food and bioproducts processing 9 6 ( 2 0 1 5 ) 20–26

Contents lists available at ScienceDirect

Food and Bioproducts Processing

journal homepage: www.elsevier.com/locate/fbp

Isolation of quinoa protein by milling fractionation

and solvent extraction

Maike Föste, Dana Elgeti, Anna-Katharina Brunner,

Mario Jekle ∗ , Thomas Becker
Technische Universität München, Institute of Brewing and Beverage Technology, Research Group Cereal Process
Engineering, Weihenstephaner Steig 20, 85354 Freising, Bavaria, Germany

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

Article history: Quinoa is an attractive non-animal protein source, because of its absence of gluten and the
Received 13 March 2015 favorable amino-acid profile. The objective of this study was the extraction of quinoa protein
Received in revised form 1 June 2015 by two consecutive approaches. Initially, the protein-rich bran was isolated by optimizing
Accepted 15 June 2015 the conditioning parameters of a milling fractionation procedure. In contrast to the protein
Available online 23 June 2015 content of the whole grain, 11.75% in dry base (db), the bran fraction contained 27.78% (db)
protein when conditioned with 15% moisture at 20 ◦ C for 20 h. Subsequently, an aqueous
Keywords: extraction was developed, resulting in a protein solubility of 60% at pH 10 and 20 ◦ C for
Pseudocereals 1 h. For purification at the isoelectric point, the pH-value and the separation method were
Bran varied. After drying, an extraction of proteins from bran yielded 68% (db) as compared to
Particle size 52% (db) protein from quinoa whole grain flour.
Protein concentration © 2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Solubility
Gluten-free

1. Introduction Protein from these sources can be isolated by solvent

extraction to serve as an ingredient in dietary supplements.
The uptake of proteins is one of the major factors for main- Additionally, for gluten-free products the application of pro-
taining human health. Because of the higher costs for animal tein isolates can be beneficial for processing and quality
protein production and the huge amount of resources which improvement of bread, pasta, dietary supplements, baby
are required, plant-based protein sources are of growing inter- foods and beverages. According to Segura-Nieto et al. (1999)
est. In addition, the incidence of allergies and intolerances and Gorinstein et al. (2002) proteins of amaranth, soybean,
concerning proteins from egg or milk has been rising. Cur- buckwheat, and quinoa are highly soluble and applicable in
rently, cereals and legumes represent the main alternatives, functional foods.
for both human and animal nutrition. However, the alcohol The type of milling product, processing parameters and
soluble gliadins and glutenins in wheat are attributed to the suitable solvents have to be taken into consideration to
celiac disease (Thompson, 2001). Thus, the identification of achieve a maximum yield. Previously, a lot of studies have
novel and sustainable protein sources is of great relevance. focused on the extraction of proteins from whole grain flours
The pseudocereal quinoa might be a suitable candidate, since for e.g. amaranth (Salcedo-Chávez et al., 2002), while only
it features high protein and mineral content with a well bal- few reported extraction from byproducts such as rice bran
anced amino acid profile and lacks gluten (Aufhammer et al., (Jiamyangyuen et al., 2005). The separation of different grain
1995). tissues enables the selection of a fraction which is naturally

Abbreviations: AACC, American Association of Cereal Chemists; db, dry base; ICC, International Association for Cereal Science and
Technology; CIE, Commission Internationale d’Eclairage.
∗
Corresponding author. Tel.: +49 8161 71 5327; fax: +49 8161 3883.
E-mail addresses: [email protected] (M. Föste) , [email protected] (M. Jekle).

http://dx.doi.org/10.1016/j.fbp.2015.06.003
0960-3085/© 2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
 food and bioproducts processing 9 6 ( 2 0 1 5 ) 20–26 21

rich in target substances. The germ of quinoa, for example, Switzerland). Due to the mechanical pretreatment (involv-
contains most of the protein (Valencia-Chamorro, 2003), while ing washing and friction) by the manufacturer, the quinoa
starch is present in negligible levels. Additionally, protein bran fraction in this work was lacking the pericarp. Quinoa
can be solubilized more easily when less starch is avail- seeds were ground to whole grain flour in an ultra-centrifugal
able, because especially small starch granula are difficult to mill Retsch ZM 200 (Haan, Germany) with a mesh screen of
separate. Since amaranth and quinoa starch granula are par- 500 ␮m. Prior to fractionation, the moisture content of whole
ticularly small (Aufhammer et al., 1995) typical separation grain flour was determined thermo-gravimetrically accord-
methods based on gravimetric or size differences are difficult ing to AACC standard method 44-01 (American Association of
to realize. Cereal Chemists (AACC), 2002). Conditioning was carried out
Nevertheless, in the case of quinoa and amaranth, milling in an airtight box and the initial moisture content of 12.3% was
presents a technological challenge because of their small ker- elevated to 14%, 15% and 16% by spraying water onto the seeds,
nel sizes: for quinoa 1.0–2.5 mm (Taylor and Parker, 2002) followed by manual stirring. Resting time lasted 20 h and was
and for amaranth 1.0–1.5 mm (Bressani, 2003), respectively, shortened to 16 h for 15% seed moisture samples at room tem-
depending on the plant variety. Thus, only few studies have perature (20 ◦ C). Milling trials were performed in a Brabender
focused on the production and application of milling fractions Quadrumat Junior mill (Duisburg, Germany), which is a labo-
(Chauhan et al., 1992; Ando et al., 2002). Chauhan et al. (1992) ratory roller mill, separating the quinoa seeds into bran and
compared two methods for reducing the saponine content white flour by sieving in a rotating sifter (mesh of 200 ␮m).
by a pretreatment before milling the grains into whole grain The milling yield as percentage was determined by dividing
flour, bran and white flour. However, there are no available the mass of the respective fraction by the total mass of milling
data, focusing on the duration and the water content during a products.
conditioning step. Similarly US patent 20100184963 from 2010
suggests the use of protein-rich fractions for the production 2.2. Quantitation of protein, ash and moisture
of quinoa protein concentrate (Scanlin et al., 2010). Scientific
data and validation of the proposed methods are not available. Milling fractions were analyzed for their moisture and ash
In the case of wheat and other typical cereals, conditioning is content by employing standard methods of analysis, AACC
said to be one of the most important steps prior to grinding, 44-01 and 08-12, respectively (American Association of Cereal
which renders the outer cell tissues more elastic, resulting in a Chemists (AACC), 2002). The crude protein content was
better separation from the endosperm. Depending on the mor- determined by the Kjeldahl procedure according to standard
phology of wheat kernels, the target moisture content during method AACC 46-10 using a conversion factor of N × 5.54
conditioning varies between 15% and 16% with a resting time (American Association of Cereal Chemists (AACC), 2002;
between 6 and 36 h (Cauvain, 2003; Fang and Campbell, 2003). Fujihara et al., 2008). Each test was performed in quadruplicate
After choosing a suitable milling fraction for protein con- and expressed on dry weight basis (db).
centration, an extraction procedure has to be developed. Color characterization of milling fractions was measured
Salcedo-Chávez et al. (2002) varied processing parameters for using a Spectrophotometer fitted with an optical sensor from
alkaline extraction from amaranth whole grain flour, deter- BYK Gardner GmbH (Geretsried, Germany) on the basis of the
mining solubilization at pH 8 to 11 and acidic precipitation at CIE L* , a* , b* color system as describe by Edney et al. (2002)
pH 4.5 to 5.0 as optimum conditions. Depending on the extrac- The instrument was calibrated using a white standard color
tion conditions, physicochemical changes of the proteins may calibration plate. The brightness, indicated by L* (0 = black,
affect their intended functionality and bioavailability. Scilingo 100 = white) was recorded. Three measurements from three
et al. (2002) analyzed the effect of extraction and precipitation different samples were performed.
conditions on the electrophoretic and calorimetric behavior
of amaranth proteins. Results from this study indicated that 2.3. Particle size distribution of milling fractions
pH-values >9 and <5 reduced thermal stability and increased
protein denaturation. Similarities regarding protein solubil- The particle size distribution of flour was measured accord-
ity of amaranth and quinoa can be expected, because of ing to ICC standard method No 207 (International Association
the same family (Amaranthaceae) and a related seed struc- for Cereal Science and Technology (ICC), 1998). Milling prod-
ture. ucts (each 100 g) were sifted in a sieving chamber from Bühler
The aim of this study was to optimize a fractionation pro- GmbH (Braunschweig, Germany) with a set of graded standard
cess in order to separate quinoa flour from the bran fraction. sieves (45 ␮m, 90 ␮m, 125 ␮m, 180 ␮m, 250 ␮m, 355 ␮m, 500 ␮m,
Second, proteins of the bran fraction were further concen- 1000 ␮m, 1250 ␮m) for 10 min in the case of flours and for 5 min
trated by optimizing an extraction process, aiming to reach when sieving bran. The retained product on each sieve was
a protein content between 50% and 90%. Different pH condi- weighed and expressed as percentage retention. Three mea-
tions for extraction were analyzed and the resulting solution surements from three different samples were conducted.
was further purified and dried to obtain a stable product which
can be used as dietary supplements and functional food ingre- 2.4. Extraction of proteins from quinoa bran
dients.
After enriching the proteins in the bran fraction through
conditioning, milling and sieving, further concentration was
2. Materials and methods conducted by solvent extraction followed by an optional
purification procedure as described in Fig. 1. Proteins were
2.1. Origin and fractionation of quinoa solubilized (see Fig. 1a) by preparing a suspension of quinoa
bran and water in a ratio of 1:10 (w:v). The impact of the
Organic Royal Quinoa grains (Chenopodium quinoa) from Bolivia pH-value on protein solubility was determined by adding 1N
were purchased from Ziegler & Co. GmbH (Wunsiedel, NaOH to obtain a pH-range from 7 to 12 according to the
22 food and bioproducts processing 9 6 ( 2 0 1 5 ) 20–26

Fig. 1 – Solubilization (a) and purification (b) of quinoa bran proteins. Varied parameters are listed below the arrows with
the variation range given in brackets.

method of Bera and Mukherjee (1989). Furthermore, the dura- with LMC-1, Martin Christ Gefriertrocknungsanlagen GmbH,
tion of the extraction process, the temperature as well as the Osterode, Germany) for a minimum of 48 h reaching a mois-
particle size of the raw material (quinoa bran) were varied ture content below 5% before storage at −20 ◦ C until further
in order to determine the conditions yielding in the maxi- analyses.
mum protein solubility. The suspensions were stirred on a
shaker for 1 h, 2 h, 3 h or 4 h at 20 ◦ C, 30 ◦ C, 40 ◦ C, or 50 ◦ C. 2.6. Statistical analysis
In order to vary the particle size, the bran was milled in an
ultracentrifugal mill Retsch ZM 200 (Haan, Germany) with dif- Statistical analysis was performed with the software Prism
ferent mesh sizes (750 mm, 500 ␮m, 250 ␮m). Separation of 5 (Version 5.03, GraphPad Software, Inc.) on all data. To
soluble proteins from the insoluble residue was performed by detect significant differences between means, one-way anal-
centrifugation at 4500 × g for 20 min at 4 ◦ C. Solubility of pro- ysis of variance (ANOVA) with separation of means by the
teins from quinoa bran was calculated according to Formula Tukey–Kramer test was applied with a significance level of
(1) and expressed as the percentage of crude protein (N × 5.54) p < 0.05.
present in the supernatant (mp,sup. ) in relation to the total
protein content of bran (mp,bran ) (Paredes-López et al., 1994).
3. Results and discussion
Determination of protein nitrogen was performed as for the
milling fractions in Section 2.2.
3.1. Impact of conditioning on yield and protein
mp,sup. content of milling fractions
Solubility (%) = × 100 (1)
mp,bran
The results of varied conditioning settings are presented in
2.5. Purification and drying of protein extracts Table 1. Separation of unconditioned seeds resulted in a yield
of nearly 71% quinoa white flour and 28% bran. With regard to
The preparation of quinoa bran protein isolates was conducted 14% moisture, the yield of quinoa white flour was elevated to
by using different pH values for extraction and precipitation a maximum of 68%. The higher moisture content led to bet-
according to Paredes-López et al. (1994). For protein purifica- ter separation of seed tissues, resulting in significantly higher
tion (see Fig. 1b), supernatants from the solubilization step protein content in bran (26.18% db). By elevating the seed
were adjusted in the pH-value range of 2 to 6 with 1 N HCl. moisture up to 15%, the protein content reached 27.78% db,
Precipitated proteins were centrifuged at 4500 × g, 9000 × g or which represents an enrichment of over 50% (rel.) in compar-
15,000 × g at 4 ◦ C for 20 min. Precipitates were re-suspended ison to whole grain flour. Moreover, the ash content of the
in water, neutralized to pH 7 and freeze-dried (Beta 2–8 white flour was 0.63% db, indicating that most of the bran

Table 1 – Impact of seed conditioning on milling yield, moisture, ash and protein content of quinoa milling fractions.
Milling fraction Conditioned moisture (%)a Resting time (h) Yield (%) Moisture (%) Ash (% db) Protein (% db)b

Whole grain flour – – – 12.25 ± 0.03b 2.33 ± 0.03d 11.75 ± 0.12c

White flour – 0 71.27 ± 1.19f 12.26 ± 0.10b 1.47 ± 0.08c 9.18 ± 0.54b
Bran – 0 28.18 ± 1.16a 10.24 ± 0.20a 4.70 ± 0.17e 23.97 ± 1.78d
White flour 14.0 20 68.19 ± 0.06e 14.06 ± 0.03d 0.91 ± 0.03b 6.33 ± 0.11a
Bran 14.0 20 31.55 ± 0.08b 12.04 ± 0.05b 5.22 ± 0.06f 26.18 ± 1.04e
White flour 15.0 20 64.43 ± 0.55d 14.71 ± 0.08e 0.63 ± 0.01a 5.86 ± 0.53a
Bran 15.0 20 35.29 ± 0.57c 12.93 ± 0.40c 5.44 ± 0.08g 27.78 ± 1.10f
White flour 15.0 16 62.78 ± 0.38d 14.91 ± 0.07e 0.85 ± 0.01 5.96 ± 0.25a
Bran 15.0 16 35.97 ± 0.01c 13.24 ± 0.26c 5.24 ± 0.07f 25.82 ± 0.49e

Mean values of two independent milling trials with standard deviation: yield (n = 2) moisture, ash and protein content (n = 4). Different letters
denote statistically significant differences (ANOVA, p < 0.05). Protein and ash content calculated on dry base.
a
Conditioned moisture amount added before milling.
b
Determined with Kjeldahl (N × 5.54).
 food and bioproducts processing 9 6 ( 2 0 1 5 ) 20–26 23

particles were separated. Higher amounts of conditioning

Table 2 – Particle size distribution and protein content of
water (16%) were not applicable since the high moisture con- quinoa milling fractions in comparison to gluten-free
tent caused an obstruction of particles between the milling flours.
rolls. In comparison, Chauhan et al. (1992) obtained bran with
Sizea (␮m) Whole grain White flour Bran (%)
a protein content of 20.4% and 24.3% (db) after using different flour (%) (%)
dehulling procedures, respectively. The authors chose a con-
version factor of 5.7 and the native whole grain had a higher >0 3.93 ± 1.43 8.40 ± 3.44 5.06 ± 0.11
>45 28.19 ± 0.68 43.27 ± 2.90 –
protein content (13.7% db). Generally, literature uses different
>90 11.16 ± 0.38 11.13 ± 0.96 –
conversion factors for calculating the amount of crude pro- >125 13.36 ± 0.62 27.27 ± 0.29 10.34 ± 0.28
tein from the total nitrogen determined via Kjeldahl or Dumas >180 15.83 ± 1.44 6.33 ± 0.23 15.95 ± 0.83
compromising the comparability. To sum up, conditioning to >250 15.83 ± 0.11 2.50 ± 0.20 20.03 ± 0.46
15% moisture for 20 h improved the separation of the bran tis- >355 12.46 ± 0.59 0.00 ± 0.00 23.24 ± 0.12
sues from the starch rich perisperm significantly. Whereas the >500 0.00 ± 0.00 – 24.47 ± 0.27
>1000 – – 2.73 ± 0.10
protein content in whole grain flour was 11.75% db, this was
concentrated up to 2.3 fold in the bran fraction. a
Particles with different sizes were classified by weight after siev-
ing. Means are presented with standard deviation (n = 3).
3.2. Correlation of ash and protein content with the
coloring of fractions particular importance for the extractability of components or
other applications. As an example, de la Hera et al. (2013)
The coloring of flour is often used to evaluate its quality observed that coarser maize flour particles resulted in higher
and pureness. Especially the whiteness, which is composed loaf volume and Raghavendra et al. (2006) found the hydration
of brightness and yellowness, is of interest for characterizing properties of coconut residue were affected by their particle
milling products. However, for untypical grains like quinoa the size.
predictability of the pureness by these values is not guaran- The particle size distributions of the quinoa milling frac-
teed because of considerable differences regarding the seed tions are presented in Table 2. Due to the milling of quinoa
structure. According to Oliver et al. (1993) the CIE LAB color whole grain flour in an ultra-centrifugal mill with a sieve of
space co-ordinates L* and b* are useful for measuring bright- 500 ␮m mesh size, flour particles varied between 45 to 355 ␮m.
ness and yellowness of wheat flour. The resulting brightness As expected quinoa white flour consisted of considerably
is influenced largely by the particle size and bran inclusion smaller particles (>50% below 90 ␮m) than bran (>50% above
effects. In Fig. 2 the brightness of milling fractions is plot- 355 ␮m). In comparison to wheat bran, which can achieve a
ted against ash (see Fig. 2a) and protein content (see Fig. 2b) mean size of 1239 ␮m (Stewart and Slavin, 2009), the parti-
to determine correlations. Indeed, a correlation between the cle size of quinoa bran was smaller. This is not surprising,
L* value and the ash content (R2 = 0.9909, p < 0.0001) was considering the different grain sizes. Since the quinoa germ
observed. By elevating the seed moisture content, flour got is surrounding the starch-rich perisperm, a maximum length
purer as indicated by lower ash content and higher L* val- of 3 mm can be expected if the germ is pulled straight (for
ues. Similarly, the protein content correlated with the L* value ∼1.5 mm grain diameter).
(R2 = 0.9898, p < 0.000.1). Thus, it can be concluded that the
brightness measured with an LAB system can be used to deter- 3.4. Protein extraction by solubilization
mine the pureness of quinoa milling fractions.
The solubility of proteins depends on the physicochemical
3.3. Particle size distribution of quinoa milling state of the molecules, which can be favorably or adversely
products affected by processing treatments such as heating or dry-
ing during processing and storage. Further factors are the
The particle size distribution of milling fractions can be ratio of polar to non-polar groups as well as the primary,
of interest for technological applications, because of differ- secondary, tertiary and quaternary structure of the protein.
ences regarding the reactive surface of particles, which is of In alkaline solutions the conformation alters, exposing more

Fig. 2 – Correlation between ash (a) and protein content (b) with the brightness of quinoa milling fractions. Deviations regarding
ash and protein content of quinoa bran and white flour derive from different conditioning procedures before milling. Whole
grain flour was produced in an ultracentrifugal mill. L* values represent the brightness. Db: dry base. Plotted are means
with standard deviation (n = 6). Dotted lines represent linear correlations of the data with R2 = 0.9909 (p < 0.0001) for the ash
content (a) and R2 = 0.9898 (p < 0.0001) for the protein content (b).
24 food and bioproducts processing 9 6 ( 2 0 1 5 ) 20–26

Fig. 3 – Impact of pH-value and particle size on the protein solubility of quinoa bran. (a) Solubility: g proteins in solvent per
100 g proteins in bran. (b) Solubility of proteins at pH 10 depending on the particle size, as obtained after centrifugal milling.
The protein content of the supernatant was measured after centrifugation at 4500 × g (N = 5.54). Ref: Bran without milling.
Means are presented with standard deviation (n = 4). Different letters denote statistically significant differences (ANOVA,
p < 0.05).

hydrophobic groups, which promotes higher protein solubility result, there was no significant variation regarding the pro-
by increasing the interactions with water (Aceituno-Medina tein solubility after varying the duration of the extraction (see
et al., 2013; Salcedo-Chávez et al., 2002). Fig. 4a). Fortunately, an elongation of the extraction over 1 h,
Alkaline solubilization of proteins was performed in the which would have provoked additional manufacturing costs
range of pH 7 to 12 revealing that solubility increased by ele- and favored enzymatic degradation processes, was not neces-
vating the amount of sodium hydroxide (see Fig. 3a). Similarly, sary.
plant-based proteins from soybean, rice, pea and amaranth As a final factor, temperature is known to influence con-
become more soluble in alkaline medium (Maruyama et al., formation and hydrophilic behavior of proteins. In order to
1999; Wang et al., 1999; Anon et al., 2001; Tömösközi et al., avoid irreversible protein denaturation during extraction, tem-
2001; Salcedo-Chávez et al., 2002). A pH-value of 10 was cho- perature was kept below 60 ◦ C. As can be observed in Fig. 4b
sen for further extraction trials. Higher alkaline values were the heating to 25 ◦ C increased solubility by 6% (absl.) which
avoided in order to reduce a possible negative impact on essen- was not significantly elevated by further heating. As for
tial amino acids such as lysine. These are said to generate the duration, lower temperatures are advantageous to limit
lysinoalanine, resulting in a loss of protein digestibility and manufacturing costs and chemical modifications. Thus, fur-
biological value (Finot, 1997). ther extractions were performed at room temperature (20 ◦ C)
The impact of the particle size (250 ␮m, 500 ␮m, 750 ␮m) on because these factors were considered superior to the slightly
the protein solubility of quinoa bran is presented in Fig. 3b. elevated protein yield.
The objective was to study the protein accessibility depend-
ing on the degree of comminution. The reduction in particle 3.5. Protein purification by precipitation
size up to 250 ␮m resulted in significantly higher protein sol-
ubility (67.7%). In comparison to the non-milled bran this was Protein–protein interactions increase in the area of the iso-
an increase in protein content of 13% (rel.). A possible expla- electric point as a consequence of minimum net electrostatic
nation might derive from the accessibility of proteins through charges of the molecules. This isoelectric point of the quinoa
larger particle surfaces. proteins (extracted at pH 10) was evaluated in a pH-range
Another critical factor for extraction processes is the dura- between 2 to 6. After acidification, the solubility of proteins
tion, since diffusion and enzymatic degradation processes are was determined in proportion to the amount of previously
time-dependent. Therefore, the extraction was elongated up extracted proteins. Fig. 5a reveals that proteins which were
to four hours to test the effect on protein solubility. As a initially extracted at pH 10, had the lowest solubility at pH

Fig. 4 – Impact of extraction time and temperature on the protein solubility of quinoa bran. (a) Extraction in water over time
(30 min to 4 h). (b) Extraction in water with varying temperature (20 ◦ C to 50 ◦ C). The protein content of the supernatant was
measured after centrifugation at 4500 × g (N = 5.54). Means are presented with standard deviation (n = 4). Different letters
denote statistically significant differences (ANOVA, p < 0.05).
 food and bioproducts processing 9 6 ( 2 0 1 5 ) 20–26 25

Fig. 5 – Impact of purification by isoelectric precipitation and centrifugation on protein solubility and content. (a) Protein
solubility (g protein per 100 g protein in bran) depending on the pH-value after aqueous extraction at pH 10. N = 5.54. (b)
Protein content after freeze drying when using different centrifugal forces for the isolation of the precipitate. db = dry base.
Means are presented with standard deviation (n ≥ 4). Different letters denote statistically significant differences (ANOVA,
p < 0.05).

4, indicating the isoelectric point. Thus, precipitation at pH 4 volume. In this case, the separation of quinoa bran resulted in
can be used to exclude non-protein solutes but only 41.2% of beneficial effects on the gas retention capacity of gluten-free
the extracted bran proteins were yielded after this purification dough. Moreover quinoa white flour improved bread qual-
step, while the remaining proteins remained in solution. ity regarding crust color, loaf volume and sensory attributes.
Both, the separation of insoluble products from the alkaline Considering the nutritive value and economic aspects also
protein solution after the initial extraction stage, as well as the milling byproducts such as quinoa bran feature a high poten-
separation of the precipitated proteins from soluble residues tial for the application in gluten-free products.
after acidification, were performed by centrifugation. Focus- Since gluten-free flours are mostly lacking protein when
ing on maximizing the protein yield, these centrifugation based on corn (6.58% db) or rice (6.99% db), the targeted appli-
steps were performed with three different centrifugal forces: cation of small quinoa bran amounts can improve the nutritive
4500 × g, 9000 × g and 15,000 × g. The protein content (% db) of value without impairing gas holding properties and bread vol-
quinoa bran extracts is presented in Fig. 5b. Without purifi- ume (Föste et al., 2014). In addition, the protein extracts can
cation (=reference) the extract contained 45.83% db protein, be of relevance as functional food ingredients, when featu-
while precipitation stage for removing soluble non-protein ring stabilizing or foaming properties as observed for isolated
compounds increased the protein content up to 58.92% db, amaranth proteins (Fidantsi and Doxastakis, 2001). Further
as expected. By elevating the centrifugal speed, protein yield analyses should cover the impact of quinoa proteins as ingre-
was further improved, reaching its maximum at 67.93% db at dients in gluten-free bread or dietary supplements in order to
15,000 × g. Moreover as a comparison to the use of quinoa bran enrich protein content and improve food quality.
as raw material the same extraction process was performed
with quinoa whole grain flour (at maximum centrifugal speed) Acknowledgements
resulting in a lower protein content (−22.89% rel.).
To sum up, quinoa bran is a suitable byproduct of the This work is based on a research project (16847 N), which
milling process, which facilitates the concentration of the pro- was supported by the German Ministry of Economics and
teins. The results indicate that conditioning of quinoa is the Technology (via AIF Project GmbH, Berlin, Germany) and the
first step to separate the outer parts of the seed from the FEI (Forschungskreis der Ernährungsindustrie e. V., Bonn,
starch-rich perisperm by choosing the right pre-conditioning Germany).
settings (15% seed moisture and 20 h resting time). Compared
to whole grain flour, in particular protein and minerals were
enriched in the bran fraction up to 136% (rel.) and 133% (rel.), References
respectively. The protein content of the extract obtained in this
study is comparable to the results of Jiamyangyuen et al. (2005) Aceituno-Medina, M., Lopez-Rubio, A., Mendoza, S., Lagaron, J.M.,
2013. Development of novel ultrathin structures based in
who produced extracts from non-defatted and defatted rice
amaranth (Amaranthus hypochondriacus) protein isolate
bran ending up with 72% protein. The presented combination
through electrospinning. Food Hydrocolloids 31, 289–298.
of fractionation and solvent extraction provides an efficient American Association of Cereal Chemists, 2002. Approved
method for the concentration of proteins from quinoa. methods. In: AACC International: Method: No 46-10, No 08-12,
No 44-01, 10th ed. American Association of Cereal Chemists,
St. Paul, MN.
4. Conclusion Ando, H., Chen, Y.C., Tang, H., Shimizu, M., Watanabe, K.,
Mitsunaga, T., 2002. Food components in fractions of quinoa
Depending on the industrial purpose, milling fractions fea- seed. Food Sci. Technol. Res. 8 (1), 80–84.
ture their own technological characteristics, which can be Anon, M.C., Sorgentini, D.A., Wagner, J.R., 2001. Relationships
between different hydration properties of commercial and
individually applied for modifying gluten-free foods, bakery
laboratory soybean isolates. J. Agric. Food Chem. 49,
products or dietary supplements. More recently observed by
4852–4858.
Elgeti et al. (2014), quinoa white flour exhibits considerably Aufhammer, W., Lee, J.H., Kübler, E., Kuhn, M., Wagner, S., 1995.
higher substrate availability than rice and corn flour, result- Anbau und Nutzung der Pseudocerealien Buchweizen
ing in elevated carbon dioxide formation and increased bread (Fagopyrum esculentum Moench) Reismelde (Chenopodium
26 food and bioproducts processing 9 6 ( 2 0 1 5 ) 20–26

quinoa Willd.) und Amarant (Amaranthus spp.) als function relationships of soybean ␤-conglycinin constituent
Körnerfruchtarten. Bodenkultur 46 (2), 125–140. subunits. J. Agric. Food Chem. 47, 5278–5284.
Bera, M.B., Mukherjee, R.K., 1989. Solubility, emulsifying, and Oliver, J.R., Blakeney, A.B., Allen, H.M., 1993. The colour of flour
foaming properties of rice bran protein concentrates. J. Food streams as related to ash and pigment contents. J. Cereal Sci.
Sci. 54 (1), 142–145. 17 (2), 169–182.
Bressani, R. Amaranth, 2003. In: Caballero, B. (Ed.), Encyclopedia Paredes-López, O., Guzman-Maldonado, H., Ordorica-Falomir, C.,
of Food Sciences and Nutrition 2. Academic Press, Oxford, pp. 1994. Food proteins from emerging seed sources. In: Hudson,
166–173. B.J.F. (Ed.), New and Developing Sources of Food Proteins.
Cauvain, S.P., 2003. Breadmaking: an overview. In: Cauvain, S.P. Springer, US, pp. 241–279.
(Ed.), Breadmaking: Improving Quality. Woodhead Publishing, Raghavendra, S.N., Ramachandra Swamy, S.R., Rastogi, N.K.,
Cambridge, pp. 8–28. Raghavarao, K.S.M., Kumar, S., Tharanathan, S.R.N., 2006.
Chauhan, G.S., Eskin, N.A.M., Tkachuk, R., 1992. Nutrients and Grinding characteristics and hydration properties of coconut
antinutrients in quinoa seed. Cereal Chem. 69 (1), 85–88. residue: a source of dietary fiber. J. Food Eng. 72 (3),
de la Hera, E., Talegón, M., Caballero, P., Gómez, M., 2013. 281–286.
Influence of maize flour particle size on gluten-free Salcedo-Chávez, B., Osuna-Castro, J.A., Guevara-Lara, F.,
breadmaking. J. Sci. Food Agric. 93 (4), 924–932. Domínguez-Domínguez, J., Paredes-López, O., 2002.
Edney, M.J., Rossnagel, B.G., Ozawa, S., Brophy, M., 2002. Pealing Optimization of the isoelectric precipitation method to obtain
quality of Canadian barley varieties and the potential use as protein isolates from amaranth (Amaranthus cruentus) seeds. J.
rice extenders. J. Cereal Sci. 36, 295–305. Agric. Food Chem. 50 (22), 6515–6520.
Elgeti, D., Nordlohne, S.D., Föste, M., Besl, M., Linden, M.H., Heinz, Scanlin L.A., Stone, M.B, Burnett, C. Quinoa protein concentrate,
V., Becker, T., 2014. Volume and texture improvement of production and functionality. US Patent application
gluten-free bread using quinoa white flour. J. Cereal Sci. 59 (1), 12/748,986. Patent application number: 20100184963,
41–47. 2010-07-22.
Fang, C., Campbell, G.M., 2003. On predicting roller milling Scilingo, A.A., Molina Ortiz, S.E., Martıı́nez, E.N., Añón, M.C., 2002.
performance. V: Effect of moisture content on the particle size Amaranth protein isolates modified by hydrolytic and thermal
distribution from first break milling of wheat. J. Cereal Sci. 37 treatments. Relationship between structure and solubility.
(1), 31–41. Food Res. Int. 35, 855–862.
Fidantsi, A., Doxastakis, G., 2001. Emulsifying and foaming Segura-Nieto, M., Shewry, P.R., Paredes-López, O., 1999. In: Shevry,
properties of amaranth seed protein isolates. Colloids Surf., B: P.R., Casey, R. (Eds.), Globulins of the Pseudocereals:
Biointerf. 21 (1), 119–124. Amaranth, Quinoa, and Buckwheat in Seed Proteins. Springer
Finot, P.K., 1997. Effects of processing and storage on the Verlag, Netherlands, pp. 453–475.
nutritional value of food proteins. In: Damodaran, S., Paraf, A. Stewart, M.L., Slavin, J.L., 2009. Particle size and fraction of wheat
(Eds.), Food Proteins and Their Applications. Dekker, New bran influence short-chain fatty acid production in vitro. Br. J.
York, NY, pp. 551–577. Nutr. 102 (10), 1404–1407.
Fujihara, S., Sasaki, H., Aoyagi, Y., Sugahara, T., 2008. Taylor, J.R., Parker, N., Quinoa., M.L., 2002. In: Belton, P.S., Taylor,
Nitrogen-to-protein conversion factors for some cereal J.R.N. (Eds.), Pseudocereals and Less Common Cereals: Grain
products in Japan. J. Food Sci. 73 (3), C204–C209. Properties and Utilization. Springer Verlag, Berlin,
Föste, M., Nordlohne, S.D., Elgeti, D., Linden, M.H., Heinz, V., Jekle, pp. 93–122.
M., Becker, T., 2014. Impact of quinoa bran on gluten-free Thompson, T., 2001. Case problem-questions regarding the
dough and bread characteristics. Eur. Food Res. Technol. 239 acceptability of buckwheat, amaranth quinoa, and oats from a
(5), 767–775. patient with celiac disease. J. Am. Diet. Assoc. 101 (5),
Gorinstein, S., Pawelzik, E., Delgado-Licon, E., Haruenkit, R., 586–587.
Weisz, M., Trakhtenberg, S., 2002. Characterisation of Tömösközi, S., Lasztity, R., Haraski, R., Baticz, O., 2001. Isolation
pseudocereal and cereal proteins by protein and amino acid and study of the functional properties of pea proteins.
analyses. J. Sci. Food Agric. 82 (8), 886–891. Nahrung 45, 399–401.
International Association for Cereal Science and Technology, Valencia-Chamorro, S., 2003. Quinoa. In: Caballero, B. (Ed.),
1998. ICC: Method No. 207. ICC. Encyclopedia of Food Science and Nutrition, vol. 8. Academic
Jiamyangyuen, S., Srijesdaruk, V., Harper, W.J., 2005. Extraction of Press, Amsterdam, The Netherlands, pp. 4895–4902.
rice bran protein concentrate and its application in bread. Wang, M., Hettiarachchy, N.S., Qi, M., Burks, W., Siebenmorgen, T.,
Extraction 27 (1), 56. 1999. Preparation and functional properties of rice bran
Maruyama, N., Sato, R., Wada, Y., Matsumura, Y., Goto, H., Okuda, protein isolate. J. Agric. Food Chem. 47 (2), 411–416.
E., Nakagawa, S., Utsumi, S., 1999. Structure–physicochemical