Food Research International 44 (2011) 893 – 896
Contents lists available at ScienceDirect
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / f o o d r e s
a
b
c
a
a
,
⁎
a b c
Department of Food Science, University of Guelph, 50 Stone Road East, Guelph, Ontario, Canada N1G 2W1
Food Division, Brabender GmbH & Co. KG, 51 – 55 Kulturstrasse 47055, Duisburg, Germany
Dipartimento di Scienze Molecolari Agroalimentari, Università degli Studi di Milano, Via Celoria, 2 – 20133, Milan, Italy a r t i c l e i n f o
Article history:
Received 9 November 2010
Accepted 20 January 2011
Keywords:
Gluten Peak Tester (GPT)
Hofmeister series
Gluten aggregation
Gliadin
Glutenin
Flour a b s t r a c t
This study investigated differences in gluten aggregation time and gluten strength using the Hofmeister series in high shear-treated slurries. Two fl ours (15.1% and 10.6% protein) were evaluated by using a Gluten Peak
Tester (GPT). Hofmeister anions including NaF, NaCl, NaBr, NaI, and NaSCN at concentrations ranging from 0.1
to 1.0 M and cations including KCl, NaCl, NH
4
Cl, MgCl
2
, and CaCl
2 at concentrations ranging from 0.0625 to
1.0 M were used. The instrument applies high shear to a fl our/salt solution slurry and measures torque and aggregation time to form gluten. Aggregation time using the GPT followed the order of the Hofmeister series, with minor effects at salt concentration b 0.3 M and increasing differences at higher salt concentrations.
Torque increased with increasing concentration. Creating models of the trends using second and third order equations demonstrated that gluten aggregation follows a distinct natural law in the slurries. The study con fi rmed the potential of the high shear based method to be used a research tool to investigate gluten aggregation properties and to potentially predict functionality in baked product systems.
© 2011 Elsevier Ltd. All rights reserved.
1. Introduction
Gluten is the aggregated product of gliadin and glutenin when wheat fl our is mixed with water. Rheological aspects of gluten aggregation have been studied for many years, with research focusing on the speci fi c interactions that play a role in gluten formation
( Butow, Gras, Haraszi, & Bekes, 2002; Kinsella & Hale, 1984; Preston,
1981
). The salts of the Hofmeister series have also been used to study gluten aggregation and these studies have provided information on the interactions involved in gluten formation. Preston studied the effect of different anions on gluten solubility and aggregation and
reported distinct trends that follow the Hofmeister series ( Preston,
1981
). Kinsella and Hale further observed that different anions at the same concentration greatly affected farinograph consistency and
stability ( Kinsella & Hale, 1984 ).
In dough systems where limited water is available, the effect of ions on gluten is due to their ability to alter water structure in doughs
( Zhang & Cremer, 2006 ). The chaotropic ions are
‘ water structure breakers ’ , causing alterations in water structure that facilitate protein unfolding and solubilization. The kosmotropes have the opposite effect, enhancing water structure to facilitate proteins to remain in their native form through hydrophobic interactions (
Zhang & Cremer,
Abbreviations: GPT, Gluten peak tester; PMT, Peak maximum time; MT, Maximum torque; rpm, Rotations per minute; HMW, High molecular weight.
⁎ Corresponding author. Tel.: + 1 519 824 4120x52204; fax: + 1 519 824 6631.
E-mail address: kseethar@uoguelph.ca
(K. Seetharaman).
0963-9969/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi: 10.1016/j.foodres.2011.01.053
2006
). In doughs, at low salt concentration (0.1
– 0.3 M), all ions have a similar effect on protein aggregation, by shielding electrostatic forces that then allow or prevent protein interaction. At higher salt concentrations ( N 0.3 M), protein behavior is dictated by the salt type and the different effect that the chaotropes and kosmotropes have on water structure (
He, Roach, & Hoseney, 1992; Preston, 1981
).
It is thought that at high salt concentrations, electrostatic forces are negligible, and hydrophobic interactions are responsible for gluten
aggregation behavior ( Preston, 1989
).
Previous studies that have investigated gluten aggregation behavior by using Hofmeister ions have used pre-formed gluten
( Bruun, Sondergaard, & Jacobsen, 2007; Clements, 1973; Preston,
1981; Preston, 1984 ). Other researchers have investigated aggrega-
tion behavior in fl our using large sample size or long mixing time
( Butow et al., 2002; Galal, Varrianomarston, & Johnson, 1978; He et al.,
1992; Kinsella & Hale, 1984; Preston, 1989; Ukai, Matsumura, &
Urade, 2008 ). The present study evaluates the aggregation properties
of gluten in situ using small sample sizes (8 g), high water content slurries (125% water), and high shear (2750 rpm) and the output is recorded as the time to maximum peak (PMT) and maximum torque
(MT).
2. Materials and methods
Two commercial fl ours of different protein content were used; Ibis, a hard wheat blend from western Canada, with a protein content of
15.1% and Diamant, a hard/soft blend, with a protein content of 10.6%.

894 J.P. Melnyk et al. / Food Research International 44 (2011) 893 – 896 create a gluten network. Flour (8 g) and water or salt solution (10 g) were weighed into the sample cup. Sample temperature was maintained at 35 °C by circulating water through the jacketed sample cup. The paddle was set to rotate at 2750 rpm and the torque reading was recorded over time with a peak occurring as gluten aggregates, and the torque falling off when gluten breaks down (
Fig. 1
). Tests ran for a maximum of 10 min whether or not a peak was observed. Salt solutions of the fi ve anions were made at concentrations of 0.1 M,
0.2 M, 0.3 M, 0.4 M, 0.6 M, 0.8 M, and 1.0 M. Cations were tested at
0.0625 M, 0.125 M, 0.25 M, 0.5 M, and 1.0 M. The control for all tests was double distilled water and all runs were done in duplicate. Error bars are included on all graphs and represent the standard deviation.
Fig. 1.
Sample curve produced by Gluten Peak Tester (GPT) software during a test.
During mixing, the paddle does not experience any torque until gluten aggregates. The variables of importance are highlighted: peak maximum time (PMT) and maximum torque (MT). (BE, Brabender Equivalents).
3. Results
3.1. Effect of salt type and concentration on peak maximum time is K
The order of the Hofmeister series from kosmotrope to chaotrope
+ , Na + , NH +
4
, Mg 2+ , and Ca 2+ and F
−
, Cl
−
, Br
−
, I
−
, and SCN
− for cations and anions, respectively. These salt solutions were chosen in order to cover the entire range of the Hofmeister series. Chloride salts
(KCl, NaCl, NH
4
Cl, MgCl
2
, and CaCl
2
) were used to test the effect of
Hofmeister cations, and sodium salts (NaF, NaCl, NaBr, NaI, and
NaSCN) were used to test the effect of anions. All salts were of analytical grade (Fisher Scienti fi c, New Jersey; Merck KGaA,
Germany).
The Gluten Peak Tester (Brabender GmbH and Co KG, Duisburg,
Germany) is an instrument designed to measure the in fl uence of high shear when mixing a high moisture content fl our/water slurry to
The effect of anions and cations of the Hofmeister series on gluten aggregation is evident since Peak maximum time (PMT) and
Maximum torque (MT) followed distinct trends. At low concentration
(0.0625 M to 0.3 M), all salts had a similar effect on PMT, which was signi fi cantly lower than for the control (
Fig. 2 ). As the ionic
concentration increased, ion-related differences in PMT became evident, indicating that each ion had a different effect on the fl our with these effects being more pronounced at higher concentration. At each concentration, changes in PMT followed the Hofmeister series; i.e., decreasing from the most kosmotropic to the most chaotropic species. At salt concentrations greater than 0.3 M, ions at the kosmotropic end of the series resulted in increasing PMT as salt concentration increased; i.e., gluten network formation was delayed.
The opposite effect (faster gluten formation) was observed with
Fig. 2.
Peak Maximum Time (PMT) of Ibis and Diamant in the presence of anionic salt solutions of the Hofmeister series at a range of concentrations (0.1 M, 0.2 M, 0.3 M, 0.4 M, 0.6 M,
0.8 M, and 1.0 M) and in the presence of cationic salt solutions of the Hofmeister series at a range of concentrations (0.0625 M, 0.125 M, 0.25 M, 0.5 M, and 1.0 M). The data shows that at each concentration the time followed the Hofmeister series with kosmotropes increasing aggregation time and chaotropes decreasing aggregation time as concentration increases. All data points have error bars although small and dif fi cult to view.

4. Discussion
J.P. Melnyk et al. / Food Research International 44 (2011) 893 – 896 chaotropic ions, which resulted in decreased PMT with increasing ions concentration. These trends were observed in both Ibis and Diamant fl ours (
Fig. 2
). The graphs do not display data for F
− except for 0.1 M in the Ibis graph. The strong kosmotropic effect of fl uoride at concentrations greater than 0.1 M prevented aggregation and the observation of a peak within the 10 min test time. NaCl and KCl also prevented aggregation within 10 min at 1 M in the low-protein
Diamant fl our due to their strong kosmotropic effect.
3.2. Effect of salt type and concentration on torque
Fig. 3
displays the effect of Hofmeister ions on the maximum torque. Addition of ions resulted in increased torque compared to control. A similar torque value was observed for all ions at the lowest concentration (0.1 M), compared to the larger differences observed at higher salt concentrations. The trends followed the Hofmeister series, with the most kosmotropic anion producing the lowest torque
(weakest gluten network) and the most chaotropic ion producing the greatest torque (strongest gluten network). These trends were observed in both Ibis and Diamant; however NaF, the most kosmotropic ion, only produced a peak at 0.1 M in Ibis, while NaCl and KCl did not produce peaks at 1 M in Diamant within the 10 min test time. When comparing the effect of one salt on both fl ours, Ibis produced a greater torque than Diamant and this trend held true for all fi ve anions.
The Hofmeister series, from kosmotrope to chaotrope, is ordered
K + , Na +
SCN
−
, NH +
4
, Mg for anions (
2+ , and Ca 2+ for cations and F
−
, Cl
−
Baldwin, 1996; Zhang & Cremer, 2006
, Br
−
, I
−
, and
). Ions of the
Hofmeister series clearly affect gluten aggregation in dough
895
according to this order ( Preston, 1989 ). The high shear based
approach using the GPT was able to detect and con fi rm these aggregation trends in slurries.
4.1. Effect of low salt concentration on peak maximum time
At low salt concentration ( b 0.3 M), the difference in aggregation time (PMT) between the salts was much smaller compared to the difference at high concentration. Ibis at 0.3 M had a difference in aggregation time of 150 s (NaSCN = 50 s, NaCl = 200 s), whereas this difference increased to 380 s (NaSCN = 20 s, NaCl = 400 s) at 1.0 M.
The observed trend can be explained by changes in electrostatic interactions in the presence of these ions. At low concentration, ions neutralize charges on amino acids allowing for inter-protein aggregation of hydrophobic surface polypeptides and hydrogen bonding, while preserving the native organization of the gluten proteins (
Fu,
Sapirstein, & Bushuk, 1996; Galal et al., 1978; Ukai et al., 2008
). This was demonstrated by Preston who showed that at low anionic salt concentration (0.05 M), gluten solubility decreased compared to water independent of ion type (
Preston, 1981 ). The low gluten
solubility stems from salt-dependant neutralization of otherwise repelling similar charges on individual proteins, facilitating retention of the native form (
Galal et al., 1978; Preston, 1989
).
4.2. Effect of high salt concentration on peak maximum time
At higher salt concentrations ( N 0.3 M), electrostatic forces become negligible and the type of ion dictates the aggregation behavior. This was highlighted in
Fig. 2
where greater differences in PMT between kosmotrope and chaotrope are observed at high salt concentration compared to low salt concentration. The concentration-dependant increase in PMT above 0.3 M due to kosmotropic anions (F
− and Cl
−
)
Fig. 3.
Maximum torque (MT) of Ibis and Diamant in the presence of anionic salt solutions of the Hofmeister series at a range of concentrations (0.1 M, 0.2 M, 0.3 M, 0.4 M, 0.6 M,
0.8 M, and 1.0 M) and in the presence of cationic salt solutions of the Hofmeister series at a range of concentrations (0.0625 M, 0.125 M, 0.25 M, 0.5 M, and 1.0 M). The data shows that at each concentration the torque followed the Hofmeister series with all salts increasing torque as concentration increased. All data points have error bars although small and dif fi cult to view.

896 and cations (Na + , K + , and NH +
4
) can be explained by the effect these ions have on water structure, and consequently, on gluten aggregation (
Zhang & Cremer, 2006
). The structure-enhancing effects that kosmotropes have on water result in less water availability for gluten which in turn enhances hydrophobic interactions between gluten proteins in their native form. Less unfolding and hydrogen bonding between unfolded chains are experienced. Therefore, more mixing time is required to unravel proteins, and allow for inter-protein interactions through hydrogen bonding and hydrophobic interactions as well as inter-disul fi de bonding and physical entanglements that
contribute to fully developed gluten ( Kinsella & Hale, 1984
). This results in a longer PMT. The opposite effect is evident in the presence of the chaotropic anions (Br
−
, I
−
, and SCN
−
) and cations (Mg 2+ and
Cl + ), resulting in a concentration-dependant decrease in PMT.
Chaotropes, the ‘ water structure breakers ’ , result in disordering of water, allowing faster penetration into the protein structure to aid in more rapid protein unfolding. The unfolding of the protein and disruption of native hydrophobic polypeptide interactions within the core of individual proteins allow for kinetically facilitated interprotein associations through hydrophobic, hydrogen bond, and disul fi
de bond interactions ( Kinsella & Hale, 1984 ). The rapid protein
unfolding and interactions result in a faster PMT.
Some fl our-speci fi c differences can be observed in PMT that are attributed to total protein amount and the gliadin/glutenin fractions in the individual fl ours. At low concentrations ( b 0.3 M), Ibis fl our had a slightly lower PMT (faster aggregation) than Diamant which is attributed to its higher protein content. This results in a slightly faster aggregation time as a consequence of more binding sites being available for intermolecular interactions. Another observation worth noting is that the PMT of both fl ours was similar in the presence of chaotropes but different in the presence of kosmotropes. These differences in PMT suggest that the hydrophobic composition of each of the fl our is different. Preston reported that different Hofmeister ions have the ability to extract different protein fractions and amounts of each fraction (
Preston, 1984
). More recent work highlights the molecular weight distribution and effects of various protein fractions as a determinant for gluten aggregation and dough properties (
Gupta,
Batey, & Macritchie, 1992; Kim & Bushuk, 1995; Southan &
MacRitchie, 1999
). Further studies on the composition of protein fractions within each fl our are required to draw further conclusions as to how varying protein fractions affect GPT readings.
4.3. Effect of salt concentration on torque
J.P. Melnyk et al. / Food Research International 44 (2011) 893 – 896
The presence of all salts resulted in an increasing torque as salt concentration increased. The torque is likely related to the total amount of protein available to aggregate and contribute to a fully developed network once unfolding is complete. Differences between salts within fl ours may be due to the effect that each salt has on the various gluten fractions, while differences between fl ours are due to total gluten. More work must be conducted to determine how different salts and concentrations affect torque.
The gluten aggregation trends due to Hofmeister salts were fi tted to second order and third order equations with the majority of R-squared values being greater than 0.9. This highlights that gluten aggregation follows a pattern that can be predicted using models. Previous gluten aggregation work utilizing the Hofmeister series has not been able to demonstrate this. The models demonstrate the ability to predict gluten aggregation with reliability.
The current study demonstrates the ability to detect changes in gluten aggregation properties and highlights the potential of the high shear-based technique to investigate gluten aggregation properties using small samples sizes in slurries. The GPT highlights differences in gluten aggregation time and relative gluten strengths of slurries simultaneously which has not been previously investigated nor has previously been possible. Further studies will try to determine how
GPT readings can predict protein quality and baking performance according to standard baking tests.
Acknowledgements
Funding from MITACS with ciences is greatly appreciated. Funding from the Grain Farmers of
Ontario is also greatly appreciated.
References fi nancial support from Dow AgroS-
Baldwin, R. L. (1996). How Hofmeister ion interactions affect protein stability.
Biophysical Journal , 71 (4), 2056 − 2063.
Bruun, S. W., Sondergaard, I., & Jacobsen, S. (2007). Analysis of protein structures and interactions in complex food by near-infrared spectroscopy. 2. Hydrated gluten.
Journal of Agricultural and Food Chemistry , 55 (18), 7244 − 7251.
Butow, B. J., Gras, P. W., Haraszi, R., & Bekes, F. (2002). Effects of different salts on mixing and extension parameters on a diverse group of wheat cultivars using 2-g mixograph and extensigraph methods.
Cereal Chemistry , 79 (6), 826 − 833.
Clements, R. L. (1973). Effects of prior salt treatment on gluten dispersibility.
Cereal
Chemistry , 50 (1), 87 − 100.
Fu, B. X., Sapirstein, H. D., & Bushuk, W. (1996). Salt-induced disaggregation/ solubilization of gliadin and glutenin proteins in water.
Journal of Cereal Science ,
24 (3), 241 − 246.
Galal, A. M., Varrianomarston, E., & Johnson, J. A. (1978). Rheological dough properties as affected by organic-acids and salt.
Cereal Chemistry , 55 (5), 683 − 691.
Gupta, R. B., Batey, I. L., & Macritchie, F. (1992). Relationships between proteincomposition and functional-properties of wheat fl ours.
Cereal Chemistry , 69 (2),
125 − 131.
He, H., Roach, R. R., & Hoseney, R. C. (1992). Effect of nonchaotropic salts on fl our breadmaking properties.
Cereal Chemistry , 69 (4), 366 − 371.
Kim, H. R., & Bushuk, W. (1995). Salt sensitivity of acetic acid-extractable proteins of wheatfl our.
Journal of Cereal Science , 21 (3), 241 − 250.
Kinsella, J. E., & Hale, M. L. (1984). Hydrophobic associations and gluten consistency —
Effects of speci fi c anions.
Journal of Agricultural and Food Chemistry , 32 (5),
1054 − 1056.
Preston, K. R. (1981). Effects of neutral salts upon wheat gluten protein-properties.1.
Relationship between the hydrophobic properties of gluten proteins and their extractability and turbidity in neutral salts.
Cereal Chemistry , 58 (4), 317 − 324.
Preston, K. R. (1984). Gelfi ltration and characterization of neutral salt extracted wheat gluten proteins varying in hydrophobic properties.
Cereal Chemistry , 61 (1), 76 − 83.
Preston, K. R. (1989). Effects of neutral salts of the lyotropic series on the physical dough properties of a Canadian red spring wheatfl our.
Cereal Chemistry , 66 (3),
144 − 148.
Southan, M., & MacRitchie, F. (1999). Molecular weight distribution of wheat proteins.
Cereal Chemistry , 76 (6), 827 − 836.
Ukai, T., Matsumura, Y., & Urade, R. (2008). Disaggregation and reaggregation of gluten proteins by sodium chloride.
Journal of Agricultural and Food Chemistry , 56 (3),
1122 − 1130.
Zhang, Y. J., & Cremer, P. S. (2006). Interactions between macromolecules and ions: The
Hofmeister series.
Current Opinion in Chemical Biology , 10 (6), 658 − 663.