronics Department
Tab. 1 shows that the agreement of voltage measurement of both platforms is very high (< 1%). In
addition, the linearity of the measurement results over three orders of magnitude indicates that a
very wide range of input currents give a precise measurement result.
Next, the precision and range of measurement with applied magnetic field is investigated. For this
purpose the high mobility sample has been chosen. Fig. 2 shows the transport data of twodimensional
electron gas. As the thickness d is unknown, the sheet carrier concentration ns
= n⋅
d
(cf. eq. 1) is determined. Both software platforms were investigated in a wide range of applied
magnetic fields B. The high precision of the measurement set-up enables a very reliable
measurement at a magnetic field strength as low as B = 50 mTesla. There should be no physical
dependence of measurement data up to approximately µ ⋅ B ≤ 1 which holds here because the
mobility doe not exceed 10,000 cm²/Vs (cf. Fig. 2). Therefore, the higher variation of the evaluated
data in comparison to table 1 is attributed to the precision of the set-up of the magnetic field. The
relative deviation of the mobility with the basic program is 5.88 % while with the new LabView
program a strongly reduced deviation of 1.38 % is achieved.
mobility µ [cm²/Vs]
9500
9200
8900
8600
HP-BASIC
LabView
µ
0 200 400 600 800
magnetic field B /mT
n s
3.5
3.0
2.5
sheet carrier concentration n s [cm²]
Fig. 2
Transport data of a two-dimensional electron gas determined at various magnetic field
strength. (I = 20 µA, T=300K).
Summary
A new LabView program is developed, enabling user friendly van-the-Pauw/Hall measurements
with extended data storage utilities. The program has been tested with exemplary samples like a
high mobility two-dimensional electron gas and a low mobility p-type doped layer. The
measurements shows a good agreement to the previous basic program. In addition, a wide
measurement parameter regime for precise HALL measurements and an improved data acquisition
could be demonstrated.