1. Introduction
Coffee is a popular beverage and arguably one of the most traded commodities in the world. Due to the high demand for coffee, large amounts of by-products such as pulp, silver skin, and parchment are generated. According to the International Coffee Organization, in 2019, about 10 million tons of fresh coffee were produced globally. It implies the generation of approximately 3 million tons of coffee cherry pulp (CCP) as a by-product and represents a severe environmental problem. Additionally, CCP contains fiber, minerals, amino acids, and polyphenolic compounds potentially beneficial for human nutrition [
1]. Several studies have analyzed CCP’s chemical composition, reporting that the dried material has about 10% crude protein, 21% crude fiber, 8% ash, and 44% nitrogen-free extract; these values change according to coffee variety, location, and agricultural practices [
2]. Furthermore, hydroxycinnamic acids as chlorogenic, caffeic, and ferulic acid in CCP are of interest for their antioxidant properties, which neutralize excess free radicals to prevent cell damage from free radicals [
3,
4].
On the other hand, bakery products are highly consumed worldwide, usually elaborated with refined flours, deficient dietary fiber, and other nutrients [
5]. So, CCP could be used as an ingredient in fiber-deficient bakery products. Bakery products are technologically demanding, and changes in their formulation or process could affect their quality. Fiber modifies the rheological properties of dough and bread quality [
6]. It is well known that the food matrix’s viscoelasticity or rheological behavior is related to the composition, structure, and stability; therefore, the rheological characterization of a food matrix and its components is overriding for predicting food quality [
7].
In the breadmaking process, the dough is subjected to different deformation types during mixing, fermentation, rolling and shaping, proofing, and baking [
8], and even when we chew it. This is why, rheological studies are helpful research tools to determine the dough’s behavior and estimate the interrelations between flour composition, ingredients functionality, process parameters, and loaf characteristics [
8].
Empirical tests conducted with the farinograph, alveograph, and extensograph are widely used in the baking industry as tools to predict baking quality while fundamental rheological tests continue probing it [
9,
10,
11,
12,
13]. The creep recovery test measures the material’s viscoelasticity by applying constant stress during a determined time [
14]. Many researchers have tried to understand gluten and dough’s viscoelastic behavior by modeling experimental data using the Burgers model with Maxwell and Kelvin elements [
10,
12,
13,
15,
16,
17]; the model’s values were correlated with breadmaking quality [
6,
13]. Because of the complexity of the protein network in gluten, the viscoelastic behavior for a breadmaking quality can change among wheat cultivars, new bread formulations, and storage.
Creep and recovery tests have been widely used to measure the impact of dietary fiber on the viscoelastic behavior of rice flour doughs [
18], the effect of high molecular weight glutenin subunits in gluten and its relationship with the quality of bread [
13], rheological, and functional properties of red bean flour of gluten-free batter for cupcakes [
17]. Other authors investigated the effect of flour substitution by bran with different particle sizes on dough rheology. They reported a significant effect of substitution level and a minor effect of particle size [
19]. In general, fiber added to flour affects the gluten-starch interaction, reduces swelling of starch granules and its availability for gelatinization, compels gas cells to expand, and adversely affects dough viscoelastic behavior constraining dough availability to trap gas [
20]. It is of interest to assess the CCP by-product in food applications and specifically in baked products. By assessing the influence of CCPP in gluten and non-gluten components of dough, a comparison can be made on the changes in dough’s viscoelasticity and stiffness, both essential attributes in the baking process [
13]. Thus, the objective of this study was to investigate the effect of CCPP on the wheat dough and gluten rheological properties and baking quality without any other additives.
2. Materials and Methods
2.1. Materials
Decaffeinated CCPP (Pulphari) was obtained from Grupo Techver S.A. de C.V. (Veracruz, Mexico). Commercial Spring wheat flour (enriched, bleached) was donated by Shawnee Milling Company (Shawnee, OK, USA).
The particle size of CCPP was further reduced with an impact-type mill (Kitchen Mill, Blendtec, Orem, UT, USA), sieved (screen number 100) to obtain a more homogeneous powder then stored in closed containers at room temperature until needed for analyses. In total, three levels of wheat flour substitution by CCPP were used 0, 1.25, 2.5, and 5% wet basis (0, 1.2, 2.3, and 4.7% dry basis). These levels were selected from results of preliminary experiments with levels up to 30% CCPP. At CCPP substitution levels greater than 5%, the bread had a grassy and earthy aftertaste, and in samples with substitution levels greater than 10%, the dough was difficult to handle.
2.2. Flour and CCPP Analysis
CCPP and wheat flour were evaluated according to the Association of Official Agricultural Chemists (now AOAC International) Official Methods [
21] for protein (920.87), moisture (925.10), ash (923.03), and fat (945.16). Soluble and insoluble dietary fiber were evaluated by the American Association of Cereal Chemists International (AACCI) Official Method 32-07.01 [
22] and AOAC 991.43 [
21]. The proximate composition of flour was provided by Shawnee Milling Co. (Shawnee, OK, USA). The particle size of CCPP flour was analyzed with AACCI method 66-20.01 [
22] to obtain particles lower than 150 µm. Flour blends were thoroughly mixed, and their rheological behavior determined by AACCI Approved Method 54-21.02 [
22] to obtain water absorption at 500 Brabender Units (BU) of consistency, development time, stability time, and mix tolerance index during mixing using a Farinograph-E (C.W. Brabender Instruments, South Hackensack, NJ, USA). High values of development and stability times are key indicators of dough strength.
Caffeine Concentration in CCPP
Caffeine was analyzed by high-performance liquid chromatography (HPLC) Shimadzu (Model LC-20AD/T LPGE Kit Shimadzu Co., Kyoto, Japan) using the method of the International Organization for Standardization (ISO) 20481 [
23], and an aqueous extraction of 1 ± 0.001 g of CCPP in 250 mL of water. The mixture was heated in a water bath for 20 min at 90 °C and filtered through 0.20 µm nylon membrane (SUNSri, Wilmington, NC, USA). The mobile phase was 24% methanol and 76% water (
v/
v) at a flow rate of 1.0 mL/min, and the injection volume was 10 µL, using a RP-18 (Reverse Phase) column with isocratic elution and the absorbance recorded at 272 nm with an UV detector. The caffeine content was calculated with anhydrous caffeine (C
8H
10O
2N
4) as external standard (Fermont Chemical Company, Monterrey, Nuevo Leon, Mexico).
2.3. Sample Preparation for Rheology Test
2.3.1. Gluten Samples Preparation
According to AACCI Approved Method 38-12.02 [
22], the extraction of wet gluten was carried out with a Glutomatic System (Perten Instruments AB, Huddinge, Sweden) model 2200 with 20 g NaCl/L aqueous solution. Briefly, a 10-g flour sample was made into a dough; starch and other components were removed by washing the dough with the salt solution. The wet gluten obtained was at its intrinsic water-binding capacity.
2.3.2. Dough Sample Preparation
The dough was prepared at constant water absorption using the value of flour without CCPP obtained in the farinograph. Samples were prepared in a modified Mixograph (National Manufacturing, Lincoln, NE, USA) equipped with a 10 g bowl and mixed for 4 min.
2.4. Compression-Recovery, Bi-Axial Deformation
Gluten samples were shaped as previously reported [
17] using a Perten Centrifuge 2015 at 2430×
g for 1 min (Perten Instruments AB, Huddinge, Sweden). A compression-recovery based on biaxial deformation was performed using a gluten compression-relaxation (Gluten CORE) Analyzer (Perten Instruments AB, Huddinge, Sweden) with the following conditions: velocity start 20 mm/s, compression rate 4 mm/s, target force 0.5 N, target force compression 8 N, compression time 5 s, target force recovery 0.2 N, and recovery time 55 s. Gluten elastic recoverability was reported as elastic recovery in percentage.
2.5. Dough and Gluten Sample Loading for Creep and Recovery Test
Freshly extracted gluten and prepared dough samples with different substitution levels of CCPP were relaxed under a press of 2.5 kg top plate with a 2.5 mm gap for 40 min at room temperature as reported [
13]. A 25 mm gluten and dough disc were loaded into an AR1000 rheometer (TA Instruments, New Castle, DE, USA). A parallel hatched plate (diameter = 25 mm) geometry was used with a 2.5 mm gap, and the temperature was kept constant at 25 °C. The gluten and dough discs were retrimmed, and the edges covered with mineral oil to prevent moisture loss during the test. The constant stress of 100 Pa was applied for 100 s in the creep phase, and the strain was recorded in the recovery phase (without stress) during 100 s to measure the deformation response of gluten and dough samples (within the linear viscoelastic region).
2.6. Modeling of Viscoelastic Properties of Bread Dough and Gluten
An analog mechanical model composed of the concepts of springs (Hookean bodies) and dashpots (fluid bodies) was used to study the viscoelastic behavior of bread dough and gluten. The Kelvin–Voigt generalized model of six elements was used to describe experimental data:
where
γ(
t) is the strain as a function of time,
σ0 is the applied stress,
G0 is the elastic modulus for the Hookean body response:
G1 and
G2 are elastic moduli of the first and second retarded elastic response (first and second Kelvin–Voigt elements, respectively).
η0 is the steady state viscosity [
13].
λ1 =
η1/
G1 and
λ2 =
η2/
G2 are the first and second retardation time required to achieve 63.21% of deformation in the expected creep of the specific Kelvin–Voigt element. The
η1 and
η2 are the viscous response of the first and second retarded elastic deformation, respectively, of the Kelvin–Voigt element’s dashpot. In terms of shear compliance:
where
J(
t) is the compliance (1/Pa),
J0 is the compliance at zero time,
J1 and
J2 are the compliances in the first and second retarded elastic response, respectively. After removing the shear stress, the recovery curve was analyzed with a similar equation for recovery compliance
Jr(
t); as there is no viscous flow in the recovery phase, the equation only has five elements [
13].
where
Jr(
t) is the compliance (1/Pa) at the recovery phase,
Jr0 is the compliance at zero time at the recovery phase,
Jr1 and
Jr2 are the compliance in the first and second retarded elastic response at the recovery phase and
λr1,
λr2 are the first and second retarded elastic response at the recovery phase.
2.7. Bread Baking Quality
AACC Approved Method 10-10.03 [
22] for optimized bread-baking test for evaluating wheat flour quality was carried out with 1-pound loaves. Loaf height, weight, and volume were measured 1 h after baking. Volume was measured by rapeseed displacement following Approved Method 10-05.01 [
22]. Bread crumb firmness was measured on days 1, 4, and 7 of storage at room temperature (25 °C) according to the American Institute of Baking (now AIB International) Standard Procedure for White Pan Bread [
24] using 25% of height compression in two bread slices of 12.5 mm thickness. A 25 mm round probe and 10 g force of compression were used.
2.8. Statistical Analysis
Creep recovery curves were fitted with Origin Pro 9.1 software (OriginLab, Northampton, MA, USA) using nonlinear regression analyses. Analysis of variance, multiple means comparison (p < 0.05, Tukey), and Pearson’s correlations were performed with JMP Pro 14.0.0 (SAS Institute Inc., Cary, NC, USA).