ESTIMATING WINE YEAST CONCENTRATION BY FAR FIELD
CONTRAST MEASUREMENT
DAN CHICEA1,2, OVIDIU TIŢA2,3, MIHAELA TIŢA3, ECATERINA LENGYEL3
1
Department of Environmental Sciences, University Lucian Blaga, Dr. Ion Ratiu str., no 5–7,
Sibiu, 550012, Romania
2
Pediatric Respiratory Medicine Research Center (CCMRP), Str. Pompeiu Onofreiu Nr. 2–4,
Sibiu, Romania
3
Faculty of Agricultural Sciences, Food Industry and Environmental Protection,
University Lucian Blaga, Dr. Ion Ratiu str., no 5–7, Sibiu, 550012, Romania
Received August 20, 2013
A coherent light scattering experiment on Saccharomyces Cerevisia suspension with a
concentration that covers four orders of magnitude was performed. The far field first
order statistics was computed and the variation of the average contrast with the yeast
particle concentration was analyzed. A fast procedure for monitoring the yeast
concentration is suggested.
Key words: Sacharomyces Cerevisiae – ellipsoideus, Coherent Light Scattering, Suspension,
Speckle Contrast, Particle Concentration.
1. INTRODUCTION
Yeasts are a eukaryotic microorganisms classified in the kingdom Fungi.
Approximately 1500 species of yeasts have been described [1]. Most of the yeast
species reproduce asexually by budding, although in a few cases by binary fission
[1]. Yeasts are unicellular, although some species with yeast forms may become
multicellular through the formation of a string of connected budding cells known as
pseudohyphae, or true hyphae as seen in most moulds. Yeasts are classified into
two groups, ascomycetous and basidiomycetous yeasts [2]. For these two groups
the asexual reproduction is different from one to the other. Anamorphic
ascomycetous yeasts stretch all wall layers during asexual reproduction, process
that is called holoblastic budding, while anamorphic basidiomycetous yeasts grow
out the inner wall layer, thus producing an enteroblastic bud with a collarette [3].
Ascomycetes yeasts are classified into the Saccharomycetes and
Schizosaccharomycetes, while the basidiomycetes yeasts are distributed into
Hymenomycetes, Urediniomycetes and Ustilaginomycetes [1]. Yeasts occupy different
Rom. Journ. Phys., Vol. 59, Nos. 3–4, P. 346–354, Bucharest, 2014
2
Estimating wine yeast concentration by far field contrast measurement
347
habitats, as aquatic, atmospheric, plant and soil. Consequently, they play various
ecological and biotechnological roles such as bioremediation of toxic compounds
[4], fermentation of beverages [5] and food processing [6].
Nowadays, comprehensive research is being done with respect to the methods
of obtaining lipids and phospholipids from lipid biocomponents, in order to
identify new methods for obtaining liposomal substances, needed by the
pharmaceutical, cosmetic and medical industry. At present, egg lecithin is
commonly used instead. The use of this source has several drawbacks though, one
of them laying in the fact that it oxidizes easily. Eukaryotes (yeasts, fungi, algae)
are the main microorganisms that produce lipids and phospholipids. The researches
conducted in the field of phospholipids synthesis, of obtaining phospholipids from
microorganisms and of optimizing culture media for their cultivation is of interest
for microbiology and biotechnology. Monitoring fermentation by monitoring yeast
concentration is therefore of interest during a biotechnological process.
The work described here presents an unconventional method based on
coherent light scattered on a suspension and on analyzing the scattered far field,
which might be used in assessing in a fast manner the yeast concentration variation
in a process that carries on in aqueous suspension, as fermentation of beverages is.
2. EXPERIMENTAL SETUP
The schematic of the experiment is presented in Fig. 1. The coherent light
source was a laser diode that had a wavelength of 633 nm and a constant power of
20 mW. No polarizer was placed between the laser and the glass cuvette, which
had the optical pass equal to 1 cm; D was 0.162 m. The camera was a Philips
SPC900NC CCD with the optical system removed, therefore the far field speckle
was recorded. The center of the CCD array was placed 0.007 mm apart from the
beam making the measurement angle to be 2o 28’ 27.1’’.
Fig. 1 – The schematic of the experiment.
348
Dan Chicea et al.
3
A recording in MPAG4 format was acquired for each sample. The resolution
was 640×480 pixels, the frame rate was 10 frames per second (fps) and the color
depth was 16 bits.
3. SAMPLE PREPARATION
The wine yeast that was used in this study is Saccharomyces cerevisia –
ellipsoideus. The yeast was shipped in dry state. Initially 0.2 g of dry yeast was
dissolved in 20 ml of deionised water and allowed to become moisturized for 24
hours. This initial sample had a concentration of 10 g/l, expressed in grams of dry
yeast per litre and the concentrations mentioned in this work are consistent in this
respect. From this initial diluted sample 4 ml were extracted and placed in a
transparent glass cuvette having a square section of 1x1 cm2 and was used as a
target for the light scattering experiment. The other samples were obtained by
successive dilution. 29 samples were used in the study reported here and the
concentration range was [7.6·10-03 - 10] g/l. A recording lasting for 30 seconds was
acquired for each sample and processed later on.
4. DATA PROCESSING PROCEDURE
A typical approach to relate the particle concentration with the first order
statistics is presented in [7–9]. The average speckle size is currently calculated as
the normalized autocovariance function of the intensity speckle pattern acquired in
the observation plane [7, 8]. In this paper a different approach is proposed to relate
the particle concentration with one of the first order far field statistics parameter,
which is the average contrast rather than the speckle size.
The recording was processed with a computer program written for this
purpose. The program reads the recording, extracts frame by frame in a 640×480
array of intensity levels. We can note I(i,j)=I(xi, yj) the intensity recorded by the
cell (i,j) of the CCD, hence by the pixel (i,j) of the array of pixels the image
consists of.
A frame taken from the recording having a bigger concentration, which is
2.38 g/l, is presented in Fig. 2. Another frame taken from the sample having a
smaller concentration, of 0.193 g/l is presented in Fig. 3.
Examining Figs. 2 and 3 we notice the speckle aspect, given by a succession
of maxima and minima randomly distributed across the far interference field. We
also notice, without any statistics, that the contrast we can see is bigger for the
sample with a smaller concentration.
4
Estimating wine yeast concentration by far field contrast measurement
349
Fig. 2 – A frame taken from the recording of the sample having the yeast concentration of 2.38 g/l
(color on-line).
Fig. 3 – A frame taken from the recording of the sample having the yeast concentration of 0.193 g/l
(color on-line).
The average contrast of the image, either acquired as a bitmap or extracted
from the frames of the recording, is currently calculated [10, 11] as:
K=
( I ( i, j ) − I )
I
2
(1)
350
Dan Chicea et al.
5
Once the array of intensities is extracted for each frame, the contrast of each
particular frame is calculated using (1). In (1) K is a space contrast and not a time
contrast, as pointed out in [12, 13]. In (1) the angular brackets stand for the average
over the entire 640x480 pixels collection of intensity values for an image that is
processed. Further details on computing the speckle size are presented [13].
The collection of contrast values, one value for each of the frames, is
averaged and the average contrast is considered to be the contrast for that particular
recording, hence for that particular sample. The standard deviation for the
collection of values computed for one sample is considered to be the relative error
in assessing the contrast for that particular sample.
The procedure presented in this article is an improvement of the procedure
previously described [13] where a CMOS was used rather than a CCD. The
recording used in this work was a MPEG4 format and not an uncompressed AVI
[13], which considerably reduces the amount of free space required for data saving
and processing on the hard drive of the PC. The recording processing program was
written in MATLAB. The recording was converted to AVI format with CINEPAK
compression prior of processing it. Moreover, each recording lasted for 30 s,
therefore the total number of frames, hence images that were processed for each
sample was 900, considerably bigger than 60 frames that were processed in [13].
The direct consequence is that the error bars are tremendously smaller, as
compared with the error bars in [13], as can be noticed in the Figs. presented in the
following section.
5. RESULTS AND DISCUSSION
The samples were prepared using the procedure described in section 3
starting with a concentration of 10 g/l and decreasing it by successive dilution
down to 7.59·10-3 g/l. This concentration range is quite extended, as it covers four
orders of magnitude. The average contrast was computed for each sample and the
results are presented in Fig. 4.
We notice that the average contrast presents an increasing trend in the very
small concentration range, up to 4.8·10-2 g/l followed by a plateau. From 8.6·10-2
g/l to 10 g/l the contrast exhibits a decreasing trend, that does not appear to be
smooth. This is due in part to the fact that data is presented in a logarithmic manner
on the X axis, as the data covers four orders of magnitude.
This result is similar with the results reported in [7]. At a first look [7]
presents an almost linear decrease. Moreover, in [7] the decrease for 0.2 µm
diameter microspheres is similar with the contrast decrease reported in this work.
This can be explained by the much bigger concentration range used in this work.
The work reported in [7] represents just a small area of the extended decreasing
part of the curve in Fig. 4.
6
Estimating wine yeast concentration by far field contrast measurement
351
160
140
120
K
100
80
60
40
20 -2
10
-1
10
0
concentration, g/l
1
10
10
Fig. 4 – The average contrast versus yeast concentration, for the entire concentration range.
Moving forward, we can extract the extended range where the variation is
monotone, which is from 8.6·10-2 g/l to 10 g/l and data is presented in Fig. 5,
having a linear data representation on the concentration axis though. Examining the
plot of the average contrast variation with the concentration we notice that the
curve can be used as a calibration line for measuring the concentration in the
8.6·10-2 - 10 g/l concentration range.
160
140
120
K
100
80
60
40
20
0
1
2
3
4
5
6
concentration, g/l
7
8
9
10
Fig. 5 – The average contrast versus dry yeast concentration in the range
with a monotone decreasing trend.
For this purpose we consider the data in a different manner, which is the
variation of the concentration with the contrast and we can fit a third degree
polynomial to the data. The coefficients are presented in Eq. (2) and the R2 was
352
Dan Chicea et al.
7
found to be 0.99, which suggests as reasonably good fit. Fig. 6 presents the data,
with circles and line described by Eq. (2), which can be used as a recipe for
assessing the yeast concentration from the measured contrast K.
c ( K ) = – 1.5183 ⋅ 10 −05 K 3 + 0.0055907 K 2 – 0.68956 K + 28.677
(2)
10
9
8
concentration, g/l
7
6
5
4
3
2
1
0
20
40
60
80
K
100
120
140
Fig. 6 – Concentration versus contrast; circles are experimental data, smooth line is the fit
with the polynomial in Eq. (2).
The procedure was tested for three different samples that were prepared from
dry yeast of known mass, by dilution, having the concentration within the range of
monotone variation. The contrast was measured using the procedure described in
this paper. The results are presented in Table 1, where the first column is the yeast
concentration, the second column contains the value of the measured contrast, the
third column contains the assessed concentration, using Eq. (2) and the fourth
column presents the relative error, assuming that the mass was measured precisely,
with a negligible error, with an analytic scale that had a sensitivity of 0.1 mg.
Examining Table 1 we notice that the recipe provides the sample concentration
with a measured error less than 10%, which is a reasonable value considering that
the technique is intended to be, first of all, rapid.
Table 1
Yeast concentration, the measured contrast, the assessed concentration, and the relative error
c, g/l
K
0.35
2.41
5.23
106
73
53.4
assessed c,
g/l
0.315672
2.22415
5.483688
error, %
9.8
7.3
–5.5
8
Estimating wine yeast concentration by far field contrast measurement
353
6. CONCLUSION
The work described in this article states that a simple physical procedure
using the far interference field of a coherent light beam incident on a suspension
can be used to assess the yeast concentration in aqueous suspension. The far
interference field is recorded using a CCD and the recording is processed later on
using a program written for this purpose that uses an algorithm previously tested
[13, 14]. The average speckle size variation with the concentration was found to be
monotone (decreasing with the concentration) over an extended concentration
range, from 8.6·10-2 to 10 g/l, which is the range that is currently found during vine
fermentation. This behavior suggests a simple yet very fast procedure for assessing
the yeast concentration from the measured value of the contrast K. This procedure
lasts around three to four minutes, including sample extraction from the reactor,
recording the far field, computing the contrast K and finding the concentration
using Eq. (2).
As time passes and fermentation carries on, an initial yeast concentration
increase occurs and a maximum is reached. The fermentation process intensity
decreases afterwards and the yeast concentration decreases as yeast dyes.
Monitoring the yeast concentration enables thus monitoring the fermentation
process and this might be useful in vine technology.
REFERENCES
1. C.P. Kurtzman, J.W. Fell, Yeast Systematics and Phylogeny–Implications of Molecular
Identification Methods for Studies in Ecology, Biodiversity and Ecophysiology of Yeasts, The
Yeast Handbook, Edited by C. Rosa & G. Péter, Springer-Verlag, 11–30 (2006).
2. N. J. W. Kreger-van Rij, General classification of the yeasts. In The yeasts: a Taxonomic Study.
3rd edition. Edited by N. J. W. Kreger-van Rij, Elsevier, 1–44, Amsterdam (1984).
3. N. J. W. Kreger-van Rij, M. Veenhuis, A comparative study of the cell wall structure of
basidiomycetous and related yeasts, Journal of General Microbiology 68, 87–95 (1971).
4. H. K. Kwon, S. H. Woo, & J. M. Park, Degradation of tetracyanonickelate (II) by Cryptococcus
humicolus MCN2, FEMS Microbiology Letters 214, 211–216 (2002).
5. E. J. Lodolo, J.L.F. Kock, B.C. Axcell, M. Brooks, The yeast Saccharomyces cerevisiae: the main
character in beer brewing. FEMS Yeast Research 8, 1018–1036 (2008).
6. B. C. Viljoen, A. M. Knox, P. H. de Jager, A. Lourens-Hattingh, Development of yeast populations
during processing and ripening of blue veined cheese. Food Technology and Biotechnology,
41, 291–297 (2003).
7. Y. Piederrière, J. Cariou, Y. Guern, B. Le Jeune, G. Le Brun, J. Lotrian, Scattering through fluids:
speckle size measurement and Monte Carlo simulations close to and into the multiple
scattering, Optics Express 12, 176–188 (2004).
8. Y. Piederriere, J. Le Meur, J. Cariou, J.F. Abgrall, M.T. Blouch, Particle aggregation monitoring by
speckle size measurement; application to blood platelets aggregation, Optics Express 12,
4596–4601 (2004).
354
Dan Chicea et al.
9
9. Y. Piederriere, F..Boulvert, J. Cariou, B. Le Jeune, Y. Guern, G. Le Brun, Backscattered speckle
size as a function of polarization: influence of particle-size and concentration, Optics Express
13, 5030–5039 (2005).
10. J.W. Goodman, Statistical Properties of Laser Speckle Patterns, in Laser speckle and related
phenomena, Vol.9 in series Topics in Applied Physics, J.C. Dainty, Ed., Springer-Verlag,
Berlin, Heidelberg, New York, Tokyo (1984).
11. J.D. Briers, G. Richards and X.W. He., Capillary blood flow monitoring using laser speckle
contrast analysis (LASCA), J. Biomed. Opt. 4, 164–175 (1999).
12. R. Nothdurft, G. Yao, Imaging obscured subsurface inhomogeneity using laser speckle, Optics
Express, 13(25), 10034–10039 (2005).
13. D. Chicea, Biospeckle Size and Contrast Measurement Application in Particle Sizing and
Concentration Assessment, Romanian Journal of Physics, 52, 581–587 (2007).
14. D. Chicea, Speckle Size, Intensity and Contrast Measurement Application in Micron-Size Particle
Concentration Assessment, European Physical Journal Applied Physics, 40, 305–310 (2007),
doi: 10.1051/epjap:2007163.