Talanta 147 (2016) 162–168

Contents lists available at ScienceDirect

Talanta
journal homepage: www.elsevier.com/locate/talanta

Spectrophotometric total reducing sugars assay based on cupric

reduction
Kevser Sözgen Başkan n, Esma Tütem, Esin Akyüz, Seda Özen, Reşat Apak
Department of Chemistry, Faculty of Engineering, Istanbul University, 34320 Istanbul, Turkey

art ic l e i nf o a b s t r a c t

Article history: As the concentration of reducing sugars (RS) is controlled by European legislation for certain specific food
Received 7 July 2015 and beverages, a simple and sensitive spectrophotometric method for the determination of RS in various
Received in revised form food products is proposed. The method is based on the reduction of Cu(II) to Cu(I) with reducing sugars
16 September 2015
in alkaline medium in the presence of 2,9-dimethyl-1,10-phenanthroline (neocuproine: Nc), followed by
Accepted 20 September 2015
Available online 25 September 2015
the formation of a colored Cu(I)-Nc charge-transfer complex. All simple sugars tested had the linear
regression equations with almost equal slope values. The proposed method was successfully applied to
Keywords: fresh apple juice, commercial fruit juices, milk, honey and onion juice. Interference effect of phenolic
Reducing sugar compounds in plant samples was eliminated by a solid phase extraction (SPE) clean-up process. The
Spectrophotometry
method was proven to have higher sensitivity and precision than the widely used dinitrosalicylic acid
Cu(II)-Nc
(DNS) colorimetric method.
DNS
Glucose & 2015 Elsevier B.V. All rights reserved.
Fructose

1. Introduction of sugar in different foods and beverages, the lactose content in

Sugar is the general term for a class of sweet-flavored substances
used as food, and a relevant ingredient in many food and beverage
products; it also occurs naturally in fruits, vegetables, milk, and
honey. Simple sugars such as glucose, fructose, and galactose are
called monosaccharides. When two simple sugars undergo dehy-
dration and link to each other by a covalent bond, disaccharides
such as sucrose (derived from glucose and fructose), maltose (from
two glucose units) and lactose (from galactose and glucose) are
formed. All monosaccharides, oligosaccharides (with a few excep-
tions) and other glucose polymers such as starch and cellulose are
named ‘reducing sugars' (RS) because they contain reducing ends
[1]. Nowadays, trendy diets focus on the reduction of sugar content
of foods in order to attenuate the adverse health effects of high
sugar consumption. Due to their presence in many natural products,
RS have received considerable attention. The sugar concentration is
controlled by European legislation for some specific products, like
honey which should contain no less than 60 g of reducing sugar per
100 g, and wine which is classified depending on the RS con-
centration as dry (less than 5 g L
1), semi-dry (5–15 g L
1), semi-
sweet (12–25 g L
1) and sweet (25–125 g L
1) [2].
Measurements of the concentration of reducing ends in a
sample can give very valuable information, such as concentrations
n
Corresponding author.
E-mail address: sozgen@istanbul.edu.tr (K.S. Başkan).
milk, the activity of an enzyme on cellulose and starch [3]. RS are
the end products of many biological processes and enzymatic re-
actions. Many polysaccharide-degrading enzymes are commonly
assayed by quantifying the amount of RS released at the end of an
incubation period [1,4]. Total sugar determination in fruits is also
an important parameter to evaluate their ripening stage [5–7]. The
determination of total reducing sugars is frequently performed by
one of a number of colorimetric procedures [2,4–20]. Most of them
involve colorimetric detection based on the oxidation of the car-
bonyl groups, followed by reaction with a UV–vis absorbing mo-
lecule [3,9]. The dinitrosalicylic acid (DNS) method [9] is widely
used to estimate the concentration of reducing sugars in different
samples [2,10,11]. This method is based on the reduction of 3,5-
dinitrosalicylic acid to the colored 3-amino-5-nitrosalicylate in
alkaline medium. There are the frequent methods based on Cu(II)
or Fe(III) reduction by RS, and complexation of reduced transition
metal ions with suitable ligands (such as 2,2'-bicinchoninate (BCA)
[13,14], neocuproine (Nc) [15,16] for Cu(I) and o-phenanthroline
for Fe(II) [17]), or back-oxidation of the sugar-reduced products
(such as Cu(I) and ferrocyanide) with a colorless reagent (e.g., ar-
seno-molybdate) giving a colored product (such as heteropoly
molybdenum blue [18,19]) for the spectrophotometric determi-
nation of total RS content. Nelson [18] replaced the phosphomo-
lybdate reagent (used by Folin and Wu [20]) with arsenomo-
lybdate solution for color development in the estimation of the
reduced copper. It should be borne in mind that back-oxidation

http://dx.doi.org/10.1016/j.talanta.2015.09.049
0039-9140/& 2015 Elsevier B.V. All rights reserved.
 K.S. Başkan et al. / Talanta 147 (2016) 162–168
methods have an important disadvantage such as the possibility of
air oxidation of Cu(I) to Cu(II) during the heating period before
arsenomolybdate reagent addition. Also these methods are time
consuming and less sensitive than those forming colored com-
plexes with metal ions of lower valency formed during the re-
duction process. 2,2'-Bicinchoninate (BCA) forms a deep blue
complex with Cu(I) in alkaline solution. This chromogenic reaction
found application in automated clinical glucose analysis and was
adapted by McFeeters [13] to manual spectrophotometric de-
termination of sugar samples. Waffenschmidt and Jaenicke [14]
used L-serine instead of aspartic acid for Cu(II)-amino acid com-
plex formation before the reduction with RS, enabling approxi-
mately 10-times sensitivity enhancement of the BCA assay with
the exception of fructose (with only 2.7 times enhancement).
Neocuproine was used for the first time by Brown [15] as Cu(I)-
complexing reagent for ultra-micro sugar determination in blood,
and Dygert et al. [16] modified this assay by using glycine instead
of tartrate in alkaline copper solution, bringing about a substantial
increase in the precision and an extension of the working range of
the assay. Flow injection (FI) and sequential analysis techniques
using the above mentioned colorimetric detections were also de-
veloped, especially instead of the time consuming titration
methods based on Cu(II) reduction [21–23]. Maqueira et al. [21]
developed an automated flow injection analysis (FIA) method for
the determination of reducing sugars in wine based on the clas-
sical Cu(II)-neocuproine reaction, where the samples were deco-
lorized with charcoal. Peris-Tortajada et al. [22] stated that this
method had some problems of heat exchange at the alkaline re-
duction stage, and instead proposed a dialysis procedure avoiding
charcoal, in which the sample was injected into the upper stream
letting reducing sugars pass across the membrane while ob-
structing larger molecular-size colored compounds. Araújo et al.
[23] used the same reaction in a sequential injection (SI) system
for the spectrophotometric determination of reducing sugars in
wines. In this study, the pH of sugar standards was adjusted to 3.4
(using a hydrogen phthalate–hydrochloric acid buffer) to ensure
passage through the dialysis membrane, while alkaline medium
was maintained with 0.075 M NaOH solution as carrier in heated
(to 80 °C) reaction coil for color development.
Spectrophotometric total sugar determination methods are in
current use because of their good sensitivity, low analysis cost,
easy handling and accessible equipments. For example, the 1956
article describing the famous phenol-sulfuric acid colorimetric
method for determining submicro amounts of sugars and related
substances [8] was one of the most read articles of the ACS journal
Analytical Chemistry in 2014. It should be mentioned that the
currently used colorimetric methods of selective sugars assay
suffer from certain disadvantages. For example, the oldest and
most famous phenol-sulfuric acid method requires a high con-
centration of hazardous sulfuric acid and a wide range of analytical
wavelengths for measurement [8]. Poor standardization of the
DNS method [9] with respect to reagent-to-analyte mole ratio,
reaction temperature/incubation time, and reading wavelengths
(ranging from 490 to 580 nm) may have adverse effects on the
accuracy of RS estimation [10]. It is obvious that plant polyphenols
also acting as reducing agents should show an interference in the
determination of reducing sugars with redox-based colorimetric
methods. Xu et al. [24] compared the accuracy of spectro-
photometric DNS along with two enzymatic assays for glucose
measurement in the presence of tea polyphenols (TPLs) against a
HPLC method referred to as the standard control. The results of the
DNS method from a model system containing only glucose and
TPLs showed a TPLs' amount-dependent variation of measured
glucose with a relative error of 5–35%. Thus, the main objective of
present work is to develop a simple and precise spectro-
photometric method for the determination of RS in various food
163
products. With this aim, a method comprised of simultaneous RS-
reduction of Cu(II) to Cu(I) and [Cu(Nc)2] þ colored chelate for-
mation in alkaline medium was utilized. In our study, the inter-
ference of phenolic compounds in fruit juices and honey was
eliminated by a solid phase extraction (SPE) process. Solid phase
extraction is a widely used sample preparation technique for
analyte isolation, concentration, clean-up and medium exchange
[25]. The proposed method was successfully applied to fresh apple
juice, onion juice, commercial fruit juices, honey and milk. The
results were compared with those of the DNS method [9] widely
used for the colorimetric determination of total sugars in various
matrices.
2. Materials and methods
2.1. Chemicals and standards
The following chemical substances of analytical reagent grade
were supplied from the corresponding sources: sucrose, D-
(
)-fructose, D-( þ)-maltose monohydrate, D-lactose mono-
hydrate, (þ )-catechin hydrate, (
)-epicatechin, chlorogenic acid:
Sigma (Steinheim, Germany); 3,5-dinitrosalicylic acid (DNS) and
anhydrous 2,9-dimethyl-1,10-phenanthroline (neocuproine: Nc):
Aldrich (Steinheim, Germany); D-( þ)-glucose, D-( þ)-galactose,
gallic acid monohydrate, potassium sodium tartrate tetrahydrate,
and potassium disulfite: Sigma-Aldrich (Steinheim, Germany);
copper(II) chloride dihydrate: Merck (Darmstadt, Germany); so-
dium carbonate, sodium hydroxide, phenol, L-( þ)-ascorbic acid,
and hydrochloric acid: Riedel-de Haën (Steinheim, Germany).
Discovery DPA-6S (250 mg, 3 mL) and Discovery DSC-18 (1 g,
6 mL) SPE cartridges: Sigma- Aldrich (Steinheim, Germany)
2.2. Instrumentation
Spectrophotometric measurements were made with a Varian
Cary 100 Bio UV–vis spectrophotometer (Sydney, Australia). Cen-
trifugal separations were performed with an Elektro-Mag labora-
tory centrifuge apparatus (Istanbul, Turkey). The pH measure-
ments were made with the aid of a Hanna HI 221 pH-meter using
a glass electrode (Canada). An Elektro-Mag vortex apparatus (Is-
tanbul, Turkey) was used to stir the incubation solutions. A
Memmert water bath was used for incubation. Agilent Vac Elut 12
position manifold (California, USA) and Isolab vacuum pump (Is-
tanbul, Turkey) were used for solid phase extraction (SPE) pro-
cesses. Pure water used for all solutions was obtained from Milli-
pore Simpak1 Synergy 185 (France) ultra-pure water system.
2.3. Reagent and standard solutions
Alkaline Cu(II)-Nc Assay Solutions: CuCl2 solution
(1.0  10
2 M), alkaline solution as 0.5 M NaOH containing 2% (w/
v) Na2CO3 and 0.1 M sodium potassium tartrate were prepared in
distilled water. Neocuproine (Nc) solution (1.5  10
2 M) was
prepared daily in absolute ethanol.
DNS Assay Solutions: 1% (w/v) DNS solution, 40% (w/v) sodium
potassium tartrate solution, 0.2% (w/v) phenol solution and 0.5%
(w/v) potassium disulfite solution were prepared in 1.5% (w/v)
NaOH solution.
Preparation of DNS Reagent: DNS, sodium potassium tartrate,
phenol and potassium disulfite solutions were mixed at 1:1:1:1 (v/
v/v/v) ratio.
All reducing sugar standards and polyphenolic compound
standards as 1.0  10
3 M were freshly prepared in distilled water.
164 K.S. Başkan et al. / Talanta 147 (2016) 162–168
2.4. Analytical methods for the determination of RS
2.4.1. Proposed alkaline Cu(II)-Nc spectrophotometric method
One mL of CuCl2, 1 mL of Nc, (x) mL standard or sample, (1
x)
mL H2O, 1 mL alkaline solution, and 1 mL of sodium potassium
tartrate solution (for the prevention of copper(II) hydroxide pre-
cipitation in alkaline medium) were added to a test tube such that
the total volume was 5.0 mL. The mixture in the stoppered tube
was incubated in a thermostated water bath to determine the
optimum temperature and time. Therefore, this mixture was in-
cubated at varying water bath temperatures (40, 50, 60, 70 °C) for
different time periods (10, 20, 30, 40, 50, 60 min). After the opti-
mal time (20 min) and temperature (60 °C) were specified, the
optimum pH (achieved with 1 mL 0.5 M NaOH solution in 5 mL
final volume) was determined by adding NaOH solution at differ-
ent concentrations. A standard glucose solution was used to build
the calibration curve.
2.4.2. Dinitrosalicylic acid (DNS) method
This method [9] has been selected for comparison because it is
widely used to estimate the RS content of different samples
[2,24,26–28]. In this method, DNS reacts with the free carbonyl
group of RS under alkaline conditions, forming 3-amino-5-ni-
trosalicylate, an aromatic compound with maximum absorption at
540 nm; the optical absorbance is directly proportional to the
amount of RS. In our application, x mL standard or sample, 2 mL of
DNS reagent and (2
x) mL H2O were added to a test tube, and the
mixture was vortexed and incubated in a boiling water bath for
5 min. After cooling in ice bath, the absorbance of sample was
recorded at 540 nm against a reagent blank. Linear regression
equation was obtained using a glucose standard solution.
2.5. Hydrolysis of sucrose
Sugars present in studied food samples are mainly glucose,
fructose and sucrose for fruit juices [29,30], lactose for milk, and
sucrose, glucose and fructose for honey [31]. Lactose, or milk sugar,
is composed of galactose and glucose. Lactose is a RS because it
contains a hemiacetal group, whereas sucrose derived from glu-
cose and fructose is not a RS because of the absence of a hemi-
acetal group in its structure [32].
Disaccharides can be readily hydrolyzed under weakly acidic
conditions, producing their constitutive monomers in equivalent
quantities. Since hydrolysis of each molecule of sucrose yields one
molecule glucose and one molecule fructose [33], it is possible to
indirectly determine sucrose.
Hydrolysis process was performed on 15 mL of 1.0  10
3 M
sucrose or real sample solution, using 0.5 mL concentrated HCl for
10 min at 60 °C. At the end of this time, the hydrolyzate was
neutralized with 2 M NaOH, and diluted to 25 mL, as described
elsewhere [34].
2.6. Preparation of real samples
The apple, onion, commercial fruit juices, honey, soft drink
(cola), UHT whole milk samples were supplied from a local market
of Istanbul.
The apple fruit was washed, peeled and deseeded. The apple flesh
and onion peeled were separately grated with the aid of a glass
grater. The apple and onion juices were separated from the pulps by
filtration using cheesecloth, clarified by passing through a GF/PET
(glass fiber/polyethyleneterephtalate) 1.0/0.45 μm microfilter, and
one milliliter of filtrate diluted to 100 mL with distilled water.
Honey sample was prepared by weighing 1.0 g of honey, made
homogeneous by a certain amount of distilled water and diluted to
100 mL with distilled water.
Milk sample was prepared by adding 2 mL 70% (w/v) tri-
chloroacetic acid (TCA) to 2 mL milk to precipitate the proteins,
followed by centrifugation for 5 min at 4500 rpm. One milliliter of
supernatant was filtered through the GF/PET microfilter, and di-
luted to 100 mL with distilled water.
Commercial fruit juices samples were prepared by filtration
through the GF/PET microfilter, and one milliliter of filtrate diluted
to 50 or 100 mL with distilled water.
2.7. Preparation of synthetic mixtures
Synthetic mixtures (1–3) were prepared using reducing sugar
and certain polyphenolic compound standard solutions. The mix-
ture components were selected on the basis of sugars and phenolic
compounds found in natural samples (milk, fruit and fruit juices).
Also, these sugar mixtures were prepared without polyphenolics
to compare SPE recovery values. They were prepared to contain
standards at the final concentrations declared below;
(1) 3.33  10
5 M glucose, catechin, and chlorogenic acid,
separately.
(2) 2.86  10
5 M glucose, fructose, galactose, maltose, lactose,
catechin and chlorogenic acid, separately.
(3) 1.00  10
5 M glucose, fructose, galactose, maltose, lactose,
ascorbic acid, gallic acid, catechin, epicatechin and chlorogenic
acid, separately.
2.8. Solid phase extraction (SPE) for sample clean-up
C18 and polyamide SPE cartridges were used to eliminate
phenolic compounds and other possible interferents of fruit juices
and honey sample to clean-up sugars. This process was performed
in two stages. At the first step for eliminating phenolic com-
pounds, 4 mL of sample was passed through a C18 SPE cartridge,
previously conditioned with 4 mL 80% MeOH (v/v) and then
equilibrated with 4 mL water. Phenolic compounds (basically fla-
vonoids) in the sample were retained on the column while sugars
passed. The cartridge was then washed with 4 mL of water. The
column effluent and wash water were combined. This solution
contained sugars and other unretained components. At the second
step, the combined solution was passed through a polyamide SPE
cartridge in order to retain other phenolics not held by the C18
column. Thus sugars were set free from all phenolic compounds
after two column separations with respect to polarity differences.
The phenolic compounds can be separately determined (if desired)
by elution from the two SPE cartridges with the use of 80% me-
thanol aqueous solution. All these procedures summarized in Fig. 1
were applied to both natural mixtures and synthetic samples.
2.9. Statistical analysis
Spectrophotometric assays were carried out in triplicate for
each sample and standard. Descriptive statistical analyses were
performed using Excel software (Microsoft Office 2010) for calcu-
lating the mean and the standard deviation of the mean. Results
were expressed as {mean value 7standard deviation (SD)} (n ¼3).
The precision and accuracy of two methods were compared by F-
test and Student's t test. Both methods were applied to five stan-
dard glucose solutions (at the same concentration, 216 mg L
1)
which were prepared from five different stock solutions and four
different samples obtained from the same milk (declared sugar
content, 4.70 g per 100 mL).
 K.S. Başkan et al. / Talanta 147 (2016) 162–168 165

Fig. 1. SPE procedures for sugar analysis in the presence of phenolics.

3. Results and discussion 1

3.1. Analytical method optimization 0.8

Absorbance
To optimize the procedure for the spectrophotometric total 0.6
sugar assay, varying temperatures between 40 and 70 °C and time
intervals between 10 to 60 min were tested. Absorbances obtained
0.4
under both conditions (i.e. 60 °C and 20 min; 70 °C and 10 min)
were very similar (Fig. 2) and seemed to be equally acceptable. But
0.2
the combination of 60 °C temperature and 20 min time was found
more appropriate, because the absorbances measured at 60 °C
0
almost remained stable until the end of the 50th min, i.e. less
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
temperature-sensitive than the other combination. This method
requires neither high acidity like the phenol-sulfuric acid method NaOH Concentration (M)
[8] nor high temperatures such as the DNS method [9] requiring Fig. 3. The absorbance change with NaOH concentration in the Cu(II)-Nc assay. (For
the samples be placed in a boiling water bath for 5 min. The cur- 2.0  10
5 M glucose, for 20 min at 60 °C).
rent method was tested for pH with the addition of NaOH, the
optimal concentration of which was determined as 0.1 M, obtained
Table 1
by using 1 mL of 0.5 M NaOH solution in a final volume of 5 mL Comparison of analytical performance parameters of the proposed (alkaline Cu(II)-
(Fig. 3). Nc) and reference (DNS) methods tested on glucose standards.

Parameter Alkaline Cu(II)-Nc DNS

3.2. Analytical performance
Linear range (mol L
1) 4.0  10
6–3.0  10
5 3.0  10
4–8.0  10
4
Analytical performance data of the proposed and reference Limit of detection (mol L
1)a 6  10
7 8  10
5
Limit of quantitation 2  10
6 3  10
4
methods are depicted in Table 1, showing the higher sensitivity
(mol L
1)b
(high molar absorptivity and low detection limit), wider linear Linear regression equationc y¼ 3.6  104 c þ0.013 y¼ 1.8  103 c
0.402
range and linear correlation coefficient closer to 1.0, and higher Correlation coefficient (R) R¼ 0.9999 R¼ 0.9995
precision (low RSD) of the developed method. The detection limit RSD (%) (n ¼5) 0.6 1.3

is defined by the equation: LOD ¼3sbl/m, where sbl is the standard a

LOD¼ 3 s m
1 (m is slope).
deviation of a blank, and m is the slope of the calibration line. b
LOQ ¼10 s m
1.
c
Since the DNS method had a large intercept value, sbl was ne- y stands for absorbance and c for molar concentration.
cessarily high, shifting LOD values of this method to upper levels.
In addition, the slope (m) of the calibration line of DNS method (i.
1
e. molar absorptivity) was an order of magnitude lower than that
of the Cu(II)-Nc method. The combination of these two factors (sbl
0.8
and m) made the LOD (and also LOQ) values of the two methods
differ by two orders of magnitude (Table 1), rendering the pro-
Absorbance

0.6 40°C
posed method much more sensitive.
50°C
Both methods were applied to certain reducing sugar stan-
0.4 60°C
dards, and the linear regression equations with correlation coef-
70°C
ficients for the proposed and reference methods are presented in
0.2 Table 2.
Comparing the data in Table 2, it was apparent that the alkaline
0 Cu(II)-Nc method yielded higher slopes for linear regression
0 10 20 30 40 50 60
equations (and thus higher analytical sensitivity) than the DNS
Time (min) method for all tested sugars. The linear correlation coefficient was
Fig. 2. The absorbance change with temperature and incubation period in the Cu closer to unity and intercept value considerably lower for the
(II)-Nc assay (For 2.0  10
5 M glucose). proposed method (Tables 1 and 2). It was earlier stated that the
166 K.S. Başkan et al. / Talanta 147 (2016) 162–168

Table 2 Table 4
Linear regression data for certain sugar standards obtained by the alkaline Cu(II)-Nc Glucose equivalent sugar contents of sucrose hydrolyzates and of synthetic mix-
and DNS methods. tures of glucose and fructose (1:1) (prepared at equivalent concentrations to pos-
sible hydrolysis products) determined by alkaline Cu(II)-Nc and DNS methods.
Linear regression equationsa
Glucose equivalent sugar content (mg L
1)
Reducing sugar Alkaline Cu(II)-Nc DNS methodc
methodb Sample Alkaline Cu(II)-Nc Method DNS Method
(mean 7 SD) (mean7 SD)
4 3
D-( þ)-Glucose y¼ 3.6  10 c þ 0.013 y¼1.8  10 c
0.402
(R¼ 0.9999) (R ¼0.9995) Sucrosea hydrolyzate 361 72 (0)c 340 7 1 (
6)
D-(
)-Fructose y¼ 3.4  104 c þ 0.009 y¼2.0  103 c
0.340 (Glucose þFructose)b 350 72 (
3) 3307 2 (
8)
(R¼ 0.9999) (R ¼0.9998) Sucrosea hydrolyzate 706 71 (
2) 7107 8 (
2)
D-( þ)-Galactose y¼ 2.5  104 c þ 0.004 y¼1.8  103 c
0.370 (Glucose þFructose)b 693 71 (
4) 684 7 5 (
5)
(R¼ 0.9999) (R ¼0.9998)
D-( þ)-Maltose y¼ 4.1  104 c
0.003 y¼2.5  103 c
0.307 a
Sucrose concentrations are 1.0  10
3 M (360.3 mg glucose eq. L
1) and
monohydrate (R¼ 0.9999) (R ¼0.9999) 2.0  10
3 M (720.6 mg glucose eq. L
1), respectively.
b
D-Lactose monohydrate y¼ 3.4  104 c þ 0.007 y¼2.8  103 c
0.400 The mixtures were prepared as theoretically equivalent to the hydrolysis
(R¼ 0.9999) (R ¼0.9999) products of sucrose.
c
Error % values were given in paranthesis.
a
y stands for absorbance, c molar concentration, and R correlation coefficient.
b
Linear ranges are 4.0  10
6–3.0  10
5 for all sugar standards
evaluated by applying the proposed and reference methods to the
c
Linear ranges are 3.0  10
4–8.0  10
4 for all sugar standards
hydrolyzate as well as to the possible hydrolysis products of glu-
cose and fructose in a standard mixture. Glucose equivalent sugar
DNS method is not as sensitive as other test methods (e.g., So-
contents of sucrose hydrolyzate and of a synthetic mixture of
mogyi–Nelson, phenol-sulfuric acid, anthrone (9(10H)-anthrace-
glucose and fructose standards are shown in Table 4. The findings
none) and Prussian blue) and therefore the calibration graph ob-
of the proposed method were found very close to the expected
tained for this method could not be plotted on the same scale [28].
values (relative errors between 0 and
4%). These results de-
For statistical comparison of the proposed and DNS methods,
monstrate that the present method can be used accurately for the
sugar contents of glucose standards and milk samples (described
determination of total reducing sugars in samples containing su-
in Section 2.9) were calculated using glucose calibration curves.
Student t- and F-test results are presented in Table 3. It is observed crose hydrolyzate or equivalent constituents (Table 4).
that there was no significant difference in accuracy and precision
of two methods at the 95% confidence level. 3.4. Analysis of synthetic mixtures

3.3. Hydrolysis of sucrose The synthetic mixtures were prepared to assess the inter-
ference effects derived from polyphenolic compounds expected to
As sugars having acetal or ketal linkages are not reducing su- exist in fruit juices and honey, and to evaluate recovery yields of
gars having free aldehyde chains, they do not react with any of the sugars by sequentially applying SPE to samples containing poly-
reducing-sugar test reagents. However, a non-reducing sugar can phenolic compounds. SPE is an increasingly useful sample pre-
be hydrolyzed using dilute hydrochloric acid. After hydrolysis and paration technique, the products of which are widely used for
neutralization of the acid, the products react with the test solu- sample extraction, concentration and clean-up [25,35]. C18 SPE is
tions in the same way as RS, exemplified with a sucrose sample. So also useful for separating sugars, and to some extent, organic acids
it is possible to indirectly determine sucrose. from phenolic compounds in complex samples [36]. Therefore, the
To determine the hydrolysis yield, a certain amount of sucrose alkaline Cu(II)-Nc method was applied twice to the synthetic
was hydrolyzed. Performance of hydrolysis procedure was mixtures before and after the SPE procedure (the steps of which
are shown in Fig. 1) so as to observe the reagent color arising from
Table 3 (sugars þphenolic interferents) versus sugars alone, respectively.
Statistical comparison of alkaline Cu(II)-Nc and DNS methods findings of glucose Also, SPE was applied to lone sugar mixtures which contain the
standards (216 mg L
1) and milk samples (declared sugar content 4.70 g per 100 same type and ratio of sugar components as in synthetic mixtures.
mL) (at the 95% confidence level).
The results (given as μg glucose equivalent per mL sample) and
Sample Parameter Alkaline Cu(II)-Nc DNS recoveries are presented in Table 5.
The synthetic mixtures (1–3) gave positive errors before SPE
Glucose standards No. of samples 5 5 due to the interference of polyphenolics to the proposed assay.
Average 218.4 220.4
However, the recovery values after SPE (between 96 and 104%)
Standard deviation 5.6 5.3
Variance 31.8 28.3 confirmed that RS can be quantitatively separated from common
Degrees of freedom 8
t-test 0.577
Critical value 2.306
Table 5
F-test 1.124
Total reducing sugar content (as μg glucose eq. mL
1) of synthetic mixtures with
Critical value 6.390
respect to the alkaline Cu(II)-Nc method.

Milk samples No. of samples 4 4 Sample Expected Before SPE After SPE
Average 4.04 4.0
a
Standard deviation 0.27 0.25 Syn mix 1 24 957 1 (296) 25 71 (4)
Variance 0.07 0.06 Sugar mix 1 24 25 71 (4)
Degrees of freedom 6 Syn mix 2 103 1697 2 (64) 1027 1 (
1)
t-test 0.204 Sugar mix 2 103 99 71 (
4)
Critical value 2.447 Syn mix 3 72 1567 1 (117) 737 1 (1)
F-test 1.166 Sugar mix 3 72 727 1 (0)
Critical value 9.280
a
Error % values were given in paranthesis.
 K.S. Başkan et al. / Talanta 147 (2016) 162–168 167

Table 6 Table 7
Total reducing sugar content (as g glucose eq. 100 mL
1) of studied samples, in- Recovery values of added glucose or lactose, determined by alkaline Cu(II)-Nc
cluding the declared and found values by alkaline Cu(II)-Nc and DNS methods. method.

Sample Found Declared Sample Sugar content Added Found Recovery (%)
(mg L
1) (mg L
1) (mg L
1)
Alkaline Cu(II)-Nc DNS
Fresh apple juice 31.17 0.6 3.6 35.070.8 108
Fresh apple juice 13.9 7 1.0 11.2 7 1.0
7.2 38.2 70.6 99
a
Onion juice 4.0 7 0.1 3.8 7 0.1
UHT Whole Milk 41.3 7 0.9 7.2 48.4 71.6 99
UHT Whole Milk 4.0 7 0.1 4.0 7 0.2 4.7 14.4 55.9 70.6 101
Grape juice 15.2 7 1.6 11.6 7 0.1 15.2 Apple juice 19.0 7 0.6 3.6 22.4 70.6 94
Apple juice 10.4 7 0.2 10.0 70.1 11.5 7.2 25.9 70.9 96
Orange juice 8.2 7 0.3 (10.17 0.4)a 7.0 7 0.1 (9.07 0.1)a 10.9 Orange juice 17.8 71.0 3.6 21.7 71.0 108
Soft drink (Cola) 10.5 7 0.6 10.5 70.1 10.6 7.2 25.3 70.8 104
Honeyb 70.57 4.2 (72.6 7 2.3)a 69.8 7 0.4 (72.5 7 0.4)a 82.3 Grape juice 22.4 7 1.3 3.6 25.7 70.3 92
7.2 29.8 70.8 103
a
Results found after hydrolysis. Soft drink (Cola) 15.17 0.7 3.6 18.4 71.0 92
b
Honey results expressed as (g glucose eq. 100 g
1). 7.2 21.9 70.6 94
Honey 13.8 7 0.8 3.6 17.3 70.4 97
7.2 21.3 70.6 104
polyphenolic interferents and accurately determined in synthetic
a
samples (Table 5). Experimental findings (not reported) also Added sugar is lactose.

showed that polyphenolic compounds (i.e. flavonoids and other

phenolics) can be effectively separated from sugars and be de-
termined after elution from the two SPE cartridges with 80% me-
thanol aqueous solution.
3.5. Analysis of real samples
Some real samples (such as fruit juices) were especially se-
lected, being rich in polyphenolics which interfere with RS de-
termination methods based on the reducing ability of sugars. In-
terference effects of phenolic compounds were eliminated using
two different SPE cartridges (C18 and polyamide) sequentially. The
findings (as g glucose equivalent per 100 mL or 100 g) and de-
clared amounts of samples are presented in Table 6. Contents
determined by both methods were significantly lower than those
declared for orange juice and honey samples. Hydrolysis process
was also applied to these samples because they may contain su-
crose. As a result, a significant difference was not observed in the
sugar content of honey but the content of orange juice was higher
than the previous value. These values are shown in parentheses in
Table 6. As can be seen from Table 6, there are differences (7–14%)
between the declared and found sugar values for some samples
(milk, orange and apple juices and honey). However, the methods
employed in determining the reducing sugar contents of these
commercial samples were not declared on their packing labels.
Moreover, the declared values may not be very accurate. Results
obtained by the DNS method were lower than those obtained by
the proposed method, reminiscent of a negative bias of the former
that may be attributed to the non-stoichiometric reaction of DNS
with the hemiacetal reducing groups of sugars [27,28]. If HPLC or
enzymatic (especially hexose/kinase) methods were used, accurate
results could be obtained [24]. It should be remembered that the
aim of this study was to develop a highly accurate, rapid and low-
cost spectrophotometric assay for total reducing sugars as an al-
ternative to chromatographic methods (which are time consuming
and more waste generating) and to high-cost enzymatic methods
that may be inhibited in complex samples of an unknown matrix.
So the redox reaction between reducing sugars and Cu(II)-Nc re-
agent was utilized for this aim.
Recovery assessments for alkaline Cu(II)-Nc method were per-
formed by spiking the samples with two known amounts of glu-
cose (with the exception of the milk sample, to which lactose in-
stead of glucose was added), where glucose standard solutions
were added to diluted sample solutions. The total sugar contents
were determined using the calibration curve of glucose and lactose
standard solutions. Recoveries (between 92 and 108%) in Table 7
are satisfactory for RS determination in beverages and honey, and
show that there are no chemical deviations from Beer's law con-
cerning RS determination with the proposed method in complex
matrices. Phenolic compounds in these samples could also be
eluted from the two SPE cartridges with 80% methanol aqueous
solution (recoveries not reported). Recovery experiment of DNS
method was performed only for the diluted fresh apple juice
sample. Two known amounts of glucose standard solutions (0.5
and 0.9 g L
1) were added to sample solution. The added sugar
amounts were determined as 0.5 70.1 and 0.9 70.2 g L
1 using
the calibration curve of glucose standard solutions. Recoveries
(100 720%) showed that essentially the DNS method was also free
from chemical deviations (in conformity to Beer's law) in complex
matrices after SPE clean-up step.
4. Conclusions
The present study showed that the proposed alkaline Cu(II)-Nc
assay, having superior properties such as stable and relatively
nontoxic reagents, simplicity, rapidity and higher sensitivity and
reproducibility than the widely-applied reference DNS method,
can be used to determine the total reducing sugar content of food
samples such as fruit juices, soft drinks and honey. The less sen-
sitivity and reproducibility of the DNS method in comparison to
that of alkaline Cu(II)-Nc may be attributed to the non-stoichio-
metric reaction affinity of DNS toward the hemiacetal reducing
groups of sugars, with the extent of the reaction depending on the
degree of polymerization of the residue to which the reducing
group is attached [27,28]. When the color reaction is non-stoi-
chiometric and/or side reactions accompany the primary reaction,
Beer's law is not strictly obeyed yielding poorer linearity of ab-
sorbance versus concentration. An additional advantage of the
method is that the pre-separated plant polyphenols (via two
successive SPE cartridges) showing an obvious interference effect
on redox-based reducing sugar determination assays [24] due to
their antioxidant properties can be eluted with 80% aqueous
MeOH and determined separately if desired. The method does not
require hazardous chemicals (like sulfuric acid) and does not give
side-reactions yielding a range of wavelengths for the measure-
ment of absorbance maxima (i.e. a single product, Cu(I)-Nc, is
obtained).
168 K.S. Başkan et al. / Talanta 147 (2016) 162–168
Acknowledgments
The authors would like to express their gratitude to the Sci-
entific and Technological Research Council of Turkey (TÜBİTAK-
KBAG-115Z004) for its financial support.
References
[1] M.K. Campbell, S.O. Farrell, Biochemistry, Mary Finch, USA (2012), p. 459.
[2] A. Negrulescu, V. Patrulea, M.M. Mincea, C. Ionascu, B.A. Vlad-Oros, V. Ostafe,
Adapting the reducing sugars method with dinitrosalicylic acid to microtiter
plates and microwave heating, J. Braz. Chem. Soc. 23 (2012) 2176–2182.
[3] C. Melander, E. Andersson, S. Axelsson, Determination of reducing ends with
flow injection analysis with amperometric detection: application to enzyme-
hydrolysed methyl cellulose, Anal. Bioanal. Chem. 387 (2007) 2585–2593.
[4] F. Green III, C.A. Clausen, T.L. Highley, Adaptation of the Nelson–Somogyi re-
ducing-sugar assay to a microassay using microtiter plates, Anal. Biochem. 182
(1989) 197–199.
[5] H. Gautier, V. Diakou-Verdin, C. Bénard, M. Reich, M. Buret, F. Bourgaud, J.
L. Poëssel, C. Caris-Veyrat, M. Génard, How does tomato quality (sugar, acid,
and nutritional quality) vary with ripening stage, temperature, and irradiance?
J. Agric. Food Chem. 56 (2008) 1241–1250.
[6] M. Zarei, M. Azizi, Z. Bashir-Sadr, Evaluation of physicochemical characteristics
of pomegranate (Punica granatum L.) fruit during ripening, Fruits 66 (2011)
121–129.
[7] A.T. Tilahun, Analysis of the effect of maturity stage on the postharvest bio-
chemical quality characteristics of tomato (lycopersicon esculentum Mill.) fruit,
Int. Res. J. Pharm. Appl. Sci. 3 (2013) 180–186.
[8] M. Dubois, K.A. Gilles, J.K. Hamilton, P.A. Rebers, F. Smith, Colorimetric method
for determination of sugars and related substances, Anal. Chem. 28 (1956)
350–356.
[9] G.L. Miller, Use of dinitrosalicylic acid reagent for determination of reducing
sugar, Anal. Chem. 31 (1959) 426–428.
[10] I.P. Wood, A. Elliston, P. Ryden, I. Bancroft, I.N. Roberts, K.W. Waldron, Rapid
quantification of reducing sugars in biomass hydrolysates: improving the
speed and precision of the dinitrosalicylic acid assay, Biomass Bioenergy 44
(2012) 117–121.
[11] C. Gonçalves, R.M. Rodriguez-Jasso, N. Gomes, J.A. Teixeira, I. Belo, Adaptation
of dinitrosalicylic acid method to microtiter plates, Anal. Methods 2 (2010)
2046–2048.
[12] S. Honda, Y. Nishimura, M. Takahashi, H. Chiba, K. Kakehi, A manual method
for the spectrophotometric determination of reducing carbohydrates with
2-cyanoacetamide, Anal. Biochem. 119 (1982) 194–199.
[13] R.F. McFeeters, A manual method for reducing sugar determinations with 2,2′-
bicinchoninate reagent, Anal. Biochem. 103 (1980) 302–306.
[14] S. Waffenschmidt, L. Jaenicke, Assay of reducing sugars in the nanomole range
with 2,2′-bicinchoninate, Anal. Biochem. 165 (1987) 337–340.
[15] M.E. Brown, Ultra-micro sugar determinations using 2,9-dimethyl-1,10-phe-
nanthroline hydrochloride (Neocuproine), Diabetes 10 (1961) 60–62.
[16] S. Dygert, L.H. Li, D. Florida, J.A. Thoma, Determination of reducing sugar with
improved precision, Anal. Biochem. 13 (1965) 367–374.
[17] F.E. Prado, J.A. González, C. Boero, A.R. Sampietro, A simple and sensitive
method for determining reducing sugars in plant tissues. Application quantify
the sugar content in Quinoa (Chenopodium quinoa Willd.), Phytochem. Anal. 9
(1998) 58–63.
[18] N. Nelson, A photometric adaptation of the Somogyi method for the de-
termination of glucose, J. Biol. Chem. 153 (1944) 375–380.
[19] S.V. Ting, Rapid colorimetric methods for simultaneous determination of total
reducing sugars and fructose in citrus juices, J. Agric. Food Chem. 4 (1956)
263–266.
[20] O. Folin, H. Wu, A system of blood analysis, J. Biol. Chem. 38 (1919) 81–110.
[21] A. Maquieira, M.D. Luque de Castro, M. Valcarcel, Determination of reducing
sugars in wine by flow injection analysis, Analyst 112 (1987) 1569–1572.
[22] M. Peris-Tortajada, R. Puchades, A. Maquieira, Determination of reducing su-
gars by the neocuproine method using flow injection analysis, Food Chem. 43
(1992) 65–69.
[23] A.N. Araújo, J.L.F.C. Lima, A.O.S.S. Rangel, M.A. Segundo, Sequential injection
system for the spectrophotometric determination of reducing sugars in wines,
Talanta 52 (2000) 59–66.
[24] H. Xu, X. Leng, M. Wang, G. Zhang, Glucose measurement in the presence of
tea polyphenols, Food Anal. Method. 5 (2012) 1027–1032.
[25] A. Żwir-Ference, M. Biziuk, Solid phase extraction technique-Trends, oppor-
tunities and applications, Pol. J. Environ. Stud. 15 (2006) 677–690.
[26] M. Jeng-Leun, W. Kaun-Tzer, W. Yu-Hsin, L. Yen-Pin, Nonvolatile taste com-
ponents of ear mushrooms, J. Agric. Food Chem. 46 (1998) 4583–4586.
[27] C. Breuil, J.N. Saddler, Comparison of 3,5-dinitrosalicylic acid and Nelson-So-
mogyi methods of assaying for reducing sugars and determining cellulose
activity,, Enzyme Microb. Technol. 7 (1985) 327–332.
[28] S.P. Deng, M.A. Tabatabai, Colorimetric determination of reducing sugars in
soils, Soil Biol. Biochem. 26 (1994) 473–477.
[29] A. Stój, Z. Targoński, Use of sugar content analysis for the estimation of berry
juice authenticity, Pol. J. Food Nutr. Sci. 14 (2005) 403–409.
[30] F.J. Rambla, S. Garrigues, M. De la Guardia, PLS-NIR determination of total
sugar, glucose, fructose and sucrose in aqueous solutions of fruit juices, Anal.
Chim. Acta 344 (1997) 41–53.
[31] M.A. Kamal, P. Klein, Determination of sugars in honey by liquid chromato-
graphy, Saudi J Biol. Sci. 18 (2010) 17–21.
[32] S.L. Seager, M.R. Slabaugh, Carbohydrates, Organic and Biochemistry for To-
day, fourth ed., Brooks/Cole, New York, 2000.
[33] M. Brico-Arias, Glycosides, Synthesis, and Characterization, Springer, New
York, 2007.
[34] V.A.A. Toledo, M.C.C. Ruvolo-Takasusuki, A.J.B. Oliveira, E.D. Chambó, S.M.
S. Lopes, Spectrophotometry as a tool for dosage sugars in nectar or crops
pollinated by honeybees, in: Dr. J. Uddin (Ed.), Macro- to Nano-Spectroscopy,
In Tech, Croatia, 2012, pp. 269–290.
[35] M. Cocchi, C. Durante, M. Grandi, P. Lambertini, D. Manzini, A. Marchetti, Si-
multaneous determination of sugars and organic acids in aged vinegars and
chemometric data analysis, Talanta 69 (2006) 1166–1175.
[36] P. Kylli, Berry phenolics: isolation, analysis, identification, and antioxidant
properties (Academic Dissertation). EKT-series 1502, University of Helsinki,
Department of Food and Environmental Sciences (2011), p. 38–39.