Monitoring of the fermentation media of citric acid by the trimethylsilyl derivatives of the organic acids formed

Article abstract of DOI:10.1021/jf030751v

In this approach, a derivatization method is described for monitoring of organic acids in fermentation media without any separation step. The aqueous phase of fermentation media was evaporated and heated in a silylation reagent to form trimethylsilyl (TMS) derivatives. The silylated compounds are analyzed by ²⁹Si nuclear magnetic resonance (²⁹Si NMR) and gas chromatography-mass spectrometry (GC-MS). ²⁹Si NMR can qualitatively monitor the components produced in the Krebs cycle. Quantification of these compounds is investigated by using selected ion monitoring mode of mass spectrometry. In this mode, mass to charge (m/z) values of their [M - 15] ⁺ ions, which are 465, 275, 247, 221, 335, 251, and 313 of TMS derivatives of citric, α-ketoglutaric, succinic, fumaric, L-malic, oxaloacetic, and palmitic (as an internal standard), acids, respectively, are used. The limit of detection and the linear working range for derivatized citric acid were found to be 0.1 mg L^-1 and 10-3 × 10⁴ mg L^-1. The relative standard deviation of the method for five replicates was 2.1%. The average recovery efficiency for citric acid added to culture media was ～97.2%. Quantitative results of GC-MS are compared with those obtained by an ultraviolet-visible method.

Full text of DOI:10.1021/jf030751v

6384 J. Agric. Food Chem. 2004, 52, 6384−6388
Monitoring of the Fermentation Media of Citric Acid by the
Trimethylsilyl Derivatives of the Organic Acids Formed
ALIREZA GHASSEMPOUR,*,† SAEED NOJAVAN,§ ZAHRA TALEBPOUR,#
§
ALI ASGHAR AMIRI,§ AND NAHID MASHKOURI NAJAFI
Medicinal Plants Research Institute, Shahid Beheshti University, P.O. Box 19835-389, Tehran, Iran;
Chemistry Research Center, Faculty of Science, Shahid Beheshti University, Tehran, Iran; and
Department of Chemistry, Faculty of Science, Azzahra University, Tehran, Iran
In this approach, a derivatization method is described for monitoring of organic acids in fermentation
media without any separation step. The aqueous phase of fermentation media was evaporated and
heated in a silylation reagent to form trimethylsilyl (TMS) derivatives. The silylated compounds are
analyzed by 29Si nuclear magnetic resonance (29Si NMR) and gas chromatography-mass spec-
trometry (GC-MS). 29Si NMR can qualitatively monitor the components produced in the Krebs cycle.
Quantification of these compounds is investigated by using selected ion monitoring mode of mass
spectrometry. In this mode, mass to charge (m/z) values of their [M - 15]+ ions, which are 465, 275,
247, 221, 335, 251, and 313 of TMS derivatives of citric, R-ketoglutaric, succinic, fumaric, L-malic,
oxaloacetic, and palmitic (as an internal standard), acids, respectively, are used. The limit of detection
and the linear working range for derivatized citric acid were found to be 0.1 mg L-1 and 10-3 × 104
mg L-1. The relative standard deviation of the method for five replicates was 2.1%. The average
recovery efficiency for citric acid added to culture media was ∼97.2%. Quantitative results of GC-MS
are compared with those obtained by an ultraviolet-visible method.
KEYWORDS: Citric acid; trimethylsilyl derivatives; gas chromatography-mass spectrometry; 29Si NMR;
Aspergillus niger
INTRODUCTION
area of work involving analysis of organic compounds, par-
ticularly in work involving biological samples. Silyl derivatives
are also formed by replacement of active hydrogen on OH, SH,
and NH groups. Silylation of organic acids not only reduces
the boiling points for GC analysis but also produces high
molecular weight derivatives suitable for detection in mass
spectrometry. Also, these derivatized compounds can be moni-
tored by 29Si nuclear magnetic resonance (29Si NMR).
In a previous work, we determined citric acid in fermentation
media by a pyrolysis-mass spectrometry (Pyr-MS) method (25).
In the present work, 29Si NMR and GC-MS techniques have
been employed as simple and sensitive methods to monitor the
citric acid and other Krebs cycle compounds produced in
fermentation media. These methods require preparation of
trimethylsilyl (TMS) derivatives of these compounds. 29Si NMR
and GC-MS peaks of TMS derivatives of citric, R-ketoglutaric,
succinic, fumaric, L-malic, and oxaloacetic acids appear by their
production and are eliminated by reducing in fermentation
media. Also, TMS derivatives provide information on the
chemical structure of compounds with a number of fragment
ions, including [M - 15]+ ions specified for each metabolite
for quantitative works by GC-MS method.
Citric acid, a tricarboxylic acid intermediate in the Krebs
cycle, is widely used in food and pharmaceutical industries.
Because of its commercial and academic importance, the
biosynthesis of citric acid by submerged fermentation using the
filamentous fungus, Aspergillus niger, has been the subject of
numerous investigations (1-4). On the other hand, the monitor-
ing and controlling of the fermentation process, when a new
species of A. niger was used, are very important. In fact,
improvements in process control can be expected by complete
identification of components by increasing the reproducibility
and accuracy of measurements (5).
A large number of methods for the determination of citric
and other organic acids from various sources have been reported.
These include gravimetry analysis (6, 7), potentiometry and
biosensors (8, 9), high-performance liquid chromatography
(HPLC) (10-14), enzymatic flow injection (15-18), chemi-
luminescence methods (19, 20), capillary electrophoresis (21,
22), and spectrophotometric methods (23, 24). Analytical
derivatization reactions may find application in virtually any
* Corresponding author [telephone/fax 98 (21) 2418679; e-mail
a-ghasempour@cc.sbu.ac.ir].
MATERIALS AND METHODS
† Medicinal Plants Research Institute, Shahid Beheshti University.
§ Chemistry Research Center, Shahid Beheshti University.
# Azzahra University.
Instrumentation. All nuclear magnetic resonance measurements
were carried out on a Bruker DRX 500 Avance spectrometer with field
10.1021/jf030751v CCC: $27.50
© 2004 American Chemical Society
Published on Web 09/28/2004

Citric Acid Fermentation
J. Agric. Food Chem., Vol. 52, No. 21, 2004 6385
Table 1. Characteristic 29Si NMR Peaks of TMS Derivative Organic
Acids Produced in the Krebs Cycle
strength of 11.744 T. At this field, Si-29 resonates at 99.352 MHz.
Tetramethylsilane (TMS) was used as an external reference. All
chemical shift measurements were carried out at a probe temperature
of 25.0 ( 0.1 °C; typical acquisition parameters were 400 ppm sweep
width, 32 scans, 50 s relaxation delay, 0.83 s acquisition time, and 10
ms pulse width (90° pulse) with proton decoupling.
A Shimadzu GC-MS model QP-1100 EX was used. The gas
chromatograph was a Shimadzu 14-A equipped with a fused silica
capillary column CBP-5 (25 m × 0.22 mm i.d. with a film thickness
of 0.25 µm) as the stationary phase. Helium was used as the carrier
gas, at a constant flow rate of 1 mL min-1. The temperature of the
injection was 220 °C, and the oven was programmed from 150 to 220
°C at a rate of 10 °C min-1; the final temperature was held for 5 min.
TMS derivatives were analyzed in the EI mode using an ion source
temperature at 250 °C and pressure of 5 × 10-6 Torr. Ionization was
performed at 70 eV.
acid
oxalo-
acetic
R-keto-
glutaric
citric
12.4
fumaric
25.7
L
-malic
succinic
23.0
29Si NMR position
(ppm)
19.2
27.1
24.1
was transferred to the molasses solution (above flask), incubated at 30
°C, and agitated at 120 rpm until the fermentation process was
completed.
Preparation of Fermentation Samples for Analysis. In spectro-
photometric methods, owing to its small linear working range (1-300
mg L-1), dilution of samples is required. Thus, 1.0 mL of the aqueous
phase of the fermentation medium was accurately dispensed into 50-
mL calibrated flasks. Twenty milliliters of TCA (30% w/v) was added
to each flask and then diluted to the mark with molasses solution (20%
v/v), shaken, and left to stand for 30 min. The reason for adding TCA
is to remove plant proteins present in sample and to obtain more
reproducible results. Finally, 1 mL of this solution is analyzed by UV-
vis spectroscopy.
To perform the analysis at different times from the fermentation
process, exactly 5 mL of the aqueous phase of the fermentation medium
was transferred to test tubes and evaporated. Because silylation reagent
and TMS derivatives both are hydrolytically unstable, they must be
protected from moisture. Then they were supplemented with an internal
standard (for GC-MS), silylated, and analyzed as described above. For
the 29Si NMR analysis, 300 µL of the sample was transferred to an
NMR tube and 200 µL of deuterated DMSO added.
A Shimadzu UV-2100 spectrophotometer was employed for mea-
surement of absorption of the citric acid chromophore at 428 nm. Two
water baths were used, one thermostatically controlled at 32 °C and
the other containing an ice-water slurry.
Reagents and Chemicals. Citric acid anhydride, R-ketoglutaric,
succinic, fumaric, L-malic, oxaloacetic, and palmitic acids, hexameth-
yldisilasane (HMDS), trimethylchlorosilane (TMCS), and tetrahydro-
furan (THF) were purchased from Merck (Darmstadt, Germany).
Pyridine, acetic anhydride, potassium ferrocyanide, and trichloroacetic
acid (TCA) were obtained from Fluka (Buchs, Switzerland). Beet
molasses was obtained from Shazand Sugar Factory (Shazand, Iran).
Aspergillus niger PTCC 5010 (Persian Type Culture Collection) was
used throughout the experiment. The strain was maintained on potato
dextrose agar (PDA) slants at 4 °C and subcultured at intervals from 2
to 3 months.
Standard Solutions. A stock solution of citric acid (50 mg mL-1
)
was prepared by dissolving exactly 2.5000 g of citric acid anhydride
in 50 mL of THF solvent. Also, a stock solution containing 2.5 mg
mL-1 of palmitic acid was prepared by dissolving exactly 0.1250 g of
palmitic acid in 50 mL of THF solvent. To prepare standard working
solutions at six different concentrations in the range of 10-50000 mg
L-1 with respect to citric acid, appropriate volumes of the citric acid
stock solutions were accurately dispensed into 25-mL calibrated flasks.
Then 1 mL of palmitic acid stock solution, as an internal standard,
was added to each flask and finally diluted to the mark with THF as
solvent. Thus, concentration ratios of 0.1, 1, 10, 100, 300, and 500 of
citric acid to palmitic acid were obtained. For the UV-vis spectroscopy
method, other standard working solutions at six different concentrations
in the range of 1-300 mg L-1 with respect to citric acid were prepared
with accurately dispensed appropriate volumes of the citric acid stock
solution, into 50-mL calibrated flask. Then 20 mL of TCA was added
to each flask and finally diluted to the mark with a 20% molasses
solution.
RESULTS AND DISCUSSION
The medium of fermentation is a mixture of molasses, fungus,
and citric and other organic acids in which they are produced
and reduced based on the Krebs cycle during the fermentation
process. Citric acid has four active hydrogens as part of its
structure, and they are replaced with the trimethysilyl groups
according to the reaction
Derivatization. To perform silylation, exactly 5 mL of above
standard solutions dispensed into test tubes and appropriate amounts
of HMDS and TMCS (as catalyst) were added to each tube. Then the
resultant solutions were heated at 70 °C for 4 h and the tubes cooled
to room temperature before measurement by 29Si NMR and GC-MS.
Derivatization for Spectrophotometric Method. To perform this
experiment, 1.0 mL of related standard working solutions was pipetted
into labeled tubes that were then placed in an ice-water bath. Then
1.7 mL of pyridine was added to each tube, and subsequently 5.8 mL
of acetic anhydride was immediately added to tubes. All tubes were
placed in a 32 °C water bath for 30 min and then at room temperature
for 20 min. The absorbance of the resultant chromophore was obtained
at 428 nm. The calibration curve was constructed by plotting the
absorbance against the concentration.
Fermentation Condition. Beet molasses were diluted with distilled
water to give a concentration of 20% (w/v). A conical flask (500 mL)
containing 100 mL of beet molasses, pH 7, was autoclaved at 121 °C
and a pressure of 15 psi for 15 min. The medium was treated, while
hot, with potassium ferrocyanide (10% w/v) to precipitate the heavy
metals and then cooled to 30 °C. Spores of 5-day-old culture of A.
niger grown on PDA were harvested and suspended in distilled water
to obtain 1 × 108 spores mL-1. Five milliliters of spore suspension
The TMS-derivatized citric acid has a good volatility and
thermal stability, making it amenable to GC-MS determinations
as well as 29Si NMR spectroscopy.
TMS derivatives of citric, R-ketoglutaric, succinic, fumaric,
L-malic, and oxaloacetic acids have characteristic 29Si peaks as
shown in Table 1.
The characteristic 29Si NMR peaks may be used for monitor-
ing a fermentation medium during the process. Figure 1 shows
the 29Si NMR spectra of TMS-derivatized compounds in the
fermentation media at the 5th, 7th, and 10th days. On the 5th
day of the fermentation processes, only the 29Si NMR peaks of
TMS-citric acid at 12.4, 23.4, and 25.8 ppm are seen. On the
7th day, the TMS-R-ketoglutaric acid peak appears at 24.1 ppm
and the TMS-succinic acid peak at 23.0 ppm. On the 10th day
the TMS-R-ketoglutaric and TMS-succinic acids peaks disappear

6386 J. Agric. Food Chem., Vol. 52, No. 21, 2004
Ghassempour et al.
Figure 3. Full MS spectrum of the TMS derivative palmitic acid.
Figure 1. 29Si NMR spectra of TMS derivative components on the (a)
5th, (b) 7th, and (c) 10th days of the fermentation processes.
Figure 2. Full MS spectrum of the TMS derivative citric acid.
Figure 4. UV spectra of standard solutions of citric acid at different
concentrations: (a) 50, (b) 100, (c) 150, (d) 200, (e) 250, and (f) 300 mg
mL of citric acid.
-
1
and TMS-L-malic and fumaric acids peaks appear at 19.2 and
25.7 ppm, respectively.
Table 2. Comparison of the GC-MS and UV Results
When GC analysis of the fermentation media is difficult or
impossible due to the nonvolatile components in the fermenta-
tion media, the TMS derivatives of nonvolatile compounds can
solve this problem. When TMS-citric acid is introduced to the
ionization source, the fragmentation gives a pattern of peaks at
m/z 147, 273, 347, and 465 as shown in Figure 2, which are
used for qualitative and quantitative analysis.
characteristic
GC-MS method
10
104
2.1
0.1
UV
-
1
linearity (mg L
)
−3
×
1
−
300
1.94
13.2
RSDa
-
1
LODb (mg L
)
performance time
∼
5 h
0.9967
97.2
∼
90 min
0.9908
95.2
c
R 2
recovery (%)
As shown in the results, m/z 465 belongs to [M - 15]+, which
is a fragment ion of TMS-derivatized citric acid that has high
abundance. This ion has good sensitivity and abundance, thus
monitoring and quantification of citric acid can be performed
using this fragment ion. The specificity of the method also
increases because this ion appears at a high m/z value (the
spectrum is very simple at m/z >100).
a Relative standard deviation. b Limit of detection. c Correlation coefficient square.
produce the corresponding TMS-derivatized products. The m/z
[M - 15]+ of these organic acids could not interfere with each
other at all and can be determined for quantitative works in the
fermentation media.
We noticed that selection of an internal standard (in the
present work, TMS-derivatized palmitic acid) could recalibrate
the system continuously and overcome any drift problems of
the GC-MS data. Palmitic acid has an active carboxyl group
and is easily converted to a trimethylsilyl derivative. Figure 3
shows that when TMS-palmitic acid is introduced in the
ionization source, it is subjected to fragmentation and an ion
with m/z 313 is formed, which is assigned to the [M - 15]+.
The palmitic acid derivative is chosen as internal standard for
several reasons. First, this compound after silylation has an m/z
of 313, which does not overlap with any of the peaks of the
compounds. Second, it does not react with the citric acid
derivative and other compounds in the fermentation medium.
It has high thermal stability and volatility and also shows a good
abundance and proved to be reproducible in m/z 313.
As mentioned before, UV-vis spectroscopy is a routine
method for the measurement of citric acid, suffering from
drawbacks (23, 24). The UV-vis spectra of citric acid are shown
in Figure 4.
However, this method was used as a reference method to
compare with the performance of our GC-MS approach. Results
of this comparison are presented in Table 2.
The UV method is nonspecific and tedious for the analysis
of citric acid in a complex matrix such as fermentation media.
The requirement of large amounts of expensive, toxic solvents
(pyridine and acetic anhydride), which can be harmful to the
environment and have narrow linear working ranges, are some
of the drawbacks of this method. The results in Table 2 show
that the LOD, LDR, RSD, and recovery efficiency of the GC-
MS method are better than those of the UV method. Also, each
of the acids formed in the fermentation media has a distinctive
fragmentation pattern that allows easy identification. In addition,
without utilizing any separation steps for organic acids from
According to the Krebs cycle, A. niger fungus may produce
other organic acids such as fumaric, malic, succinic, R-keto-
glutaric, and oxaloacetic acids as well as citric acid. These
byproducts are able to react with the silylation reagent and

Citric Acid Fermentation
J. Agric. Food Chem., Vol. 52, No. 21, 2004 6387
molasses based medium by a strain of Aspergillus niger. World
J. Microbiol. Biotechnol. 1999, 1 (5), 647-649.
(2) Leangon, S.; Maddox, I. S.; Brooks, J. D. A proposed biochemi-
cal mechanism for citric acid accumulation by Aspergillus niger
Yang NO.2 growing in solid-state fermentation. World J.
Microbiol. Biotechnol. 2000, 16, 271-274.
(3) Roukas, T.; Indust, J. Citric and gluconic acid production from
fig by Aspergillus niger using solid-state fermentation. J. Ind.
Microbiol. Biotechnol. 2000, 25, 298-304.
(4) Sankpal, N. V.; Joshi, A. P.; Kulkarni, B. D. Citric acid
production by Aspergillus niger immibolized on cellose mi-
crofibrils: influence of morphology and fermentater conditions
on productivity. Process Biochem. 2001, 36, 1129-1139.
(5) Pungor, E.; Pecs, M.; Szigeti, L.; Nyeste, L. Mass spectrometric
monitoring of 2-oxoglutaric acid in fermentation broth. Anal.
Chim. Acta 1984, 163, 185-191.
(6) Bhosal, S. N.; Khopkar, S. M. Reversed phase extraction
chromatography of germanium(IV) with tributyl phosphate on
silica gel. Talanta 1985, 32, 155-159.
(7) Grases, F.; Bauza, A. C.; March, J. G. Determination of citric
acid based on inhibition of the crystal growth of calcium fluoride.
Analyst 1991, 132, 59-63.
(8) Othman, A. O.; Sweileh, J. A. Phosphat rock treatment with citric
acid for the rapid potentiometric determination of fluoride with
ion selective electrode. Talanta 2000, 51, 993-999.
(9) Pereira, A. C.; Fertonani, F. L.; Neto, G. O.; Kubota, L. T.;
Yamanaka, H. Regentless biosensor for isocitrate using one step
modified Pt-Ir microelectrode. Talanta 2001, 53, 801-806.
(10) Rodriguez, M. A. M.; Oderiz, M. L. V.; Hernandez, J. L.; Lozano,
J. S. Determination of vitamin C and organic acids in various
fruits by HPLC. J. Chromatogr. Sci. 1992, 30, 433-437.
(11) Walker, T. A. A liquid chromatographic assay for citric acid in
over the counter carbamide peroxide products using indirect UV
detection. J. Pharm. Biomed. Anal. 1995, 13, 171-176.
(12) Eiteman, M. A.; Chastain, M. J. Optimization of the ion exchange
analysis of organic acids from fermentaton. Anal. Chim. Acta
1997, 338, 69-75.
Figure 5. SIM chromatograms of the fermentation medium at the (a) 7th
and (b) 8th days of the fermentation process.
Table 3. Recovery Test for the GC-MS Method
sample
citric acid added (g)
determined (g)
recovery %
1
2
3
0.100
0.050
0.010
0.098
0.0488
0.0096
98
97.6
96.1
(13) Miwa, H. High performance liquid chromatographic determi-
nation of mono-, poly- and hydroxycarboxylic acids in foods
and beverages as their 2-nitrophenylhydrazides. J. Chromatogr.
A 2000, 881, 365-386.
(14) Cunha, S. C.; Ferreria, J. M.; Fernandes, J. O.; Faria, M. A.;
Beatriz, M.; Oliveria, P. P.; Ferreia, M. A. Determination of
lactic, acetic, succinic, and citric acids in table olives by HPLC/
UV. J. Liq. Chromatogr. 2001, 24, 1029-1038.
(15) Planta´, M.; La´zaro, F.; Puchades, R.; Maquieira, A. Determi-
nation of citric acid and oxaloacetic acid in foods by enzymatic
flow injection. Analyst 1993, 188, 1193-1197.
(16) Gajovic, N.; Warsinke, A.; Scheller, F. W. A novel multienzyme
electrode for the determination of citrate. J. Chem. Technol. Biot.
1995, 63, 337-344.
(17) Matsumoto, K.; Tsukatani, T. Simultaneous quantitation of citrate
and isocitrate in citrus juice by a flow injection method based
on the use of enzyme reactors. Anal. Chim. Acta 1996, 321, 157-
164.
(18) Ine´sMato, J. F. H.; Cendo´n, V.; Soleded, M. M.; Ferna´ndez, M.
A.; Sancho, M. T. Enzymatic determination of citric acid in
honey by using polyvinyl polypyrrolidone clarification. J. Agric.
Food Chem. 1998, 46, 141-144.
(19) Dunemann, L. Automated determination of carboxylic acids in
biological and geochemical samples by means of chemical
reaction detector. Anal. Chim. Acta 1989, 221, 19-26.
(20) He, Z.; Gao, H.; Yuan, L.; Lu, S.; Meng, H.; Li, X.; Zeng, Y.
Pulse injection analysis with chemiluminescence detection:
determination of citric acid using tris-(2,2′-bipyridine) ruthenium-
(II). Talanta 1998, 47, 301-304.
(21) Smith, A. D.; Moini, M. Control of electrochemical reactions at
the capillary electrophoresis outlet/electrospray emitter electrode
under CE/ESI-MS through the application of redox buffers. Anal.
Chem. 2001, 73, 240-246.
fermentation media, the GC-MS method with SIM mode, using
m/z 465 related to citric acid, allows a selective monitoring and
quantifying. The equation of the calibration curve, of the
fermentation medium, is y ) 0.6993x - 0.0079, with R2 )
0.9967. As an example, the SIM peak of citric acid and palmitic
acid at two different fermentation times are shown in Figure
5.
On the basis of the calibration data used to compute the
concentration of citric acid, the ratios of the peak areas
representing the TMS-derivatized citric acid (m/z 465) to those
of the internal standard (m/z 313) were calculated.
Precision and Recovery. The precision of the method was
determined using standard solutions. After derivatization,
samples were analyzed by five replications by GC-MS. Mea-
surement of the response ratios of the analytes to the internal
standard was repeatable with ∼2.1% RSD.
To determine the method recovery for citric acid, known
amounts of citric acid and known amounts of the internal
standard (palmitic acid) were added to a 20% molasses solution.
After derivatization, ∼2 µL of the samples was injected in the
GC-MS in five replicates and analyzed under the mentioned
conditions. The results are shown in Table 3. The value of
recovery is always >96%.
LITERATURE CITED
(1) Pera, L. M.; Callieri, D. A. Influence of calcium on fungal growth
and citric acid production during fermentation of a sugarcane

6388 J. Agric. Food Chem., Vol. 52, No. 21, 2004
Ghassempour et al.
(22) Saavedra, L.; Ruperez, F. J.; Barabas, C. Capillary electrophoresis
for evaluating orange juice authenticity: a study on Spanish
oranges. J. Agric. Food Chem. 2001, 49, 9-13.
(23) Indyk, H. E.; Kurmann, A. Routine spectrophotometric deter-
mination of citric-acid in milk powders. Analyst 1987, 112,
1173-1175.
(24) Luque, P. E.; Rios, A.; Valca´rcel, M. Flow-injection spectro-
photometric determination of citric-acid in beverages based on
a photochemical reaction. Anal. Chim. Acta 1998, 366, 231.
(25) Ghassempour, A.; Mashkouri, N. N.; Amiri, A. A. Determination
of citric acid in fermentation media by pyrolysis mass spec-
trometry. J. Anal. Appl. Pyrol. 2003, 70, 251-261.
Received for review November 7, 2003. Revised manuscript received
July 21, 2004. Accepted July 27, 2004.
JF030751V