Article

pubs.acs.org/JAFC

Quantitation of trans-Aconitic Acid in Different Stages of the

Sugar-Manufacturing Process
Guillermo Montoya,*,† July Londono,† Paola Cortes,† and Olga Izquierdo§
†
Facultad de Ciencias Naturales, Grupo Natura, Universidad ICESI, Cali, Colombia
§
Manuelita Sugarmill Km 7 vı ́a Palmira-Cerrito, Palmira, Colombia

ABSTRACT: The sugar cane industry has seen how biomass production in sugar mills would be converted to a readily available
source of molecules besides sugar. Properly managed, byproducts would be transformed into a sustainable source of renewable
and environmentally friendly chemical products.1 As a principal and more abundant organic acid in sugar cane juice, trans-
aconitic acid (TAA) has been studied for use as a plasticizer in the polymer industry.2 However, up to now no industrial-scale
application has been reported. As a reasonable approach to recover TAA from a sugar mill, first, an analytical method to
determine its presence in all stages of the sugar-manufacturing process is needed. A new modern method was developed to
measure TAA in seven stages in a sugar mill located in Valle del Cauca, Colombia. The stages with higher content of TAA were
syrup, with 3363.6 ± 589.3 mg/L, and honey (molasses), with 6110.05 ± 139.5 mg/L.
KEYWORDS: sugar cane, trans-aconitic acid, solid phase extraction, UPLC-ESLD, byproducts, sugar mill

■ INTRODUCTION
Sugar from sugar cane is produced in around 120 countries
worldwide, and Brazil is the biggest producer with an average
annual production of 34 million tons per year.3 Additional
byproducts obtained in the Brazilian sugar cane industry are the
bioelectricity commonly used by manufacturing plants and
biopolymers (as raw materials for the production of packaging,
bags, and bottles obtained from sugar, ethanol, and bagasse).4,5
Whereas the use of byproducts in some countries is well
established, others have not determined how to take advantage
of these residuals. Exploratory work has been carried out to
recover byproducts from the production process, and trans-
aconitic acid is no exception; its industrial retrieval has been
suggested by means of easy scaled-up techniques.6 There
is no consistent flowchart of a specific pathway to discover a
recoverable molecule from industrial waste. However, the first
step of a reasonable approach is an analytical evaluation of the
target into the matrix where it is contained.
Colombia processes >20 million tons of sugar cane per year,
obtaining around 2.4 million tons of sugar;7 its residues are
mainly employed for bioelectricity, in the paper industry, and
to fertilize the sugar cane crops. Sugar cane is also one of the
country’s most widespread crops and therefore has an
important impact in gross domestic product. Conversely, the
economic progress of this agro-industrial business during the
course of the lpast few decades has been stagnant; Colombia
does not appear in the top 10 of the world’s biggest producers.
Another well-known issue in the sugar cane industry is the
unpredictable volatile nature of the price of raw sugar,8 which
supports the idea that taking advantage of byproducts in sugar
mills would be a rational alternative to increase the industry’s
economic value.
Aconitic Acid and Special Considerations. Aconitic acid
is a tricarboxylic acid and the most abundant acid produced by
sugar cane. The predominant form is the trans-aconitic acid
isomer (TAA), which in Australian molasses averages 19.4 g/kg
of the total dry solids9 and in North American molasses
averages 9−55 g/kg in dry solid bases. Several techniques have
been used to try to recover TAA from molasses; two of the
most popular are liquid−liquid extraction10 and supported
liquid membranes.9 However, bioethanol plants operate with
discarded molasses from sugar mills, and the scientific com-
munity is making enormous efforts to use bagasse as a source of
second-generation alcohols.11−13 Scientists are also trying to
find the best matrix in the sugar-manufacturing stages to
establish an industrial TAA extraction for the food, cosmetics,
and chemical industries.
With regard to analytical methods, TAA is an organic acid
with a low molar absorptivity coefficient and high hydro-
solubility (Log P = −1); therefore, it is inadequately retained in
hydrophobic chromatographic columns. The oldest method-
ologies described the use of octadecylsilane columns with an
ion pair reagent; however, ion pair reagents make the column
spoil quickly, so this practice is no longer used. Another option
was to use amino columns under low-pH normal phase con-
dition, which become weak anion exchangers capable of
separating negatively charged molecules. Recently, polar
embedded groups in reverse phase columns are gaining more
attention for their rising hydrophilic interaction, capable of
interacting with polar compounds such as carbohydrates and
small organic acids.14 Amide columns also offer a wider set of
possibilities to separate very highly hydrophilic molecules.15
This hydrophilic interaction chromatography is available in sub-
2 μm particles, offering acceptable separation of molecules with
physicochemical properties such as those described above, in
short retention times and good peak shapes.
Received: February 24, 2014
Revised: August 1, 2014
Accepted: August 6, 2014

© XXXX American Chemical Society A dx.doi.org/10.1021/jf5008874 | J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Journal of Agricultural and Food Chemistry Article

After good separation is achieved, properly detecting small

organic aliphatic acids using chromatography is one of the
biggest concerns. Generally, academics describe mass spec-
trometry as the perfect choice to couple with a separation
technique; however, in most commercial industries, it is not
always chosen because of its high cost. The lack of light
absorption of organic acids and carbohydrates makes them
unsuitable for UV detection (at least with high sensitivity
and without derivatization), but other options are available to
detect them, such as refractive index and conductivity. Recently,
evaporative light-scattering detectors (ELSD) and charged
aerosol detectors (CAD) have undergone some improvements
that allow them to replace the refractive index in many
applications. ELSD has the advantage of having a response
independent of the solvent gradient, and despite the fact that
the ELSD response is not necessarily linear, the selection
of the appropriate concentration for calibration standards and
the best fitted equation allow the analyst to quantify a wide
range of analyte concentrations.
On the basis of these aspects, we decided to develop a
method to measure the quantities of TAA in different stages
of the sugar-manufacturing process. We suggest additional
matrices from which TAA could be recovered. It is important to
systematically identify the best stages in the sugar-manufactur-
ing process where this byproduct is wasted and, consequently,
promote its industrial recovery. The new methodology includes
a sample treatment using ion exchange cartridges in solid phase
extraction coupled with a faster LC analysis with an evaporative
light-scattering detector (SPE-UPLC-ELSD).

■ MATERIALS AND METHODS

Materials. The trans-aconitic and formic acids were purchased
from Sigma-Aldrich. Methanol and acetonitrile of HPLC grade were
obtained from Merck (Darmstadt, Germany). The sugar cane juices
were obtained from the analytical laboratory of the Manuelita sugar
cane mill (Km 7, Vı ́a Palmira, Valle del Cauca, Colombia). Ultrapure
water was obtained from a Sartorius ASTM ultrapure water system
(total organic carbon < 5 ppb, conductivity at 0.05 μS/cm3, and
filtered using 0.22 μm pore size).
Selection of Different Sugar-Manufacturing Process Stages.
In the pipelines of several sugar mills, the stage names would differ.
Nevertheless, the sequence of steps remains the same. Briefly, the
processing of sugar cane begins with crushing the stalk to obtain the
juice (extraction); then the bagasse is re-extracted with water to Figure 2. Comparison of chromatograms obtained with and without
contribute better sucrose extraction (dilution). Once extracted, the solid phase extraction: (top) sugar cane juice containing high levels of
juice is transferred to an alkalization tank where the pH is controlled impurities; (bottom) well sample subjected to the SPE procedure.
by adding lime (alkalization). After that, the juice is clarified, precipitating Both samples were diluted at equal volumes to manage the same
concentration of solids, suggesting TAA was concentrated after SPE
[by comparing light-scattering units (LSU)].

proteins, fats, and waxes to make a sludge (clarification). To make the

Figure 1. Flowchart of the sugar-manufacturing process. The arrows
follow the workflow, and the solid line boxes show the stages sampled.
Dashed boxes show the stages where TAA would be recovered.
process more efficient, the sludge is squeezed and filtered (filtration).
The clarified juice is then transferred to multieffect evaporators,
where almost 80% of the water content is eliminated to obtain the
syrup; then the process is repeated twice to obtain honey. Thus, the
stages of production are extraction, dilution, alkalization, clarification,
and filtration, producing syrup and two final products: refined white
sugar (ultrapure) and brown sugar. The juices were collected during
the last week of October 2013 during four consecutive days in
triplicate. A schematic flowchart of sugar cane manufacturing is shown
in Figure 1.
Sample Treatment by Solid Phase Extraction. Sugar cane
liquors and standards were injected onto a Strata Screen-A cartridge
(sorbent with C8 and anion exchange quaternary propylamine) con-
ditioned with methanol and water. The loading of samples and

B dx.doi.org/10.1021/jf5008874 | J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry
Figure 3. Stability test of TAA at 350 mg/L in different pH values over 80 min. Injections were made every 10 min. A Kruskal−Wallis test showed a
significant difference (p < 0.05) between 8.0 and 0.5 pH values. No significant differences were found between 8.0 value and either the 2.0 or 1.0 pH
value.
standards was performed on a solution at pH 7.0, and the washing
stage was carried out with 90:10 methanol/water preserving the same
pH value as the loading stage. The elution was performed with 20:80
methanol/water at pH 1.5 with diluted hydrochloric acid. All of the
pH values as well as organic/water proportions were obtained by
experimental design not shown. The stability of trans-aconitic acid at
various pH values is discussed below.
Equipment. The UPLC system consisted of a Waters Acquity H-
class equipped with a quaternary pump, degasser, and preheater
modules. Detection was performed with an ELSD using gas pressure at
40 psi, gain set at 25, the nebulizer on cooling mode, and drift tube
temperature set at 40 °C.
Chromatographic separation analysis was carried out on an amide
BEH (2.1 mm × 100 mm, 1.7 μm particle size) column (Waters,
USA). The mobile phase was a mixture of (A) deionized water/formic
acid (FA) (99.9:0.1% v/v) and (B) acetonitrile/formic acid (FA)
(99.9:0.1% v/v), and the flow rate was 0.5 mL/min in linear gradient
without temperature control of the column.
Statistical Analysis and Validation of the SPE-UPLC-ELSD
Method. All data processing was carried out with Empower 3.0.
Calibration curves were analyzed by quadratic equation (second-order
polynomial), and the coefficient correlations were always >0.9990
from a six-level curve ranging from 100 to 700 mg/L and relative
standard deviation (RSD) <5% by quadruplicate. To validate the
method, we used the logarithm of response (light-scattering units).16
Limits of detection (LOD) and limits of quantitation (LOQ) were
determined as 3 and 10 times the signal-to-noise ratio. LOD was 4
mg/L, and LOQ was 55 mg/L. The precision of intraday variation was
determined by analyzing five standard solutions in triplicate within
1 day by two analysts. The recovery was determined in water and
matrix (1 g of ultrapure sugar in 10 mL of pure water) at two
concentrations, 200 and 500 mg/L. The results showed that the
recovery of both water and matrix-spiked standard solutions ranged
from 90 to 105% with RSDs <7%. Statistical analyses were performed
by using GraphPad Prism 5.0 for Windows. A Kruskal−Wallis test was
used to assess the significance of the differences of TAA concentration
among pH values. All data were grouped by pH and compared against
pH 7.0 (control group). p < 0.05 was considered statistically significant.
■
RESULTS AND DISCUSSION
Recovery of Aconitic Acid by Ion Exchange SPE. We
needed to develop analytical methods that were as simple
as possible; sugar cane juice represents a complex matrix that
contains fat, waxes, organic acids, polyphenols, and high
C dx.doi.org/10.1021/jf5008874 | J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
concentrations of sugars. The average concentration of organic
acids is estimated in the range of 0.5−2%. Taking this into
consideration, we used a sample treatment method that would
extend column lifetime. Figure 2 shows the differences between
raw sugar cane juice and juice after SPE ion exchange pre-
paration on separate UPLC runs. Despite the natural con-
centration of TAA being high enough to be quantified, we
found that SPE not only concentrates the analytes but also
prevents impurities such as waxes and long-chain alcohols from
entering the column.
Stability of TAA at pH Conditions. As low pH is needed
to elute the TAA from cartridges, we verified the stability of
TAA at different pH values. Figure 3 shows a stability test of
TAA at 350 mg/L under different pH values (from 0.5 to 7.0)
over 80 min. TAA has three pKa values at 2.8, 4.6, and 6.3, and
we selected pH 1.5 due to the observed stability in combination
with the fact that it is >1 unit lower than the pKa value.
Content of TAA in Different Sugar-Manufacturing
Process Stages. As shown in Figure 1, the main goal of a
sugar mill is to whiten sugar cane juice and then crystallize
sucrose. Most sugar mill engineers suggested that we evaluate
primarily the sludge produced by alkalization tanks, because
this stage eliminates most of the acidic metabolites from the
juice, which precipitate under high pH value. These engineers
were wrong, because we proved that TAA remains in all stages
of the sugar-manufacturing process as shown in Figure 4. A
large amount of TAA was found in syrup and honey, obeying
the phenomenon of increasing concentration as water content
evaporates and also because the sucrose is extracted. That is the
same reason why honey TAA content is almost twice that of
syrup, as shown in Figure 4. Ultrapure sugar lacks TAA, but
brown sugar maintains around 158 mg/L TAA. The final
molasses that does not crystallize into additional sugar is then
used in bioethanol plants, making it off-limits for TAA recovery,
unless an intermediate stage is proposed that depletes TAA
from honey and preserves the sugar content needed in a bio-
refinery. One feasible matrix that contains TAA is vinasses
(resulting after alcohol distillation of molasses); however,
the matrix becomes more complex due to the high contents
of tannins and colored compounds. TAA is most likely to be
Journal of Agricultural and Food Chemistry Article

Figure 4. Quantification of trans-aconitic acid in seven stages of the sugar-manufacturing process. Samples were taken in triplicate over four
consecutive days. Bars are plotted with a standard deviation bar between each day. The concentration of TAA decreased from alkalized to filtered.
Brown sugar contains 158.45 ± 6.7 mg/L TAA, whereas refined white sugar (ultrapure sugar) does not have quantifiable amounts of TAA. Honey
contains 6110.05 ± 139.5 mg/L TAA. The rest of the TAA contents were 1006.4 ± 193.5 mg/L alkalized, 938.5 ± 152.9 mg/L clarified, 820.8 ±
122.2 mg/L diluted, 492.6 ± 127.5 mg/L filtered, 3363.6 ± 589.3 mg/L syrup, and 547.5 ± 136.6 mg/L extract, respectively.

recovered in the dashed-line boxes in Figure 1. Decades ago,

some scientists were dealing with the recovery of TAA from
molasses,9,17,18 but up to now, no commercial production has
been carried out.
The chromatographic conditions prove that hydrophilic
interaction by amide column represents a simple alternative to retain
highly hydrophilic compounds instead of using a complex mix of
solvents. Although light-scattering detectors have less sensitivity
compared to ultraviolet ones, it is the best simple alternative for
analytes with low absorptivity coefficients and simultaneously uses
solvents with additives well adapted to the chemistry of the column.
The honey (molasses) contains a high concentration of TAA, but a
method that does not affect sugar and alcohol production is
required to be feasible at a commercial level. Future research should
explore the economic viability of these processes.
■ REFERENCES
(1) Ekman, A.; Campos, M.; Lindahl, S.; Co, M.; Börjesson, P.;
Karlsson, E. N.; Turner, C. Bioresource utilisation by sustainable
technologies in new value-added biorefinery concepts − two case
studies from food and forest industry. J. Cleaner Prod. 2013, 57, 46−
58.
(2) Gil, N.; Saska, M.; Negulescu, I. Evaluation of the effects of
biobased plasticizers on the thermal and mechanical properties of
poly(vinyl chloride). J. Appl. Polym. Sci. 2006, 102, 1366−1373.
(3) Association, S. I. Sustainability Report 2011. In UNICA-Brazil,
2012.
(4) Harding, K. G.; Dennis, J. S.; von Blottnitz, H.; Harrison, S. T. L.
Environmental analysis of plastic production processes: comparing
petroleum-based polypropylene and polyethylene with biologically-
based poly-β-hydroxybutyric acid using life cycle analysis. J. Biotechnol.

■
2007, 130, 57−66.
(5) Agroindustrial, P. Biocycle; http://www.biocycle.com.br/.
AUTHOR INFORMATION (6) Kanitkar, A.; Aita, G.; Madsen, L. The recovery of polymerization
Corresponding Author grade aconitic acid from sugarcane molasses. J. Chem. Technol.
*(G.M.) Phone: (057)-2-5552334, ext 8856. E-mail: Biotechnol. 2013, 88, 4.
glmontoya@icesi.edu.co. (7) CENICAÑ . Informe Anual 2012. In Centro de Investigación de
la caña de azúcar en Colombia.
Funding (8) Sugar prices fall to lowest since July ’10 on Brazil crop. Bloomberg
All authors acknowledge ICESI University for all financial 2013 (May 14).
support and laboratory facilities. (9) McMurray, S. H.; Griffin, G. J. Extraction of aconitic acid from
Notes mixtures of organic acids and cane molasses solutions using supported
The authors declare no competing financial interest. liquid membranes. J. Chem. Technol. Biotechnol. 2002, 77, 1262−1268.

■
(10) Malmary, G. H.; Monteil, F.; Molinier, J. R.; Hanine, H.; Conte,
ACKNOWLEDGMENTS T.; Mourgues, J. Recovery of aconitic acid from simulated aqueous
effluents of the sugar-cane industry through liquid-liquid extraction.
We thank Manuelita Sugarmill (Km 7 vı ́a Palmira-Cerrito, Bioresour. Technol. 1995, 52, 33−36.
Colombia) for allowing us to sample and giving us access to a (11) Li, H. G.; Luo, W.; Gu, Q. Y.; Wang, Q.; Hu, W. J.; Yu, X. B.
quality control chemical laboratory. Acetone, butanol, and ethanol production from cane molasses using

D dx.doi.org/10.1021/jf5008874 | J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry Article

Clostridium beijerinckii mutant obtained by combined low-energy ion

beam implantation and N-methyl-N-nitro-N-nitrosoguanidine induc-
tion. Bioresour. Technol. 2013, 137, 254−260.
(12) de Andrade, R. R.; Maugeri Filho, F.; Maciel Filho, R.; da Costa,
A. C. Kinetics of ethanol production from sugarcane bagasse enzymatic
hydrolysate concentrated with molasses under cell recycle. Bioresour.
Technol. 2013, 130, 351−359.
(13) Rabelo, S. C.; Maciel Filho, R.; Costa, A. C. Lime pretreatment
and fermentation of enzymatically hydrolyzed sugarcane bagasse. Appl.
Biochem. Biotechnol. 2013, 169, 1696−1712.
(14) Gika, H. G.; Theodoridis, G. A.; Vrhovsek, U.; Mattivi, F.
Quantitative profiling of polar primary metabolites using hydrophilic
interaction ultrahigh performance liquid chromatography−tandem
mass spectrometry. J. Chromatogr., A 2012, 1259, 121−127.
(15) Shen, G.; Zhang, F.; Yang, B.; Chu, C.; Liang, X. A novel amide
stationary phase for hydrophilic interaction liquid chromatography and
ion chromatography. Talanta 2013, 115, 129−132.
(16) Kong, W. J.; Xing, X. Y.; Xiao, X. H.; Wang, J. B.; Zhao, Y. L.;
Yang, M. H. Multi-component analysis of bile acids in natural Calculus
bovis and its substitutes by ultrasound-assisted solid-liquid extraction
and UPLC-ELSD. Analyst 2012, 137, 5845−5853.
(17) Regna, E. A.; Bruins, P. F. Recovery of aconitic acid from
molasses. Ind. Eng. Chem. 1956, 48, 1268−1277.
(18) Bloch, E.; Ritchie, R. J. Ion exchange − operation of commercial
scale plant for demineralization of cane sirups and molasses. Ind. Eng.
Chem. 1947, 39, 1581−1584.

E dx.doi.org/10.1021/jf5008874 | J. Agric. Food Chem. XXXX, XXX, XXX−XXX