Journal of Chromatography A 1660 (2021) 462664

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

High performance liquid chromatographic method with post-column

detection for quantification of reducing sugars in foods
Esin Akyüz a, Kevser Sözgen Başkan a,∗, Esma Tütem a, Reşat Apak a,b
a
Department of Chemistry, Faculty of Engineering, İstanbul University-Cerrahpaşa, 34320 Avcılar-İstanbul, Turkey
b
Turkish Academy of Sciences (TUBA), Ankara, Turkey

a r t i c l e i n f o a b s t r a c t

Article history: A novel liquid chromatographic analysis method with post-column detection for sugars was developed
Received 27 August 2021 to improve existing methods in regard to operation time, selectivity, and sensitivity. This method in-
Revised 25 October 2021
volves separation of reducing sugars on HPLC column at 30 °C and 0.8 mL min−1 flow rate, post−column
Accepted 28 October 2021
reaction of sugars with Cu(II)-neocuproine (Nc) reagent at 80 °C and 0.3 mL min−1 flow rate, and mea-
Available online 31 October 2021
surement of Cu(I)-Nc product at 450 nm. The proposed assay was applied to glucose, fructose, maltose,
Keywords: and lactose as reducing sugars. Non-reducing sucrose was determined indirectly, after conversion to its
Saccharide analysis constitutive monomers glucose and fructose by hydrolysis, and analysis with a relative error from -2.41 to
Glucose 2.09%. Honey, apple juice, and milk samples were evaluated as commercial products. The results obtained
Fructose with the proposed assay compared to those of the alkaline Cu(II)−Nc reference method were found close
Lactose to each other, and compatible with the label values of commercial products. The accuracy of the devel-
Post-column HPLC
oped method was performed by spiking glucose to honey and lactose to milk samples using two different
Cu(II)-neocuproine
concentrations. The obtained recoveries with respect to the post-column HPLC method were between 97
and 105% for honey and 96–107% for milk. The method gave linear responses against sugar concentra-
tion with correlation coefficients greater than 0.996 for the four analytes (glucose, fructose, maltose and
lactose) in a range of 9.0 - 342.3 mg L−1 with LOD values ≤ 7.4 mg L−1 . With the developed method, it
was possible to sensitively determine reducing sugars in various food samples at a lower temperature of
post-column reaction (compared to literature values) with easy application of low cost reagents requiring
minimal preliminary operation.
© 2021 Elsevier B.V. All rights reserved.

1. Introduction disaccharides and these have a reducing form, as one of the two
units may have an open−chain with an aldehyde group. However,
Carbohydrates are primarily a combination of carbon and wa- sucrose and trehalose are non-reducing disaccharides, in which the
ter, and many of them have the empirical formula (CH2 O)n , where anomeric carbons of the two units are linked together [2]. Iden-
“n” is the number of repeated units. Some compounds may also tification and quantification of the simple carbohydrates (glucose,
contain other elements, such as nitrogen, phosphorus, or sulfur. fructose, sucrose, maltose and lactose) in food is important for the
In addition to being natural components of food, they are the food and beverage industry, because sugars serve as indicators for
source of energy for vital activities in the cell [1]. Carbohydrates food characteristics such as taste, flavor and naturalness [3]. In ad-
are examined under four groups according to the number of sim- dition, the concentrations of these compounds should be specified
ple sugar molecules in the molecule: monosaccharides (fructose, on the labels of many natural and processed products sold ready-
glucose, galactose, ribose, and xylose), disaccharides (maltose, lac- to-go.
tose, cellobiose, sucrose, and trehalose) oligosaccharides (raffinose, Sugars make up the main components of honey, which is
stachyose, cyclodextrin, etc.) and polysaccharides (cellulose, starch known as one of the natural and sweet food sources. Today,
and glycogen. All free monosaccharides have free aldehyde or ke- low-cost sweetening products, most commonly high fructose corn
tone groups, and these are referred to as ‘reducing sugars (RS)’ syrups (HFCS) and invert sugars [4], have replaced the natural car-
since they reduce oxidizing substances. Lactose and maltose are bohydrates in honey. HFCS is produced by chemical and enzymatic
hydrolysis of corn starch to glucose, followed by isomerization of
glucose to fructose with glucose isomerase. As a result of this pro-
∗
Corresponding author. cess, the products obtained containing various amounts of glu-
E-mail address: [email protected] (K.S. Başkan). cose and fructose are called HFCS-90 (90% fructose and 10% glu-

https://doi.org/10.1016/j.chroma.2021.462664
0021-9673/© 2021 Elsevier B.V. All rights reserved.
E. Akyüz, K.S. Başkan, E. Tütem et al. Journal of Chromatography A 1660 (2021) 462664

cose), HFCS-55 (55% fructose and 45% glucose), and HFCS-42 (42% ence of Nc in an alkaline medium, followed by the measurement of
fructose and 58% glucose). HFCS-90 is the major product and is absorbance of the yellow [Cu(Nc)2 ]+ chelate at 450 nm. [31]. The
blended with glucose syrup to obtain HFCS-42 and HFCS-55 [5]. Al- alkaline Cu(II)-Nc assay has superior properties when compared
though being widely used in many processed products such as car- with the widely applied colorimetric phenol-sulfuric [32] and dini-
bonated and fruit drinks, chocolate, cake, confectionery, jam, mar- trosalicylic acid [33] assays; such as the use of relatively nontoxic
malade, and jelly, HFCS may cause various pathological changes, reagents, convenience, speed, high sensitivity, and repeatability.
type-2 diabetes, glucose intolerance, oxidative stress, insulin resis- In this work, for eliminating the disadvantages of existing
tance, hypertension, obesity, and cardiovascular diseases [6]. In the methods in the literature such as time consuming operations, low
wine industry, the amount of d-fructose and d-glucose (total re- selectivity and sensitivity, we developed a cheap, reliable, sensi-
ducing sugar) is one of the most important quality parameters. Ad- tive, selective, new on−line HPLC method. The most suitable ex-
ditional requirements of carbohydrates quantification in foods can perimental conditions (temperature, pH, reaction time, etc.) of the
be listed as determination of the general composition and indi- proposed chromatographic method were specified, under which
rectly energy values and nutritional content of foodstuffs, estima- the sugar species found in natural samples were separated in an
tion of fermentable carbohydrate ratios in the production of fer- aminopropyl silica HPLC column. Each separated RS was reacted
mented foods, determination of maturity levels of some fruits, and with the Cu(II)-Nc reagent in the post-column reactor, and the re-
identification of food blending, e.g., whether plant-based foods are sulting Cu(I)-Nc product was identified by photodiode array (PDA)
added to some processed animal foods. detector at 450 nm. Chromatographic analysis data were compared
In the literature, many methods were developed to deter- with the total sugar values found by the reference spectropho-
mine sugars in foods, such as paper chromatography [7], gas tometric alkaline Cu(II)-Nc method. Solid phase extraction (SPE)
chromatography−mass spectrometry (GC−MS) [8], capillary elec- technique, requiring less solvent and described in our previous
trophoresis (CE) [9–12], near infrared (NIR) and medium infrared work [31], was used to clean-up the RS in honey, apple juice, and
(MIR) spectroscopy [13,14], Raman spectroscopy [15], and high milk samples from phenolic compounds and other possible inter-
performance liquid chromatography (HPLC) with various detectors ferences.
such as corona-charged aerosol detector (C−CAD), evaporative light
scattering detector (ELSD) [16–17], and refractive index detector
2. Materials and methods
(RID) [18]. These methods are of high cost and low selectivity, and
require more analysis time, putting forward a need for developing
2.1. Chemicals and instrumentation
more selective, simple, and fast alternative methods.
HPLC is a common technique for the analysis of sugars and
Copper(II) chloride dihydrate was supplied from Merck (Darm-
other food components, because it is simple, specific and allows
stadt, Germany); sucrose, d-(-)-fructose, d-(+)-maltose monohy-
quantitative analysis with effective separation of sugars. Although
drate, d-lactose monohydrate, and anhydrous 2,9-dimethyl-1,10-
the most common detection method for simple sugars is the re-
phenanthroline (neocuproine: Nc) were supplied from Aldrich
fractive index (RI), it is known that the RI detector (RID) suffers
(Taufkirchen, Germany); d-(+)-glucose, gallic acid monohydrate,
from disadvantages such as low sensitivity, lack of selectivity, and
potassium sodium tartrate tetrahydrate, trichloroacetic acid (TCA),
the presence of many factors that change the refractive index of
sodium carbonate, sodium hydroxide, hydrochloric acid, Discov-
the solvent [19–21]. Also, this method is not sensitive enough for
ery DPA-6S (250 mg, 3 mL), and Discovery DSC-18 (1 g, 6 mL)
trace sugar analysis and is not suitable for gradient elution as the
solid phase extraction (SPE) cartridges were supplied from Sigma-
results obtained are affected by temperature and mobile phase
Aldrich (Taufkirchen, Germany).
composition. For this reason, many studies have been carried out
Chromatographic separation and quantification of reducing sug-
in liquid chromatographic sugar analysis, both to increase the sen-
ars were performed using a Waters HPLC system (Milford, MA,
sitivity and to allow the use of conventional photometric and flu-
USA) equipped with a 1525 binary pump, a 2998 PDA (photodi-
orometric detectors. There are two modes of derivatization in liq-
ode array) detector (Chelmsford, MA, USA) and in−line degasser.
uid chromatography, i.e. pre-column and post-column. Post-column
The HPLC−coupled post−column system consisted of a reaction
derivatization requires reaction after separation, but it should be
module placed after the chromatographic column equipped with
performed before detection [22–28]. Reducing sugars can reduce
an RXN−10 0 0 reaction coil (volume 1 mL, Waters), and the car-
some metal ions, such as Fe(III) and Cu(II) ions. When the reduced
rier stream from the HPLC pump merged with the post−column
form of the metal ions [iron(II) and copper(I)] bind to the suitable
reagent flowing through a 0.50-mm i.d. knitted PTFE (Teflon) tub-
compound to form a chelate having absorption in the UV or visible
ing (total length ≈ 5 m) with reaction time ≤ 1 min. Data acqui-
region, the spectrophotometric determination of reducing sugars
sition was accomplished using Empower PRO (Waters Associates,
can be possible. The oldest application of this in chromatography
Milford, MA). The pH measurements were made with the aid of a
is based on the formation of a Cu(I)−2,2 -bicinchoninate chelate
Inolab pH7110 pH-meter using a glass electrode (Weilheim, Ger-
having an absorption maximum at 562 nm. It was stated that the
many). Varian Cary 1E UV–vis spectrophotometer (Sydney, Aus-
sensitivity of the method was about 10−10 moles for monosac-
tralia) was used to measure absorbances using a pair of matched
charides and the reaction time was 25 min at 80°C [29]. In this
quartz cuvettes of 1 cm thickness. Millipore Simpak1 Synergy185
method, an anion exchange resin in the sulfate form was used for
(France) ultrapure water system was used to obtain ultrapure wa-
the required separation and aqueous ethanol solution for elution.
ter for all solutions. To stir the incubation solutions, a Select vortex
Another method that increases the sensitivity in HPLC analysis is
mixer was used. To incubate the solutions, a Witeg water bath was
the use of fluorescence (FL) detection, which may have two (pre-
used. The SPE processes were carried out with the help of Agilent
and post-column) derivatization procedures. This method can be
Vac Elut 12 position manifold (California, USA) and Isolab vacuum
very sensitive and selective, yet misleading results may result from
pump (Istanbul, Turkey).
complex procedures and side reactions. In addition, in a study con-
ducted by binding fluorophore groups, it was stated that fructose
could not be determined precisely [30]. 2.2. Preparation of solutions
In our previous work, we developed a spectrophotometric
method for determination of total RS content of various samples. Copper(II) chloride (10 mM) and sodium potassium tartrate
The method was based on cupric ion reduction by RS in the pres- (100 mM) solutions were prepared in ultrapure water separately.

2
E. Akyüz, K.S. Başkan, E. Tütem et al. Journal of Chromatography A 1660 (2021) 462664

Nc solution (15 mM) was prepared in ethanol. The alkaline solu- 5 μm particle size) (USA) with 20 μL sample injection. For the
tion was prepared by mixing 500 mM NaOH and 2% (w/v) Na2 CO3 best separation in the HPLC column, flow rate, column tempera-
solutions in final concentrations. ture, and post−column reaction conditions were specified. In the
The post-column reagent was prepared by mixing copper(II) post−column reactor, optimal flow rate and temperature were
chloride, Nc, alkaline, and sodium potassium tartrate solutions at specified for generation of the product from alkaline Cu(II)-Nc
1:1:1:1 (v/v/v/v) ratio and diluted 4 times with ultrapure water. reagent with reducing sugars separated in the HPLC column. Af-
Glucose, fructose, maltose, lactose and sucrose solutions ter separation in the column, reducing sugars were contacted with
(1.0 mM each) were freshly prepared in ultrapure water. alkali Cu(II)-Nc reagent in the post−column reactor at 80 °C. The
flow rate of the reagent in the post−column system was chosen as
2.3. Hydrolysis of sucrose 0.3 mL min−1 . The yellow reaction product ([Cu(Nc)2 ]+ ) was mea-
sured at 450 nm using PDA detector.
Samples (apple juice, honey, and milk) analyzed for their RS A new binary gradient elution program was developed and ap-
content may contain sucrose, which requires hydrolysis before de- plied with two different solutions: solvent A (acetonitrile) and sol-
termination. After sucrose is hydrolyzed, glucose and fructose are vent B (bidistilled water). A total time of 22 min was allowed for
obtained in equal amounts (1:1 ratio), which respond to the devel- the gradient elution program, which was applied as follows: Ini-
oped method. In the hydrolysis process, 15 mL of the sucrose solu- tially 90% A – 10% B; 7 min from 90 to 80% A; 6 min from 80 to
tion at two different concentrations (1.0 and 2.0 mM) were taken, 75% A, 1 min from 75 to 70% A, 3 min from 70 to 65% A, 3 min
0.5 mL concentrated HCl solution added to the solution, and incu- from 65 to 60% A, and 2 min from 60 to 55%. The rates of change
bated at 60 °C for 10 min. After this period, the hydrolyzate was in solvent ratios we specified at each step were linear and provided
neutralized with NaOH (2.0 M), and the total volume was com- using the "curve 6 mode (Empower 2 Software, Waters Corpora-
pleted to 25 mL with ultrapure water [34]. tion). The flow rate was 0.8 mL min−1 , and the separation was per-
formed at 30 °C column oven temperature. The calibration curves
2.4. Preparation of real samples and linear regression equations of peak area vs. concentration (mg
L−1 ) were established for the reducing sugars tested with three in-
As food samples, 100% apple juice, honey, and milk, as jections.
packaged products were bought from local markets in İstanbul
(Turkey). Apple juice sample was directly filtered through the glass 3. Results and discussion
fiber/polyethyleneterephtalate microfilter (GF/PET, 1.0/0.45 μm),
and filtrate was diluted 50 times with ultrapure water. The honey The purpose of developing our on−line HPLC method with
sample was prepared by diluting 1 g portion to 100 mL with ul- post−column detection was to design an alternative assay for re-
trapure water. Milk sample was prepared by adding 2.0 mL of 70% ducing sugars in natural and commercial foods, bearing the ad-
(w/v) trichloroacetic acid (TCA) to 2.0 mL milk to precipitate the vantageous properties of ease of application, minimal use of toxic
proteins, followed by centrifugation for 5 min at 4500 rpm. In chemicals, and ability to rapidly obtain sensitive, selective and ac-
preparation of the milk sample for analysis, 2 mL were taken and curate results. In order to fill the aforementioned gap in the liter-
2 mL of 70% (w/v) TCA were added to precipitate the proteins, ature, a new method based on photodiode array (PDA) detection
and the precipitate was separated by centrifugation for 5 min at at 450 nm of the {RS+Cu(II)-Nc} reaction product, i.e. cuprous-
4500 rpm. One milliliter of supernatant was filtered through the neocuproine colored chelate, formed in a reactor coil placed after
GF/PET microfilter, and diluted to 100 mL with ultrapure water. the HPLC column is proposed. Previously, an on-line HPLC-cupric
reducing antioxidant capacity (CUPRAC) method was developed by
2.5. Spectrophotometric and chromatographic analyses of reducing Çelik et al., which has the advantage of sensitive determination of
sugars individual antioxidant constituents in complex matrices like plant
extracts. In this study, it was stated that simple sugars were not
2.5.1. Alkaline Cu(II)-Nc assay oxidized by Cu(II)-Nc reagent (pH 7.0) in the post-column reactor
This assay, which was described in detail in the published [35]. It should be remembered that reducing sugars are not an-
study by Sözgen Baskan et al. [31], is based on the reduction tioxidant compounds, and require totally different reaction condi-
of [Cu(Nc)2 ]2+ complex to [Cu(Nc)2 ]+ chelate in alkaline medium tions for being measured in a post-column colorimetric method. In
with RS and subsequent absorbance measurement of the yellow our previous study where we suggested an alternative method for
reaction product at 450 nm. To a test tube, 1.0 mL of each Cu(II) spectrophotometric total reducing sugar determination, we used
chloride, Nc, alkaline, sodium potassium tartrate (used for the pre- Cu(II)-Nc reagent in alkaline medium and obtained successful re-
vention of copper(II) hydroxide precipitation in alkaline medium) sults [31]. Even, microplate application of this method was devel-
solutions, (x) mL standard or sample solution and (1-x) mL pure oped [36]. As a result, we can say that our new study is a useful
water were added so that the total volume was 5.0 mL; the con- application that combines the advantages of the above-mentioned
tents were stirred on a vortex mixer. The incubation of mixtures methods.
was performed in a thermostatic water bath at 60 °C for 20 min,
and the absorbances were measured at 450 nm (A450 ) against a 3.1. Proposed on−line HPLC method
reagent blank.
The standard calibration curve of each sugar was obtained by With the aim of effective separation of mono- and disaccha-
using absorbance vs. molar concentration, and the molar absorp- rides, the widely used LC−NH2 (aminopropyl silica) column was
tivity of the alkaline Cu(II)-Nc method for each sugar was found preferred as the stationary phase in hydrophilic interaction based
from the slope of the calibration line concerned. HPLC analysis. Acetonitrile and water were used for the gradient
elution program. Solvent ratios were changed linearly during the
2.5.2. Proposed on−line hplc method elution process as specified in the “2.5.2.” section. The reducing
In this study, chromatographic analysis was performed using a sugar standard solutions, prepared between 9.0 - 342.3 mg L−1
Waters HPLC system with a PDA detector within the wavelength concentrations, were injected 3 times and the calibration equa-
range 20 0–60 0 nm. Separation of analytes was carried out on a Su- tions were calculated using average peak areas (Table 1). The chro-
pelcosil LC−NH2 (aminopropyl silica column; 4.6 mm × 250 mm, matogram obtained from sugar solutions prepared at different con-

3
E. Akyüz, K.S. Başkan, E. Tütem et al. Journal of Chromatography A 1660 (2021) 462664

Table 1
Calibration equations of reducing sugars with respect to the post-column HPLC method (n = 3).

Reducing sugar Calibration equation Correlation coefficient (r) Working range (mg L− 1 )

D-(+)-Glucose Peak area = 9.04 × 104 c + 4.90 × 105 0.9978 9.0 – 90.1
D-(−)-Fructose Peak area = 9.28 × 104 c – 3.31 × 104 0.9968 9.0 – 90.1
D-(+)-Maltose Peak area = 5.06 × 104 c – 1.73 × 105 0.9998 21.4 – 342.3
D-Lactose Peak area = 4.55 × 104 c – 3.90 × 105 0.9996 21.4 – 342.3

Fig. 1. The chromatogram (λ=450 nm) of the sugar solutions, prepared at different concentrations (18.0–342.3 mg L− 1 ) 1-Fructose, 2-Glucose, 3-Maltose, 4-Lactose.

centrations (18.0–342.3 mg L−1 ) is shown in Fig. 1. The LOD values precipitation stability at temperatures higher than 90 °C, added to
for glucose, fructose, maltose, and lactose were calculated as 4.8, serious interference by ferric ions [45]. Besides, Niu et al. showed
5.8, 4.2, and 7.4 mg L−1 , respectively. that Cu(I)-BCA chelate formation was very slow, especially at low
In chromatographic sugar analysis methods, linear concentra- concentrations of copper, which may pose a serious drawback to
tion ranges and LOD values show variability depending on the used an on-line HPLC-colorimetric assay having a very short residence
method and detector. Sugars can be determined using a variety of time in the post-column reactor [46]. Another post-column deriva-
chromatographic modes in conjunction with RID, ELSD, and PAD tization method is based on photometric detection (at 520 nm) of
(pulse amperometric detector). The sensitivity of PAD is 100 times a product, which is formed as a result of the reaction of sugars
better than that of RID [26,37,38]. RID detection limit can be only with tetrazolium blue reagent [47]. In this method, sugars were
100 μg mL−1 (100 mg L−1 ) for fructose, glucose, and sucrose [39]. separated by pumping 89% ethanol under 400 psi pressure through
Detection limits for glucose, fructose, sucrose, and maltose have a long narrow column filled with a fine-size strongly anionic resin
been determined with PAD as 0.07, 0.09, 0.21, and 0.27 μg mL−1 , in the sulfate form where monosaccharide mixtures were fully re-
respectively [40]. The detection limit of ELSD has been reported solved within 3–4 h. A modified method for post-column HPLC
as 44 and 11 μg mL−1 [41]. When LOD values are compared, analysis of reducing sugars with tetrazolium blue reagent was de-
it is seen that the sensitivity of our proposed method is higher scribed by Aliani et al. [27]. Detection was carried out with a pho-
(and LODs lower) than those provided by RID and ELSD detectors. tometric detector at 550 nm. It was stated that the required wa-
Chemical derivatization methods allow the conversion of carbohy- ter bath temperature should be 95 °C for optimal and reproducible
drates into derivatives that can be detected with higher sensitivity. color formation, which may pose a risk for reducing sugars degra-
Fluorometric detector, UV - Vis detector [42], or PAD can be used dation over extended retention periods [48]. In this method, limit
for the detection of derivatives or reaction products. of detection values for glucose and fructose were determined as
Sugar determination studies based on post-column UV - Vis de- 0.72 and 0.71 mg mL−1 , respectively.
tection are less in number than other methods [43], possibly be- The appropriate method to choose for sugar detection in dif-
cause sugars are hard to oxidize and different reducing sugars may ferent samples depends on the number of samples, the kind and
display a wide range of redox potentials [44] that cannot be inter- contents of the matrix. Electrochemical or fluorometric detection-
cepted by most colorimetric redox reagents (mainly used in an- based chromatographic methods can be preferred for sugar deter-
tioxidant capacity assays). The average redox potential of these minations that require selectivity and high sensitivity. However,
sugars is close to or higher than those of most antioxidant as- methods based on photometric detection can be easily used in
says’ reagents laying between 0.6–0.7 V. Thus, it is very difficult samples containing relatively high amounts of sugar such as food
to carry out a post-column colorimetric reaction that is thermo- samples (honey, fruit juices, milk, beverages, etc.). Consequently,
dynamically and kinetically feasible, and on-line with chromato- we took advantage of the ability of sugars to rapidly reduce the
graphic separation of sugars. One of the oldest applications is [Cu(Nc)2 ]2+ reagent in the post-column reactor and used a PDA
based on the reaction of the reagent with RS to form a Cu(I)−2,2 - (photodiode array detector) for the absorbance measurement of
bicinchoninate chelate and measuring the absorbance of this prod- [Cu(Nc)2 ]+ product in the visible region (at 450 nm). This can-
uct at 562 nm. It was stated that the sensitivity of the method was not be accomplished by most antioxidant capacity assays working
about 10−10 moles for monosaccharides [29]. In this method, an in electron-transfer mechanism (e.g., FRAP, ferricyanide) or mixed-
anion exchanger column was used for the pre-separation of sugars. mode hydrogen and electron transfer (e.g., ABTS, DPPH) due to the
However, it should be noted that the copper-2,2 -bicinchoninate unfavorable reaction conditions in the post-column reactor (aris-
assay suffered from reagent stability below pH 5.5, and also from ing from both thermodynamic and kinetic reasons, i.e. lack of suffi-

4
E. Akyüz, K.S. Başkan, E. Tütem et al. Journal of Chromatography A 1660 (2021) 462664

Table 2
Sugar content determined by post-column HPLC analysis of the sucrose hydrolysate and glucose−fructose mixtures at the same concentra-
tions. (n = 3).

Glucose (mg L− 1 ) Fructose (mg L− 1 )

Sample
Expected Founda RSD% Expected Founda RSD%

Hydrolysate 1 b
180.16 183.92±5.07 2.09 180.16 178.63±6.53 −0.85
Mixture 1c 180.16 182.54±8.98 1.32 180.16 180.84±4.72 0.38
Hydrolysate 2b 360.32 354.72±2.72 −1.55 360.32 357.36±3.27 −0.82
Mixture 2c 360.32 351.63±5.94 −2.41 360.32 356.97±7.91 −0.93
a
The results were expressed as {mean value ± standard deviation (SD)} (n = 3).
b
Sucrose solutions were prepared as 1.0 mM (1) and 2.0 mM (2) concentrations.
c
Synthetic mixtures (1 and 2) were prepared to contain glucose and fructose at equivalent concentrations to those of hydrolysis products.

Table 3
Analysis results of honey, apple juice and milk samples with respect to the post-column HPLC and alkaline Cu(II)-Nc methods.

Sample Alkaline Cu(II)-Nc method Post-column HPLC method Label Value

Glucose Fructose Lactose

Honey 81.50±0.81 g/100 g a 13.89±1.40 g/100 g 67.02±0.89 g/100 g – 82.30 g/100 g

Apple juice 9.41±0.45 g/100 mL a 0.45±0.08 g/100 mL 9.06±0.13 g/100 mL – 11.50 g/100 mL
Milk 7.76±0.21 g/100 mL b – – 7.32±0.40 g/100 mL 4.60 g/100 mL c
3.84±0.24 g/100 mL a 3.85±0.40 g/100 mL a

a
Calculated as glucose equivalent.
b
Calculated as lactose equivalent.
c
The value was given as carbohydrate content.

Fig. 2. Honey sample chromatogram (λ=450 nm). 1-Fructose, 2-Glucose.

cient gap between the redox potentials of the reagent and analyte, prepared solutions were analyzed by the proposed on−line HPLC
and possible incomplete reaction in the post-column reactor at the method, with results given in Table 2.
specified temperature and retention time, respectively). The glucose content of the hydrolyzate of the sucrose solu-
The order of elution of reducing sugars from the HPLC column tions and synthetic mixtures were determined with respect to the
was as follows: fructose, glucose, maltose, and lactose. As seen in post−column HPLC method, and the relative error values were
the Fig. 1, separation of RSs was performed within 22 min. The found from −2.41 to 2.09% compared to theoretical values. The rel-
linear regression analyses were performed with correlation coeffi- ative errors for determining the fructose content of the same solu-
cients greater than 0.996 for the four analytes in a range from 9.0 tions were lower and found between −0.93 and 0.38%. These re-
- 342.3 mg L−1 . sults demonstrated that the developed method could also be used
for the analysis of sucrose−containing samples.

3.2. Post−column HPLC analysis of the hydrolyzed sucrose solution 3.3. On−line HPLC analysis of food samples

The sucrose solutions were hydrolyzed at two different concen- The developed method was applied to honey, apple juice and
trations (1.0 and 2.0 mM) according to the above given procedure. milk samples. The SPE method was first applied to remove com-
Standard mixtures containing glucose and fructose at the same pounds thought to be disruptive to honey and apple juice sam-
concentrations with the hydrolyzed solution were prepared. All ples analyses. In the SPE application, two different cartridges were

5
E. Akyüz, K.S. Başkan, E. Tütem et al. Journal of Chromatography A 1660 (2021) 462664

Fig. 3. Milk and lactose-added milk sample chromatograms (λ=450 nm). 1-Milk, 2- 0.1 mM Lactose-added milk, 3- 0.2 mM lactose-added milk.

Table 4
Analysis results of the amount of sugar (glucose or lactose) added in known quantities to the honey and milk samples with respect to the post-
column HPLC and alkaline Cu(II)-Nc methods.

Post-column HPLC method Alkaline Cu(II)−Nc method

Sample
Glucose (mg L− 1 ) Fructose (mg L− 1 ) Total reducing sugar (mg glucose L− 1 )

Honey 2.78 ± 0.23 13.20 ± 0.13 16.30 ± 0.16

Honey + 9 mg L− 1 Glucose 12.15 ± 0.12 (%104) 13.16± 0.30 24.99 ± 0.32 (%97)
Honey + 18 mg L− 1 Glucose 21.68 ± 0.12 (%105) 13.87 ± 0.60 34.87 ± 0.84 (%103)
Lactose (mg L− 1 )
Milk 52.30 ± 0.55 56.20 ± 0.95
Milk + 34a ve 43b mg L− 1 Lactose 84.99 ± 1.41 (%96) 102.30 ± 1.51 (%107)
Milk + 68a ve 86b mg L− 1 Lactose 120.56 ± 1.14 (%100) 145.90 1.13 (%104)
a
Added amounts for post-column HPLC method application.
b
Added amounts for alkaline Cu(II)−Nc method application.

used, including C18 and polyamide resin. The application was car- by binding chromophore and fluorophore groups, making it pos-
ried out as described in the study of the spectrophotometric total sible to use fluorescence detectors. The aim of this study was to
sugar assay by Sözgen Başkan et al. [31]. In order to compare the develop an inexpensive, reliable, sensitive, selective, and innova-
results obtained, the spectrophotometric alkaline Cu(II)-Nc total re- tive on−line liquid chromatographic method of analysis for re-
ducing sugar determination method was applied to the same sam- ducing sugars where the disadvantages of literature methods such
ples (Table 3). The results are given as glucose, fructose, or lactose as time−consuming operations, incomplete derivatization reaction,
equivalents using the calibration equations (Table 1) determined low selectivity and sensitivity were eliminated. In this context, glu-
by HPLC with post-column detection application for standards. The cose, fructose, maltose, and lactose were tested as reducing sug-
LOD values in glucose equivalents calculated on the original sam- ars. A non-reducing sugar, sucrose, was analyzed indirectly af-
ples were given in mg/kg and mg/L units. These calculations were ter acidic hydrolysis to produce a mixture of glucose and fruc-
made on the basis of the chosen sample weights, dissolution and tose. The reaction product of HPLC-separated reducing sugars with
dilution volumes, which may be improved by decreasing the men- Cu(II)-Nc reagent in the post-column reactor produced a stable
tioned parameters. Accordingly, the LOD values calculated under single product of [Cu(Nc)2 ]+ which perfectly obeyed Beer’s law
the experimental conditions for honey, apple juice, and milk were at 450 nm wavelength in a reasonable concentration range. The
2400 mg/kg, 240 mg/L, and 960 mg/L, respectively method proved to be effective in performing a rapid reaction in
The chromatogram of honey sample is shown in Fig. 2, and the mildly-heated post-column reactor with high efficiency, owing
the chromatograms obtained before and after lactose addition at to the high redox potential of the cupric-neocuproine reagent. The
known concentrations to the milk sample are shown in Fig. 3. stable post-column reagent, cupric-neocuproine, having high oxi-
In order to evaluate the accuracy of the developed method, dation power and selectivity toward reducing sugars was appro-
sugar (i.e. glucose to the honey sample and lactose to the milk priately compared in performance with other cupric chelates of
sample) addition was made to the honey and milk in known quan- phenanthroline derivatives such as BCA. The on−line HPLC method
tities, and recovery (%) values were calculated (Table 4). was applied to honey, apple juice and milk samples after SPE sep-
As observed in Table 4, the results obtained (with 96–107% re- aration of phenolic matter, and their sugar content was identified
covery values) showed that the method developed could be used and quantified simultaneously. Spectrophotometric alkaline Cu(II)-
to determine the type and amount of sugars in commercial prod- Nc total sugar determination assay was also applied to the same
ucts. samples to compare the results obtained, which were found to be
consistent with each other. The accuracy of the developed method
4. Conclusions was performed by spiking glucose to honey sample and lactose to
milk sample at two different concentrations. The obtained recov-
Since sugars do not absorb light in the UV - Vis region, RI de- ery values at 96–107% confirmed that the proposed method could
tectors are preferred in HPLC analysis or derivatization is made be used to determine the types and amounts of sugar in commer-

6
E. Akyüz, K.S. Başkan, E. Tütem et al. Journal of Chromatography A 1660 (2021) 462664

cial products. This method can be potentially used in food fraud [17] M.M. Bernárdez, J.M. Miguelez, J.G. Queijeiro, HPLC determination of sugars
detection in relation to its capability to identify and quantify the in varieties of chestnut fruits from Galicia (Spain), J. Food Compost. Anal. 17
(2004) 63–67, doi:10.1016/S0889-1575(03)00093-0.
different composition and content of sugars in food matrices such [18] J.P. Yuan, F. Chen, Simultaneous separation and determination of sugars, ascor-
as those contained in the corn syrup unlawfully added to foods. bic acid and furanic compounds by HPLC-dual detection, Food Chem 64 (1999)
423–427, doi:10.1016/S0308-8146(98)0 0 091-0.
[19] D. Duarte-Delgadoa, C.-.E. Narváez-Cuencab, L.-.P. Restrepo-Sánchezb,
Author statements A. Kushalappac, T. Mosquera-Vásqueza, Development and validation of a
liquid chromatographic method to quantify sucrose, glucose, and fructose
Esin Akyüz: Investigation, Experimental design and application, in tubers of Solanumtuberosum Group Phureja, J. Chromatogr. B 975 (2015)
18–23, doi:10.1016/j.jchromb.2014.10.039.
Writing – original draft. [20] J. Lόpez Hernández, M.J. González-Castro, I. Naya Alba, C. de la Cruz Gar-
Kevser Sözgen Başkan: Resources, Methodology, Validation, cía, High-performance liquid chromatographic determination of mono- and
Writing – original draft. oligosaccharides in vegetables with evaporative light-scattering detection and
refractive index detection, J. Chromatog. Sci. 36 (1998) 293–298, doi:10.1093/
Esma Tütem: Resources, Supervision. chromsci/36.6.293.
Reşat Apak: Conceptualization, Review & editing. [21] D.L. Sexton, Determination of lactose by reversed-phase high performance liq-
uid chromatography, Electronic Theses and Dissertations, Paper 878 (2004)
https://dc.etsu.edu/etd/878.
Declaration of Competing Interest [22] P. Vratny, U.A.T. Brinkman, R.W. Frei, Comparative study of post-column reac-
tions for the detection of saccharides in liquid chromatography, Anal. Chem.
The authors declare that they have no known competing finan- 57 (1985) 224–229, doi:10.1021/ac00279a052.
[23] B.D. Franta, L.R. Mattick, J.W. Sherbon, The analysis of pentoses in dry wine by
cial interests or personal relationships that could have appeared to high performance liquid chromatography with postcolumn derivatization, Am.
influence the work reported in this paper. J. Enol. Vitic. 37 (1986) 269–274.
[24] S. Honda, Postcolumn derivatization for chromatographic analysis of carbo-
hydrates, J. Chromatogr. A 720 (1986) 183–199, doi:10.1016/0021-9673(95)
Acknowledgements 00212-X.
[25] H. Engelhardt, P. Ohs, Trace analysis of sugars by HPLC and post-column
The authors would like to express their gratitude to the Sci- derivatization, Chromatographia 23 (1987) 657–662.
[26] A. Coquet, W. Haerdi, R. Degli Agosti, J.-.L. Veuthey, Determination of
entific and Technological Research Council of Turkey (TÜBİTAK- sugars by liquid chromatography with post-column catalytic derivatization
KBAG-115Z004) for its financial support. This study was also and fluorescence detection, Chromatographia 38 (1994) 12–16, doi:10.1007/
funded by Scientific Research Projects Coordination Unit of Istan- BF02275720.
[27] M. Aliani, L.J. Farmer, Post-column derivatization method for determination
bul University-Cerrahpasa. Project number: FDP-2016–20114. of reducing and phosphorylated sugars in chicken by high performance liq-
uid chromatography, J. Agric. Food Chem. 50 (2002) 2760–2766, doi:10.1021/
References jf011335i.
[28] K.K. Kim, J.O. Escobedo, N.N. St. Luce, O. Rusin, D. Wong, R.M. Strongin, Post-
[1] Q. Wang, Y. Fang, Analysis of sugars in traditional Chinese drugs, J. Chromatogr. column HPLC detection of mono- and oligosaccharides with a chemosensor,
B 812 (2004) 309–324, doi:10.1016/j.jchromb.2004.09.027. Org. Lett. 5 (2003) 5007–5010, doi:10.1021/ol035978q.
[2] H.D. Belitz, W. Grosch, P. Schieberle, Carbohydrates, in: Food Chem- [29] K. Mopper, E.M. Gindler, A new noncorrosive dye reagent for automatic
istry, Springer, Berlin, Heidelberg, 2009, pp. 248–339, doi:10.1007/ sugar chromatography, Anal. Biochem. 56 (1973) 440–442, doi:10.1016/
978- 3- 540- 69934- 7_5. 0 0 03- 2697(73)90210- 8.
[3] D.B. Gomis, D.M. Tamayo, J.M. Alonso, Determination of monosaccharides in [30] A. Meyer, C. Raba, K. Fischer, Ion-pair RP-HPLC determination of sugars, amino
cider by reversed-phase liquid chromatography, Anal. Chim. Acta. 436 (2001) sugars, and uronic acids after derivatization with p-aminobenzoic acid, Anal.
173–180, doi:10.1016/S0 0 03-2670(01)0 0889-3. Chem. 73 (2001) 2377–2382, doi:10.1021/ac001402s.
[4] I.S. Arvanitoyannis, Novel quality control methods in conjunction with chemo- [31] K. Sözgen Başkan, E. Tütem, E. Akyüz, S. Özen, R. Apak, Spectrophotometric
metrics (Multivariate Analysis) for detecting honey authenticity, Crit. Rev. Food total reducing sugars assay based on cupric reduction, Talanta 147 (2016) 162–
Sci. Nutr. 45 (2005) 193–203. 168, doi:10.1016/j.talanta.2015.09.049.
[5] K. Parker, M. Salas, V.C. Nwosu, High fructose corn syrup: production, uses and [32] M. Dubois, K.A. Gilles, J.K. Hamilton, P.A. Rebers, F. Smith, Colorimetric method
public health concerns, Biotechnol. Mol. Biol. Rev. 5 (2010) 71–78. for determination of sugars and related substances, Anal.Chem 28 (1956) 350–
[6] A.P. Ross, T.J. Bartness, J.G. Mielke, M.B. Parent, A high fructose diet impairs 356, doi:10.1021/ac60111a017.
spatial memory in male rats, Neurobiol. Learn. Mem. 92 (2009) 410–416. [33] G.L. Miller, Use of dinitrosalicylic acid reagent for determination of reducing
[7] C.M. Wilson, Quantitative determination of sugars on paper chromatograms, sugar, Anal.Chem 31 (1959) 426–428, doi:10.1021/ac60147a030.
Anal. Chem. 31 (1959) 1199–1201. [34] M.C.C.Ruvolo-Takasusuki V.A.A.Toledo, A.J.B. Oliveira, E.D. Chambó,
[8] Z.F. Katona, P. Sass, I. Molnár-Perl, Simultaneous determination of sugars, sugar S.M.S. Lopes, Chapter 14-Spectrophotometry as a tool for dosage sugars
alcohols, acids and amino acids in apricots by gas chromatography–mass spec- in nectar or crops pollinated by honey bees, in: J. Uddin (Ed.), Macro
trometry, J. Chromatogr. A. 847 (1999) 91–102, doi:10.1016/S0021-9673(99) to nano spectroscopy, In Tech, Croatia, 2012, pp. 269–290, doi:10.5772/
00333-7. 48043.
[9] X. Xu, W.T. Kok, H. Poppe, Sensitive determination of sugars by capillary zone [35] S.E. Çelik, M. Özyürek, K. Güçlü, R. Apak, Determination of antioxidants by a
electrophoresis with indirect UV detection under highly alkaline conditions, J. novel on-line HPLC-cupric reducing antioxidant capacity (CUPRAC) assay with
Chromatogr. A. 716 (1995) 231–240, doi:10.1016/0 021-9673(95)0 0552-X. post-column detection, Anal. Chim. Acta 674 (2010) 79–88, doi:10.1016/j.aca.
[10] Y.H. Lee, T.I Lin, Determination of carbohydrates by high-performance capil- 2010.06.013.
lary electrophoresis with indirect absorbance detection, J. Chromatogr. B 681 [36] M. Bener, E. Akyüz, F.B. Şen, K. Sözgen Başkan, E. Tütem, R. Apak, A Simple au-
(1996) 87–97. tomated microplate method for determining reducing sugars in food extracts
[11] T. Soga, D.N. Heiger, Simultaneous determination of monosaccharides in glyco- and synthetic serum using cupric-neocuproine as reductant, Turk. J. Chem. 42
proteins by capillary electrophoresis, Anal. Biochem. 261 (1998) 73–78, doi:10. (2018) 794–807, doi:10.3906/kim- 1707- 30.
1006/abio.1998.2727. [37] C.M. Montero, M.C.R. Dodero, D.A.G. Sánchez, C.G. Barroso, Analysis of low
[12] T. Soga, M. Serwe, Determination of carbohydrates in food samples by capillary molecular weight carbohydrates in food and beverages: a Review, Chro-
electrophoresis with indirect UV detection, Food Chem 69 (20 0 0) 339–344. matographia 59 (2004) 15–30, doi:10.1365/s10337- 003- 0134- 3.
[13] L.E. Rodriguez-Saona, F.S. Fry, M.A. McLaughlin, E.M. Calvey, Rapid analysis of [38] L. Monti, S. Negri, A. Meucci, A. Stroppa, A. Galli, G. Contarini, Lactose, galac-
sugars in fruit juices by FT-NIR spectroscopy, Carbohydr. Res. 336 (2001) 63– tose and glucose determination in naturally “lactose free” hard cheese: HPAEC-
74, doi:10.1016/S0 0 08-6215(01)0 0244-0. PAD method validation, Food Chem 220 (2017) 18–24, doi:10.1016/j.foodchem.
[14] G.A. de Oliveira, F. Castilhos, C.M.G.C. Renard, S. Bureau, Comparison of NIR 2016.09.185.
and MIR spectroscopic methods for determination of individual sugars, organic [39] R. Sharma, Y.S. Rajput, G.Dogra Poonam, S.K. Tomar, Estimation of sugars in
acids and carotenoids in passion fruit, Food Res. Int. 60 (2014) 154–162, doi:10. milk by HPLC and its application in detection of adulteration of milk with
1016/j.foodres.2013.10.051. soymilk, Intl. J. Dairy Technol. 62 (2009) 514–519, doi:10.1111/j.1471-0307.
[15] B. Özbalcı, İ.H. Boyacı, A. Topçu, C. Kadılar, U. Tamer, Rapid analysis of sug- 20 09.0 0532.x.
ars in honey by processing Raman spectrum using chemometric methods and [40] T. Soga, Y. Inoue, K. Yamaguchi, Determination of carbohydrates by hydrophilic
artificial neural networks, Food Chem 136 (2013) 1444–1452, doi:10.1016/j. interaction chromatography with pulsed amperometric detection using post-
foodchem.2012.09.064. column pH adjustment, J. Chromatogr. A 625 (1992) 151–155, doi:10.1016/
[16] I. Márquez-Sillero, S. Cárdenas, M. Valcárcel, Comparison of two evaporative 0021- 9673(92)85196- Z.
universal detectors for the determination of sugars in food samples by liquid [41] U. Sharma, P. Bhandari, N. Kumar, B. Singh, Simultaneous determination of ten
chromatography, Microchem. J. 110 (2013) 629–635, doi:10.1016/j.microc.2013. sugars in Tinospora cordifolia by ultrasonic assisted extraction and LC-ELSD,
07.008. Chromatographia 71 (2010) 633–638, doi:10.1365/s10337-010-1520-2.

7
E. Akyüz, K.S. Başkan, E. Tütem et al. Journal of Chromatography A 1660 (2021) 462664

[42] H. Mikami, Y. Ishida, Post-column fluorometric detection of reducing sugars in [46] Y. Niu, J. Wang, C. Zhang, Y. Chen, Rapid determination of trace copper in ani-
high performance liquid chromatography using arginine, Bunseki Kagaku 32 mal feed based on micro-plate colorimetric reaction and statistical partitioning
(1983) E207–E210, doi:10.2116/bunsekikagaku.32.6_E207. correction, Food Chem 221 (2017) 1406–1414, doi:10.1016/j.foodchem.2016.11.
[43] S. Hase, Chapter 28 Pre- and post-column detection-oriented derivatization 012.
techniques in HPLC of carbohydrates, J. Chromatogr. Libr. 66 (2002) 1043–1169, [47] K. Mopper, E.T. Degens, A new chromatographic sugar autoanalyzer with a
doi:10.1016/S0301- 4770(02)80053- 7. sensitivity of 10−10 moles, Anal. Biochem. 45 (1972) 147–153, doi:10.1016/
[44] R.S. Blackburn, A. Harvey, Green chemistry methods in sulfur dyeing: appli- 0 0 03-2697(72)90 015-2.
cation of various reducing d-sugars and analysis of the importance of opti- [48] G. Eggleston, J.R. Vercellotti, Degradation of sucrose, glucose and fructose
mum redox potential, Environ. Sci. Technol. 38 (2004) 4034–4039, doi:10.1021/ in concentrated aqueous solutions under constant pH conditions at ele-
es0498484. vated temperature, J. Carbohyd. Chem. 19 (20 0 0) 1305–1318, doi:10.1080/
[45] M.A. Anwar, M. Iqbal, M.A. Qamar, M. Rehman, A.M. Khalid, Technical commu- 0732830 0 0 08544153.
nication: determination of cuprous ions in bacterial leachates and for environ-
mental monitoring, World J. Microb. Biot. 16 (20 0 0) 135–138, doi:10.1023/A:
1008978501177.