Automatic ice-cream characterization by electrical impedance spectroscopy

1
Automatic Ice-Cream Characterization by
Electrical Impedance Spectroscopy
Marco Grossi Department of Electrical Energy and
Information Engineering University of Bologna,
Italy http//www.researchgate.net/profile/Marco_Gr
ossi
2
Food products are routinely screened for
important organoleptic characteristics.

• Off-line screening samples shipped to a
  laboratory for analysis

• On-line screening sensors are used to
  investigate some properties of the sample

Automatic product recognition is very important
in the case of ice-cream mixes
Different types of mixes require to set different
parameters in the machines where the mixes are
stored
Ice cream optimal machine setting requires the
products to be clustered in three different
groups milk based creamy mixes, frozen yogurt
and fruit based mixes
3
Electrical Impedance Spectroscopy
A sine-wave signal Vin(t)VMinsin(?t) is applied
to the sample
The current through the sample Iin(t)
IMinsin(?tf) is measured
The impedance of the sample Z(j?)
Vin(j?)/Iin(j?) is calculated
4
Z(j?) is measured over a range of frequencies
(?2pf) of the test signal and fitted with a
suitable equivalent electrical circuit
The estimated electrical parameters are used to
evaluate characteristics of the sample under test
The sample electrical response in non-linear
the test signal amplitude VMin is low to
guarantee the response to be confined in a
pseudo-linear region
The measure of Z(j?) outside the pseudo-linear
region can give useful additional informations
about the sample properties
5
Experimental Approach
21 ice-cream samples have been tested using three
different techniques

1. Electrical Impedance Spectroscopy (EIS) in the
   linear region with sine-wave signal of amplitude
   100mV and frequency in the range 20Hz 10kHz

1. Electrical response in non-linear region with
   sine-wave signal of frequency 20Hz and amplitude
   in the range 10mV 2V

1. Measure of the sample pH with a Crison micropH
   2000

6
Measurement Setup
Incubation temperature 35 C
7
Tested Samples
1 to 10 milk based creamy
11 to 14 frozen yogurt
15 to 21 fruit based
8
EIS in the Pseudo-Linear Region
The impedance Z(j?) of the sample under test in
direct contact with the sensor electrodes can be
modeled with the circuit
9
The proposed electrical model features a very
good correlation (R2 0.998) with the measured
data
10
Rm and Q can not reliably discriminate between
milk based and fruit based ice-cream mixes
11
Electrical Response in the Non-Linear Region
The modulus of the impedance Z(j?) deviates
from its pseudo-linear region value Z(j?)10mV
for VMin gt 200mV
12
The deviation is stronger at lower frequencies,
thus the measurements have been carried out at
20Hz
The deviation of Z(j?) from its pseudo-linear
region value Z(j?)10mV can be modeled as
function of VMin as
The slope of Z(j?) vs VMin in the non-linear
region is
13
The slope of Z(j?) in the non-linear region can
reliably discriminate between milk based and
fruit based products, in particular in the case
of sensor B
14
Conclusions
The feasibility to discriminate different groups
of ice-cream mixes has been shown

• The basic discrimination between milk based and
  fruit based mixes is possible with the electrical
  characterization of the sample in the non-linear
  region

• A second level discrimination of the first group
  between creamy mixes and frozen yogurt is
  possible measuring the sample pH

If you want to know more about this, please read
Grossi M., Lanzoni M., Lazzarini R., Riccò B.
(2012). Automatic ice-cream characterization by
impedance measurements for optimal machine
setting. Measurement, 45, 1747-1754.