Food Chemistry 135 (2012) 827–831

Contents lists available at SciVerse ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Direct potentiometric determination of diastase activity in honey

Milan Sak-Bosnar, Nikola Sakač ⇑
Department of Chemistry, Josip Juraj Strossmayer University of Osijek, F. Kuhača 20, 31000 Osijek, Croatia

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

Article history: A novel method for the determination of diastase activity is reported. The method is based on a direct
Received 12 July 2011 potentiometric measurement of triiodide ion that is released when a starch–triiodide complex is hydro-
Received in revised form 20 February 2012 lysed by honey diastase. The increase of free triiodide ion concentration in a sample is found to be directly
Accepted 2 May 2012
proportional to the diastase activity of the sample.
Available online 11 May 2012
A response mechanism of the platinum redox electrode is proposed, allowing a calculation of the dia-
stase activity factor (F). The sensor and analyte parameters, including F, were obtained by least squares
Keywords:
fitting of potentiometric data using the optimisation function of the Solver add-in of Microsoft Excel. The
Diastase
Triiodide
values of F obtained by the new direct potentiometric method were compared with those obtained using
Platinum the standard Phadebas method (DN values), and the two values were found to agree within experimental
Honey error. Finally, the diastase activity of nine varieties of honey was determined using the novel method
Phadebas developed here.
Direct potentiometry Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction
Honey contains low concentrations of a number of enzymes, the
most prominent of which are diastase, invertase (a-glucosidase),
glucose-oxidase, catalase and acid phosphatase. These enzymes
originate from a number of sources, including nectar and the sali-
vary fluids and pharyngeal gland secretions of honeybees (Huido-
bro et al., 1995). Diastases are a group of starch-digesting
enzymes that include a- and b-amylase. The enzyme a-amylase
hydrolyses starch chains at random locations, producing a variety
of dextrins, and b-amylase splits the reducing sugar maltose from
the end of the starch chain (Crane, 1975). Very low diastase activity
has been shown to indicate that the honey has been exposed to
unfavourably high temperatures. (Schade, Marsh, & Eckert, 1958).
The diastase activity of honey has traditionally been measured
using the classical Schade procedure (Bogdanov & Lischer, 1993;
Hadorn & Zürcher, 1972; Schade et al., 1958; White & Pairent,
1959) or with commercial products, such as the Phadebas amylase
test (Persano Oddo & Pulcini, 1999) and the diastase activity assay
by Megazyme (McCleary & Sturgeon, 2002; American Association
of Cereal Chemists. Approved Methods, 2004).
The Schade procedure is based on the diastatic degradation of
the blue colored starch–triiodide complex (Teitelbaum, Ruby, &
Marks, 1978) under standard conditions. The decrease of the blue
color is spectrophotometrically measured at regular time intervals.
The Schade procedure suffers from many drawbacks. First, mea-
surement in a limited absorbance range can make several dilution
steps necessary to obtain accurate measurements. Second, the pro-
⇑ Corresponding author. Tel.: +385 31495534; fax: +385 31495549.
E-mail address: nikola.sakac@gmail.com (N. Sakač).
cedure for the analysis of samples with low diastase activity can be
incredibly time consuming. Additionally, repetitive measurements
are often needed to increase the accuracy of the determination.
The commercial methods are based on the use of an insoluble
dyed starch substrate (Bogdanov, 1984; Siegenthaler, 1975). Dia-
static hydrolysis of the substrate releases soluble fragments of
the dyed starch into solution. After terminating the reaction, the
remaining insoluble substrate is removed by centrifugation. The
absorbance of the supernatant solution measured at the character-
istic wavelength 620 nm is proportional to the diastase activity of
the sample. These commercial methods are easy to use, but they
are more expensive than the Schade method per sample analysed.
Moreover, if the diastase activity is too high, additional dilution is
required, resulting in a longer time of analysis. Both methods re-
quire a spectrophotometer to measure the response of samples.
The aim of this study was to develop a methodology for deter-
mining diastase activity using direct potentiometric techniques.
The chemical principle that serves as the basis of the analytical
method is based on the measurement of triiodide ion that is re-
leased from a starch-triiodide complex following biocatalytic deg-
radation of the starch by diastase, where a platinum redox
electrode is used as a detector.
2. Experimental
2.1. Reagents and solutions
2.1.1. Preparation of the stock starch solution
The stock starch solution was prepared by dissolving 0.5 g of
dried soluble starch in deionised water in a volumetric flask. After

0308-8146/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.foodchem.2012.05.006
828 M. Sak-Bosnar, N. Sakač / Food Chemistry 135 (2012) 827–831
heating and stirring the solution for approximately ten minutes,
starch was completely dissolved, and the volumetric flask was
filled with deionised water to the mark. A new starch solution
was prepared daily to avoid microbial degradation and
retrogradation.
2.1.2. Preparation of conditioning solution
A conditioning solution was prepared by dissolving solid salts to
provide final concentrations of 6 mM CaCl2 and 20 mM NaCl in
500 mL of a 0.1 M acetate buffer solution (pH 6.0). The addition
of CaCl2 and NaCl to the honey solution serves to stabilise enzyme
activity.
2.1.3. Preparation of honey sample solutions
For this study, nine varieties of honey from Croatia were used:
mandarin, acacia, salvia, linden, honeydew, chestnut, lavender,
mountain lea and Jerusalem thorn. The honey samples were stored
at 20 °C during analysis. Sample honey solutions were prepared by
adding 20 g of each honey variety to a volumetric flask and then
diluting it to the mark with conditioning solution. The resulting
solution was stirred without heating until the honey was com-
pletely dissolved. Honey solutions were used within a few hours
of preparation to avoid self-decomposition and decrease in dia-
stase activity.
2.1.4. Preparation of acetic acid triiodide solution
A potassium triiodide solution was prepared by dissolving solid
iodine to a final concentration of 100 lM and potassium iodide to a
final concentration of 0.05 M in a small volume of water. When the
solid iodine was completely dissolved, acetic acid was added to
yield a final concentration of 1.0 M. Acetic acid triiodide solution
(ATIS) was used as a diastase inhibitor and as a reagent to form
the colored starch–triiodide complex.
2.1.5. Preparation of the standard a-amylase solution
The standard a-amylase solution (0.36 U/mL) was prepared by
dissolving 5 mg of a-amylase (Fluka, Switzerland) in 500 mL of
conditioning solution. The addition of CaCl2 and NaCl to the amy-
lase solution is intended to stabilise enzyme activity over long
periods of time. The decrease in amylase activity due to self-
decomposition was found to be less than 3% after 3 days when
stored in a refrigerator.
2.2. Equipment
Direct potentiometric measurements were performed on an
eDAQ 186 Quad Amp pH/mV amplifier connected to an eDAQ e-
corder 821 8-channel data acquisition system operated by the
eDAQ EChem1.5 software (all from eDAQ, Australia). The system
was equipped with an IJ64 platinum redox electrode (Ionode, Aus-
tralia) and a Ag/AgCl reference electrode (Metrohm, Switzerland).
Spectrophotometric measurements were performed using a Shi-
madzu 1800 UV–vis spectrophotometer (Shimadzu, Japan).
2.3. Procedures
2.3.1. Measurement of the diastase activity of honey samples by direct
potentiometry
A series of test tubes with the same volume of starch solution
were submerged in a temperature-controlled water bath. The buf-
fered honey solution was incrementally added to the starch solu-
tions in test tubes at volumes ranging from 0 to 2.5 mL. The
volume of the reaction mixture was held constant by the addition
of conditioning solution when necessary. After 10 min of heating at
40 °C, the reaction was terminated by adding 5 mL of ATIS. After
allowing the mixture to cool to room temperature, the redox po-
tential was measured using the platinum redox electrode system
described above.
2.3.2. Phadebas honey diastase test
The diastase activity of samples was measured with the Phade-
bas method according to the Harmonised Methods of the European
Commission of Honey (Bogdanov, Martin, & Lüllmann, 1997). A
tablet of an insoluble blue-dyed, cross-linked starch was used as
the substrate for the degradation reaction.
After dissolving 1.00 g of honey in the acetate buffer in a volu-
metric flask, 5.0 mL of the honey solution was transferred to the
test tube and incubated in a water bath at 40 °C for a few minutes.
A blank was prepared by adding 5.0 mL of the acetate buffer solu-
tion and was treated in the same manner as a sample solution.
After placing the Phadebas tablets (Pharmacia Diagnostics AB,
Uppsala, Sweden) into both test tubes, a timer was started. The
tubes were quickly removed from the water bath, stirred and then
returned to the water bath. After 30 min, the reaction was termi-
nated by adding 1.0 mL of 1 M sodium hydroxide solution. The
mixture was stirred again and filtered. The absorbance of the sam-
ple was measured at 620 nm with deionised water as a reference.
The absorbance of the blank was subtracted from that of the sam-
ple solution (DA620). If the absorbance was higher than 1.0, the
sample was diluted, and the dilution factor was taken into account
when calculating the final activity. The measured absorbance of
the solution is directly proportional to the diastatic activity of
the sample.
The diastase activity, expressed as DN or diastase number, was
calculated from the absorbance measurements using Eqs. (1) and
(2) for high (8–40 diastase units) or low (up to 8 diastase units)
activity values, respectively:
DN ¼ 28:2  DA620  2:64 ð1Þ
DN ¼ 35:2  DA620  0:46 ð2Þ
Eqs. (1) and (2) were proposed by the Harmonised Methods of
the European Commission of Honey (Bogdanov et al., 1997). Dia-
stase activity was referred to as DN in the Schade scale, which cor-
responds to the Gothe scale number, or g, of starch hydrolysed per
hour at 40 °C per 100 g of honey. There is a strong correlation be-
tween the diastase activity expressed in Schade units (the tradi-
tional Schade method) and the absorbance measured in the
Phadebas test, so a linear relationship can be used to calculate
the diastase activity in Schade units from the UV–visible
absorbance.
2.3.3. Measurement of a-amylase concentration/activity
Starch hydrolysis by a-amylase was carried out in the test tubes
submerged in the temperature-controlled water bath. The buffered
amylase solution was added to the starch solution to produce a
range of concentrations, and the resulting mixtures were incubated
for 10 min at 40 °C. The volume of the reaction mixtures was held
constant by a periodic addition of conditioning solution. The reac-
tion was terminated by an addition of 5 mL of ATIS. After allowing
the mixture to cool to room temperature, the redox potential was
measured using the platinum redox electrode system described
above.
2.3.4. Optimisation strategy using Solver
Solver is a spreadsheet optimisation modelling system that is
incorporated into Microsoft Excel for Windows, which can be used
for solving various linear, nonlinear or integer problems. When ap-
plied to analysis of a scientific data set, it can yield the same results
as commercial analysis software packages. Solver is activated by
selecting ‘‘Add ins’’ in the Tools menu. Solver was used to compare
an array of data predicted by the model described above using an
 M. Sak-Bosnar, N. Sakač / Food Chemistry 135 (2012) 827–831 829

initial set of parameters over a range of dependent variables with a
set of experimental data. Then, the sum of the squared residuals
(SSR) between the two arrays was calculated by varying the
parameters according to an iterative search algorithm to minimise
the SSR between the theoretical and experimental data sets.
The following steps were used in the Solver optimisation
procedure:
(a) A worksheet was generated containing the data, fit with an
independent variable E (redox potential in mV) and added
volume of honey solution in mL.
(b) A column was included containing Ecalc, which was calcu-
lated by applying Eq. (9) to the results of Eqs. (5) and (13),
to describe the response of the platinum redox electrode to
the increase of triiodide concentration and to include the
appropriate number of parameters to be varied to improve
3.2. Direct potentiometric determination of diastase concentration and
activity
Diastase enzymes catalyse the hydrolysis of starch at biological
conditions. The remaining non-hydrolysed starch, the concentra-
tion of which is inversely proportional to the diastase concentra-
tion, forms a complex with triiodide. This complex formation
reaction reduces the concentration of free triiodide ion in solution.
The change of the concentration of free triiodide ion shifts the
ratio of triiodide to iodide ion in solution, resulting in a shift in
the potential of the redox couple. This shift in the redox potential
of the solution is measured by the redox sensor as shown in Eq. (5).
The measured shift in redox potential can be directly correlated
with the diastase concentration.
Eq. (5) can be modified to show that
0

the theoretical fit. These parameters include the sensor slope E ¼ E0 þ S log cI3 ð6Þ
(S), the system standard redox potential (E0’) and the propor-
where
tionality factor (F). As opposed to determining a very low F
value on the order 105, the log F value is one of the deter- 0
E0 ¼ E0  S log c3I ¼ E0  3S log cI ð7Þ
mined variables. Providing different sets of initial conditions
ensured that Solver was able to determine a global mini- Because of the large excess of iodide present in the analysed solu-
mum of the SSR. tions, its concentration can be assumed to be constant during the
(c) A column was included to calculate the squares of the resid- reaction.
uals, EEcalc, for each data point, and the SSR was calculated Before the addition of the diastase solution, Eq. (6) can be writ-
and displayed in a target cell. ten as
(d) Solver was used to minimise the SSR that was displayed in a 0

target cell by varying the selected parameters of Eq. (9). No E ¼ E0 þ S logðcI3 Þ0 ð8Þ
constraints were applied to the parameters that were varied
where E is the measured electrode potential in mV, S represents the
to minimise the error.
slope of the sensor in units of mV/log unit of concentration, and
The macro, SolvStat (Billo, 2001) provided the regression statis-
tics for Solver by calculating the standard deviations of the param- 420
eters, correlation coefficients of the linear function and standard
errors of the y estimate, SE (y).

3. Results and discussion

380
3.1. Preparation of a standard potassium triiodide solution

Potassium triiodide was prepared by dissolving iodine in a

potassium iodide solution. The formation of triiodide in this solu-
tion occurs according to the reaction equilibrium shown in Eq. (3).
E/mV

340
I2 þ I I3 ð3Þ
A large excess of iodide is required to maintain a constant con-
centration of triiodide in the solution (greater than 100 times the
initial triiodide concentration). For the redox couple in Eq. (4),

I3 þ 2e 3I ð4Þ 300

the corresponding redox potential can be described using the

Nernst equation (Eq. (5)).

 R c I  c I
E¼E þ ln 3 ¼ E þ S log 33 ð5Þ 260
2F c3I c I
0 0,5 1 1,5 2 2,5
V(HS)/mL
where E is the standard reduction potential of the system, S repre-
sents the term, 2303 log RT , cI3 and cI are the concentration of tri- Fig. 1. The response curves of the platinum redox electrode sensor as a function of
2F
varying volume of the 20% honey solutions of nine varieties of honey with a range of
iodide and iodide, respectively, in molarity. The platinum redox diastase activities. Some curves are displaced vertically for clarity. (legend: N,
electrode sensor exhibited a Nernstian response and a high correla- mandarin; d, acacia; D, salvia; x, linden; +, honeydew; j, chestnut; s, lavender; ⁄,
tion of E as a function of log ðcI3 Þ at concentrations at low as 106 M. mountain lea; h, Jerusalem thorn).
830 M. Sak-Bosnar, N. Sakač / Food Chemistry 135 (2012) 827–831

335 10 The measured diastase concentration can be expressed in units

of mass or in an alternate unit related to mass, e.g., the volume of
330 solution containing the known concentration of the diastase sam-
8 ple. In this study, the value of cD was related to the volume of 20%
325
honey solution analysed. The results can also be expressed as dia-
stase activity by performing the calibration procedure using a solu-
6 tion of known diastase activity.
320
By substitution of Eq. (12) into Eq. (10), the following equation
E/mV

RR
is obtained:
315
4
ðcI3 Þf ¼ ðcI3 Þ0 þ F  cD ð13Þ
Experimental
310
Model
Eq. (13) can be rearranged to solve for the unknown diastase con-
RR 2
centration, yielding Eq. (14).
305
ðcI3 Þf  ðcI3 Þ0
300 0
CD ¼ ð14Þ
F
0,0 0,5 1,0 1,5 2,0 2,5
V(HS)/mL
3.2.1. Direct potentiometric determination of diastase activity in honey
Fig. 2. Comparison of experimental (j) and theoretical (D) response characteristics
The diastase activity of samples of nine varieties of honey was
of a platinum redox electrode sensor to exposure to a 20% honey solution. The
values of the squared residuals (RR, s) are shown on the secondary Y-axis. determined. A plot of the response of the platinum redox electrode
sensor as a function of varying volume of the 20% honey solutions
ðcI3 Þ0 represents the initial triiodide concentration in M. After the of each variety of honey is shown in Fig. 1.
addition of diastase at a concentration of cD, the following expres- The samples containing a higher diastase concentration pro-
sion was obtained: duced the highest rate of starch biodegradation, releasing more tri-
0 iodide during the reaction period. The increase of released triiodide
E ¼ E0 þ S logðcI3 Þf ð9Þ
resulted in larger changes of the triiodide-to-iodide ratio, causing
where larger changes of the redox potential of the solutions. Therefore,
the curves that exhibit steeper changes in the measured electro-
ðcI3 Þf ¼ ðcI3 Þ0 þ ðcI3 Þl ð10Þ motive force also exhibit greater diastase activity.
The term ðcI3 Þl represents the increase in triiodide concentra- The experimental data was compared with the theoretical mod-
tion in M. After substituting Eq. (10) into Eq. (9), the following el described by Eq. (9), in which the properties of the sensor and
0

expression was obtained: analyte including sensor slope S, standard redox potential E0 and
h i proportionality factor F were optimised using the minimiser func-
0
E ¼ E0 þ S log ðcI3 Þ0 þ ðcc3 Þl ð11Þ tion of Solver. The comparison of experimental and theoretical re-
sponses of a platinum redox electrode sensor to exposure to a
where the increase in triiodide concentration is proportional to the honey solution is shown in Fig. 2. Solver may be used to determine
concentration of diastase concentration added to the solution: the values of the variables that will minimise the difference be-
ðcI3 Þl ¼ F  cD ð12Þ tween the theoretical and experimental curves.
The results suggest that the calculated theoretical model satis-
where F is a proportionality factor and cD is the diastase concentra- factorily fits the experimental values for the range of diastase
tion in solution. activity levels measured.

Table 1
Selected potentiometric measurements and model parameters for the response characteristics of a platinum redox electrode sensor to exposure to a range of volumes of 20%
honey solution. (ATIS, Acetic acid triiodide solution; HS, honey solution; ssr, sum of squared residuals; res, residuals; sr, squared residuals for the experimental data, S, sensor
0
slope;E0 , standard redox potential; F, proportionality factor; Ecalc , calculated model redox potential; E, experimental redox potential; CðI 3 Þf , free triiodide concentration
(Eq. (10))’’.

Model parameter
Starch cS(g/L) 5.00 S 38.41
0
VS/mL 2.00 Eo 429.38
F 2.705E-05
Honey solution w (HS) 20% log F 4.57
F  106 27.05
ATIS VATIS/mL 5.00
c(I3) = 1.00E-04 ssr 1.58
V (HS)/mL E/mV C (I) C (I3)f Ecalc res sr
0 308.1 0.02632 1.82E-05 307.98 0.12 0.02
0.25 313.4 0.02632 2.49E-05 313.26 0.14 0.02
0.5 316.6 0.02632 3.17E-05 317.26 0.66 0.44
0.75 320.4 0.02632 3.84E-05 320.49 0.09 0.01
1 323.1 0.02632 4.52E-05 323.19 0.09 0.01
1.25 325.8 0.02632 5.20E-05 325.51 0.29 0.08
1.5 328.3 0.02632 5.87E-05 327.55 0.75 0.56
1.75 329.1 0.02632 6.55E-05 329.37 0.27 0.07
2 331.4 0.02632 7.23E-05 331.01 0.39 0.15
2.25 332.4 0.02632 7.90E-05 332.50 0.10 0.01
2.5 333.4 0.02632 8.58E-05 333.87 0.47 0.22
 M. Sak-Bosnar, N. Sakač / Food Chemistry 135 (2012) 827–831
Fig. 3. Comparison of the results obtained using the Phadebas method (DN) and
with the direct potentiometric method described here (F). Diastase concentrations
(cD) expressed as lg of a-amylase per gram of honey are shown on the secondary
Y-axis. Sample key: 1, lavender; 2, honeydew; 3, chestnut; 4, linden; 5, Jerusalem
thorn; 6, acacia; 7, mandarin; 8, salvia; 9, mountain lea honey.
A section of the spreadsheet that displays the potentiometric
measurements and the model parameters for the response charac-
teristics of a platinum redox electrode sensor to exposure to the
range of volumes of 20% honey solution is shown in Table 1.
The sensor and analyte parameters obtained by modeling
potentiometric data using the optimization procedure with the
Solver function and the corresponding regression statistics ob-
tained from SolvStat are following (values ± standard deviation):
Slope = 38.4 ± 2.9 mV/decade, = 429.4 ± 9.5 mV, log(F) = 27.05 ±
4.2, R2 = 0.9977, SE (y) = 0.445.
The determined diastase activity values in different varieties of
honey expressed as F-values are compared with the DN values ob-
tained using the Phadebas method. Diastase concentration ex-
pressed as lg of a-amylase per gram of honey was also
determined. The sensor detection limit was found to be
0.054 lg (1.944 mU). The resulting concentrations from all three
determination methods are shown in Fig. 3.
It can be concluded that a good correlation exists between the
two methods. The differences between the varieties of honey that
were obtained using the Phadebas method follow the differences
that were obtained by means of a direct potentiometric method.
The differences between the same varieties of honey that were ob-
tained by the direct potentiometric method were greater than
those that were obtained using Phadebas method, indicating im-
proved sensitivity using the direct potentiometric method.
The differences between the results expressed as lg of a-amy-
lase per gram of honey agree with the values determined by both
the Phadebas and the direct potentiometric method.
4. Conclusions
The method described above made use of a direct potentiomet-
ric determination of diastase activity by measuring the triiodide
831
ion released during biocatalytic degradation of starch with a plat-
inum redox electrode sensor as the detector. It makes use of sim-
ple, low-cost instrumentation and is convenient for fast, reliable
determinations of diastase activity in samples. The results were
compared with those obtained using the Phadebas method, which
is usually employed for measuring the diastatic activity of honey
samples. A good correlation between the results obtained from
both methods was observed. Additionally, the mechanism of the
sensor response for the technique was proposed. The sensor and
analyte parameters were obtained by optimising a theoretical
model using the Solver add-in of Excel.
When compared with other methods, the proposed analytical
method exhibits numerous advantages. First, the manipulation of
solutions and dilutions are minimised. No additional dilutions
must be made, since there is only a fixed volume of honey required
for each analysis. Second, the incubation time is fixed at 10 min,
and no additional time must be spent for each analysis.
Finally, the diastase activity of nine varieties of honey has been
determined using the developed method.
Acknowledgements
The authors gratefully acknowledge the financial support of the
Croatian Ministry of Science, Education and Sports, given to Project
No. 291-0580000-0169.
References
American Association of Cereal Chemists. Approved Methods (2004). Method 22–05.
Measurement of a-amylase in cereal grains and flours – Amylazyme method.
First Approval November 3rd, 1999.
Billo, E. J. (2001). Excel for Chemists ((3rd ed.)). New York: Wiley (Chapter 12).
Bogdanov, S., & Lischer, P. (1993). Interlaboratory trial of the International Honey
Commission: Phadebas and schade diastase determination methods, humidity
by refractometry and invertase activity: Report for the participants.
Bogdanov, S. (1984). Honigdiastase, Gegenüberstellung verschiedener
Bestimmungsmethoden. Mitteilungen Gebiete Lebensmittelsicherheit Hygiene,
75, 214–220.
Bogdanov, S., Martin, P., & Lüllmann, C. (1997). Harmonised methods of the European
honey commission. Apidologie, (extra issue), 1–59.
Crane, E. (1975). Honey: A comprehensive survey. International Bee Research
Association, London: Heinemann.
Hadorn, H., & Zürcher, K. (1972). Eine einfache kinetische Methode zur Bestimmung
der Diastasezahl in Honig. Deutsche Lebensmittel Rundschau, 68, 209–216.
Huidobro, J. F., Santana, F. J., Sanchez, M. P., Sancho, M. T., Muniategui, S., & Simal-
Lozano, J. (1995). Diastase, invertase and alpha-glucosidase activities in fresh
honey from north–west Spain. Journal of Apicultural Research, 34, 39–44.
McCleary, B. V., & Sturgeon, R. (2002). Measurement of a-amylase in cereal, food
and fermentation products. Cereal Foods World, 47, 299–310.
Persano Oddo, L., & Pulcini, P. (1999). A scientific note on the Phadebas method for
honeys with low enzyme content. Apidologie, 30, 347–348.
Schade, J. E., Marsh, G. L., & Eckert, J. E. (1958). Diastase activity and
hydroxymethylfurfural in honey and their usefulness in detecting heat
adulteration. Food Research, 23, 446–463.
Siegenthaler, U. (1975). Bestimmung der Amylase in Bienenhonig mit einem
handelsublichen, farbmarkierten Substrat. Mitteilungen Geb
Lebensmittelsicherheit Hygiene, 66, 393–399.
Teitelbaum, R. C., Ruby, S. L., & Marks, T. J. (1978). On the structure of starch–iodine.
Journal of American Chemical Society, 100, 3215–3217.
White, J. W., & Pairent, F. W. (1959). Report on the analysis of honey. Journal of the
Association of Offshore Agricultural Chemists, 42, 341–348.