COFFEE EXTRACTION PROCESS AND COFFEE PRODUCT

Abstract

The present invention provides an instant coffee composition for forming a coffee beverage, wherein the composition comprises at least 6 wt % of an insoluble coffee sediment fraction, the insoluble coffee sediment fraction comprising, when analysed after acid hydrolysis, 1 wt % or less arabinose.

Claims

1. An instant coffee composition for forming a coffee beverage, wherein the composition comprises at least 6 wt % of an insoluble coffee sediment fraction, the insoluble coffee sediment fraction comprising, when analysed after acid hydrolysis, 1 wt % or less arabinose.

2. The instant coffee composition according to claim 1, wherein the composition comprises from 7.5 to 15 wt % of the insoluble coffee sediment fraction.

3. The instant coffee composition according to claim 1, wherein the insoluble coffee sediment fraction comprises, when analysed after acid hydrolysis, from 0.5 to 1 wt % arabinose.

4. The instant coffee composition according to claim 1, wherein the insoluble coffee sediment fraction comprises, when analysed after acid hydrolysis, less than 5 wt % galactose, preferably from 2 to 4 wt % galactose.

5. The instant coffee composition according to claim 1, wherein the instant coffee composition comprises at least 0.8 wt % coffee oils by dry weight, preferably from 1 to 5 wt % coffee oils.

6. The instant coffee composition according to claim 1, wherein the composition when analysed by wet laser diffraction at a 1.5 wt % concentration has a D50 of less than 10 microns, preferably from 2.5 to 7.5 microns.

7. The instant coffee composition according to claim 1, wherein the composition consists of coffee.

8. The instant coffee composition according to claim 1, wherein the composition is spray- or freeze-dried, or wherein the instant coffee composition is a liquid coffee concentrate.

9. A method for the manufacture of a coffee-extract product, the method comprising: (a) providing roast and ground coffee having a mean particle size of from 100 to 600 microns; (b) mixing the roast and ground coffee with water to form a first slurry containing 15 to 30 wt % coffee solids, (c) passing the first slurry through an aroma-separation step to recover a coffee aroma fraction and to form a dearomatised slurry; (d) passing the deaeromatised slurry to a first filtration device at a temperature of from 90 to 150° C. to form a first coffee extract and a first filter cake; (e) adding water to the first filter cake to form a reconstituted slurry having at least 12 wt % coffee solids; (f) thermally treating the reconstituted slurry at a temperature of from 150 to 205° C.; (g) then passing the thermally-treated reconstituted slurry to a second filtration device to form a second coffee extract and a second filter cake; (h) combining the first and second coffee extracts to form a third coffee extract; (i) concentrating the third coffee extract to form a fourth coffee extract having 35 to 70 wt % coffee solids; (j) adding the coffee aroma fraction to the fourth coffee extract to form a liquid, coffee-extract product.

10. The method according to claim 9, wherein the roast and ground coffee has a mean particle size of from 400 to 600 microns, or wherein the roast and ground coffee has a mean particle size of from 250 to 400 microns.

11. The method according to claim 9, wherein the coffee-extract product is a soluble powder, the method further comprising: (k) drying the liquid coffee-extract product to form a soluble powder.

12. The method according to claim 9, wherein the liquid, coffee-extract product has 40 to 50 wt % coffee solids.

13. The method according to claim 9, wherein the water in step (b) and/or step (e) is at a temperature of from 80 to 100° C.

14. The method according to claim 9, wherein the reconstituted slurry formed in step (e) has 12 to 30 wt % solids.

15. The method according to claim 9, wherein the second filter cake is subjected to a further high temperature extraction process to obtain a further coffee extract to be combined in step (h) with the first and second coffee extracts to form the third coffee extract.

16. The method according to claim 9, wherein step (f) is conducted in a plug-flow reactor.

17. The method according to claim 9, wherein step (i) is conducted in an evaporator unit.

18. The method according to claim 9, wherein step (c) is conducted under vacuum.

19. The method according to claim 9, wherein the method further comprises packaging the coffee-extract product.

20. The method according to claim 9, wherein the method is a continuous process.

21. A coffee-extract product obtainable by the method of any of claim 9.

Description

[0110] The invention will now be described further with respect to the figures, in which:

[0111] FIG. 1 shows a flow-chart of the steps of the present invention.

[0112] FIG. 2 shows a plot of the viscosity of various samples at different shear rates.

[0113] FIG. 3 shows sensory data from a trial.

[0114] As shown in FIG. 1, the method for the manufacture of a coffee-extract product includes a number of steps.

[0115] In step (a) roast and ground coffee is provided having a mean particle size of from 100 to 600 microns, preferably 200 to 600 microns. Within this range, larger sizes are favoured for liquid extract products, whereas smaller sizes are favoured for dried soluble coffee products.

[0116] In step (b) the roast and ground coffee with water 5 to form a first slurry 10 containing 15 to 30 wt % coffee solids. The water 5 is added at a temperature of from 80 to 100° C., and preferably from 90 to 95° C. The solids level is determined by the particle size, since a minimum amount of water 5 is used as necessary to obtain a pumpable slurry 10. The larger the particle size, the more water 5 is required (the lower the solids) to achieve a pumpable slurry 10.

[0117] In step (c) the first slurry is passed through an aroma-separation step to recover a coffee aroma fraction 15 and to form a dearomatised slurry 20. A typical approach to this method involves the addition of steam to the pumpable slurry 10 where the vapours are treated in a spinning cone treatment unit.

[0118] In step (d) the deaeromatised slurry 20 is passed to a first filtration device at a temperature of from 90 to 150° C., such as 90 to 100° C., to form a first coffee extract 25 and a first filter cake 30. The temperature is retained from the preceding step or can be further increased to increase the extraction yield. The filter cake 30 may be washed and is pressed to obtain the largest possible amount of soluble coffee solids.

[0119] In step (e) water 5 is added to the first filter cake 30 to form a reconstituted slurry 35 having at least 12 wt % coffee solids. The water 5 is preferably hot and there may be mechanical agitation to break up the first filter cake 30. The amount of water required to reconstitute a slurry tends to be higher than that required in step (b).

[0120] In step (f) the reconstituted slurry 35 is thermally treated at a temperature of from 150 to 205° C., such as 180 to 205° C. to form a thermally-treated reconstituted slurry 40. That is, it is pumped through a heat-treatment unit, such as a plug-flow reactor. Residence times in the heat treatment are typically at least 5 minutes to ensure good extraction.

[0121] In step (g) the thermally-treated reconstituted slurry 40 is passed to a second filtration device to form a second coffee extract 45 and a second filter cake 50. The second filter cake 50 may be washed and is pressed to obtain the largest possible amount of soluble coffee solids. The temperature in this step may be retained from the preceding step, or may be lowered as heat is recovered for use in step (b), such as down to a temperature of from 80 to 100° C.

[0122] The second filter cake 50 may then be burned in step M to produce heat for the process, or may be subjected to a further high temperature extraction step M to obtain a further coffee extract 52.

[0123] In step (h) the first coffee extract 25 and the second coffee extract 45 are combined to form a third coffee extract 55. Other aqueous coffee extracts may also be added in this step, such as further coffee extract 52.

[0124] In step (i) the third coffee extract 55 is concentrated to form a fourth coffee extract 60 having 35 to 70 wt % coffee solids, such as 35 to 60 wt % coffee solids.

[0125] In step (j) the coffee aroma fraction 15 is added to the fourth coffee extract 60 to form a liquid, coffee-extract product 65.

[0126] The liquid coffee extract product 65 may be treated in step K to form a dried coffee product, such as a soluble coffee powder 70. The liquid coffee extract product 65 may be diluted in step L to form a liquid coffee concentrate 80.

[0127] In FIG. 3 the current technology is represented by the smallest quadrilateral. The other two quadrilaterals represent different Prototypes with 70% current and 30% new technology products. The axis are: positive x (Viscous); positive y (Turbid); negative x (powdery); negative y (dry).

[0128] The present invention will now be described further in relation to the following non-limiting example.

EXAMPLE 1

[0129] Roast whole beans were ground to between 200 μm and 400 μm in a 3 stage roller grinder.

[0130] The roast and ground coffee was slurried with water at 20° C.-30° C. at a ratio of 25% Coffee to 75% water.

[0131] The slurry was fed forward into a heat exchanger and heated to 95° C. before moving into a spinning cone column where aroma was stripped from the slurry.

[0132] Upon exit of the spinning cone the slurry was fed forward through a heat exchanger, raising the temperature to between 120° C. and 150° C. for 2 to 5 minutes.

[0133] The slurry was then fed into a filter separating the coffee liquor from the grounds. The grounds were then subjected to 2 further washing steps at 130° C. to 150° C. to remove additional solids.

[0134] The grounds were then re-slurried at a ratio of 12% to 17% solids with fresh water. The resulting slurry was fed forward to a hydrolysis step where it was heated to between 180° C. and 205° C. (185° C.) and held for between 5 and 20 minutes.

[0135] The resulting slurry was then cooled to below 100° C. before passing through a second filtering step repeating the separation and washing of the first separation step.

[0136] The coffee extracts obtained from each filtration step were combined and concentrated. The aroma compounds stripped from the first slurry where then added to the mixture. The fully combined three components were then freeze-dried with a conventional process to obtain a soluble coffee powder.

[0137] The process recovered an incremental yield of 2% roasted coffee over current technologies with reduced water usage.

EXAMPLE 2

[0138] Arabica and/or Robusta beans were roasted and ground, using a 3-stage roller grinder, to a mean particle size of 300 um. The ground coffee was then slurried with water at 20-25° C. at a ratio of 25% coffee to 75% water.

[0139] The slurry was fed forward into a heat exchanger and heated to 70° C. before moving into a spinning cone column where aroma was stripped from the slurry.

[0140] The slurry was then fed into a filter at a temperature of 95° C. separating the coffee liquor from the grounds. The grounds were then subjected to 2 further washing steps to remove additional solids.

[0141] The grounds were then re-slurried at a ratio of 12% to 17% solids with fresh water. The resulting slurry was fed forward to a plug-flow reactor (hydrolysis step) where it was heated to 170° C. and held for 5-10 minutes.

[0142] The resulting slurry was then cooled to below 100° C. before passing through a second filtering step repeating the separation and washing of the first separation step.

[0143] The coffee extracts obtained from each filtration step were combined and concentrated. The aroma compounds stripped from the first slurry where then added to the mixture. The fully combined three components were then freeze-dried with a conventional process to obtain a soluble coffee powder.

[0144] The product of this example was found to have more body/mouthfeel than products produced using current technology.

EXAMPLE 3

[0145] A coffee slurry was prepared as described in Example 1.

[0146] The slurry was fed forward into a heat exchanger and heated to 95° C. before moving into a spinning cone column where aroma was stripped from the slurry.

[0147] Upon exit of the spinning cone the slurry was fed forward through a heat exchanger, raising the temperature to between 145-150° C. for 4 to 5 minutes.

[0148] The slurry was then fed into a filter separating the coffee liquor from the grounds. The grounds were then subjected to 2 further washing steps at 140° C. to remove additional solids.

[0149] The slurry was then fed into a filter separating the coffee liquor from the grounds. The grounds were then re-slurried at a ratio of 12% to 17% solids with fresh water. The resulting slurry was fed forward to a plug-flow reactor (hydrolysis step) where it was heated to 200° C. and held for 7-10 minutes.

[0150] The resulting slurry was then cooled to below 100° C. before passing through a second filtering step repeating the separation and washing of the first separation step.

[0151] The coffee extracts obtained from each filtration step were combined and concentrated. The aroma compounds stripped from the first slurry where then added to the mixture. The fully combined three components were then freeze-dried with a conventional process to obtain a soluble coffee powder.

[0152] The product of this example was found to have more body/mouthfeel than products produced using current technology.

EXAMPLE 4

[0153] Arabica and/or Robusta beans were roasted and ground, using a 3-stage roller grinder, to a mean particle size of 400 um. The ground coffee was then slurried with water at 20-25° C. at a ratio of 15% coffee to 85% water.

[0154] The remainder of the process was conducted as per example 1.

[0155] The resulting product has lower levels of oil than the product of example 1.

EXAMPLE 5

[0156] Samples obtained by the method described herein were assessed in comparison to a range of commercially available soluble coffee products. As can be seen from the comprehensive testing, the products obtained by the process are new and can be readily distinguished from products obtained from conventional processes.

Oil Content

[0157]

TABLE-US-00001 Sample Type Bean blend* Fat Content 1 Alta Rica Pure Instant Arabica 0.3 2 Nescafe Gold Whole Bean Arabica/Robusta 0.4 Blend Instant 3 Kenco Really Pure Instant Arabica/Robusta 0.2 Rich 4 Milicano Whole bean Arabica/Robusta 0.7 Instant 5 Percol Pure Instant Robusta 0 6 Kenco Really Pure Instant Arabica/Robusta 0.2 Rich 7 Prototype Product of Arabica 1.8 Colombian invention 8 Prototype Product of Arabica 1.8 Central invention 9 Prototype Product of Robusta 0.4 Robusta invention 10 Prototype Brazil Product of Arabica 3.9 invention *bean identity for competitor products is based on an educated guess

[0158] Examples 7, 8, 9 and 10 have been produced in accordance with the method described herein. Examples 1-6 are commercially available products, of which 2 and 4 are products supplemented with added roast and ground coffee additives (designated “whole bean instant” in the table).

[0159] It should generally be appreciated that levels of oil in Robusta beans are lower than in Arabica beans. This is reflected with the generally lower levels of oil in products comprising Robusta beans, include inventive example 9. Sample 10 is a dark Brazil known for high oil levels.

[0160] As can be seen, there are low levels of oil in the pure instant coffees, i.e. samples 1, 3, 5 and 6, which have not been supplemented with roast and ground coffee additives. The oil levels are slightly higher in samples 2 and 4 due to the oil content of the roast and ground coffee additives, with sample 2 containing approximately 5% roast and ground coffee and Sample 4 containing more roast and ground coffee.

[0161] Samples 7, 8 and 10 contain high levels of oil due to the fine grind of roasted coffee in the new process which releases more oil into the extract.

[0162] As can be seen, no conventional soluble coffee products contain significant levels of oil. Indeed, it is speculated that the levels of oil observed for some of these products is added afterwards to the surface of the dried powder to improve its aroma.

[0163] The only prior art products which contain high oil levels are a consequence of the addition of roast and ground coffee additives in the product. In contrast, the method described herein achieves high levels of oil, even for Robusta bean products.

Sediment Levels

[0164] Sediment levels were determined by taking 30 grams of a given coffee sample added to 70 grams of boiling water and shaken for 2 minutes. The sample is then centrifuged for 15 minutes at 10,000 g. After centrifugation the supernatant is decanted off and the sediment re-dissolved with 70 grams of boiling water, shaken for 2 minutes and then centrifuged again under the same conditions as above. This washing process is repeated 3 times for a total of four centrifugation steps. The sediment from the final wash is then freeze dried and then the sediment percentage is related to the starting sample of 30 g (e.g. 1.8 g of sediment represents a 6 wt % insoluble coffee sediment fraction).

TABLE-US-00002 Sediment Sample (wt %) 1 I'Or Intense 5.2 2 Kenco Rich 4.7 3 Carte Noir 3.8 4 Kenco Milicano Americano 11.5 5 Nescafé Gold 4.4 6 Nescafé Azera Americano 9.3 7 Inventive sample Robusta 11.9 8 Inventive sample Colombia Arabica 7.8 9 Inventive sample Centrals Arabica 9.2

[0165] Examples 7, 8 and 9 have been produced in accordance with the method described herein. Examples 1-6 are commercially available products, of which 4, 5 and 6 are products supplemented with added roast and ground coffee additives.

[0166] As can be seen, all commercially available instant coffee products have some level of insoluble coffee sediment fraction. This is expected to be small fragments of coffee cell walls which pass through the extraction system into the coffee extracts. The levels of the insoluble coffee sediment fraction typically increase for those products supplemented with added roast and ground coffee additives.

[0167] As can be seen, the products produced according to the method described herein all have significantly higher levels of insoluble coffee sediment fraction than instant coffee products which have not been supplemented with added roast and ground coffee additives.

Particle Size Distribution

[0168]

TABLE-US-00003 D D Dx Dx Dx [3, 2] [4, 3] (10) (50) (90) sample description μm μm μm μm μm 1 I'Or Intense 1.84 15.7 0.79 2.85 8.49 2 Kenco Rich 2.74 42.3 1.35 3.69 11.2 3 Carte Noir 1.96 4.73 0.897 2.79 8.63 4 Kenco Milicano 4.32 13.1 1.6 11.7 27.3 Americano 5 Nescafé Gold 3.4 42.5 1.1 20.1 102 6 Nescafé Azera Americano 5.91 136 2.16 31.1 197 7 Inventive sample Robusta 3.35 8.22 1.56 4.97 14.6 8 Inventive sample Colombia 2.58 33.9 1.02 4.68 37.2 Arabica 9 Inventive sample Centrals 2.67 33.4 1.03 5.32 34.5 Arabica

[0169] Examples 7, 8 and 9 have been produced in accordance with the method described herein. Examples 1-6 are commercially available products, of which 4, 5 and 6 are products supplemented with added roast and ground coffee additives.

[0170] The sediment quantification method with multiple centrifugation steps allows for a large amount of very fine particles to be recovered.

[0171] Particle size distribution has been measured with the Malvern 3000 after making a 1.5% hot brew of the dried product, for example 3 g of dried product in 200 ml of hot water.

[0172] 3 classes of sediment can be distinguished:

Class 1 L'Or Intense, Kenco Rich and Carte Noir:

[0173] Unimodal distribution D10: <1.5 and D90: <15 μm

[0174] Relative low amount of sediment <5.5% wt

[0175] The very small particle size (such as low D90) perhaps reflects the way in which these particles have escaped from the extraction column into the extract, or mannans which have sedimented in the evaporator.

Class 2: Kenco Milicano, Nescafe Gold and Azera Clearly Differs from Class 1 and 3

[0176] Bi-modal distribution (2 peaks) peak 1 between 1 and 10 μm and peak 2 between 10 and 100 μm.

Class 3: Inventive Samples

[0177] Unimodal distribution but broader distribution than class 1 D10: >1.0 and D90: >15 μm and relative higher amount of sediment such as >7.5% wt.

Carbohydrate Analysis

[0178] The analysis is of mono-saccharides after acid hydrolysis.

TABLE-US-00004 sample description Arabinose Galactose Glucose mannose Total 1 I'Or Intense 0.71 3.45 0.33 61.2 65.7 2 Kenco Rich 0.48 2.43 0.31 50.9 54.1 3 Carte Noir 0.61 3.01 0.43 69.0 73.1 4 Kenco Milicano 1.99 8.88 0.4 36.1 47.4 Americano 5 Nescafé Gold 1.34 6.36 0.37 50.7 58.8 6 Nescafé Azera Americano 1.42 7.06 0.46 60.8 69.7 7 Inventive sample Robusta 0.77 2.47 0.28 42.4 45.9 8 Inventive sample Colombia 0.74 3.08 0.41 54.5 58.7 Arabica 9 Inventive sample Centrals 0.76 3.51 0.45 55.2 59.9 Arabica

[0179] Examples 7, 8 and 9 have been produced in accordance with the method described herein. Examples 1-6 are commercially available products, of which 4, 5 and 6 are products supplemented with added roast and ground coffee additives.

[0180] As can be seen, the insoluble coffee sediment fraction of the products of the invention has a level of Arabinose broadly similar to that of a soluble coffee product which has not been supplemented with roast and ground coffee. Typically it also has a lower level of galactose than a soluble coffee product which has been supplemented with roast and ground coffee

[0181] Without wishing to be bound by theory, it is considered that the high levels of arabinose in the supplemented products is a consequence of the presence of the unextracted coffee material. In contrast, for the inventive products, the levels are lower reflecting the fact that the arabinose has already been extracted into the soluble coffee fraction by the process of the invention.

Sensory Testing

[0182] 2 prototypes of the product of the invention were combined with product from current technology in a ratio of 30 (POI):70 (current product). These were then tested in a set with an additional sample of 100% current technology product. The 3 samples were given to a panel of sensory experts who were then asked to pair the products according to similarities/dissimilarities to the third.

[0183] The results indicate that even at levels of only 30% in a blend with current products the prototype is considered more viscous/dry and powdery—all attributes contributing to mouthfeel/body. The levels correlate directly with the tribology data. More oil means more lubrication which means more mouthfeel/body. The effects are shown in FIG. 3.

Collapse Temperature

[0184] Crystalline products have a well-defined “eutectic” freezing/melting point, this point is called its collapse temperature. When freeze-drying a concentrated coffee extract, the extract is heated up from an initial frozen temperature of about −50° C. under a vacuum. This allows the water content to sublime away. The rate of heating depends on the extract and there is a collapse temperature above which the product will have melt-back and be compromised. The temperature and pressure can then be raised on subsequent cycles until evidence of collapse or melt-back is seen, indicating that the product was too warm. Surprisingly the inventors found that the collapse temperature for several samples of the inventive product were higher than that for their standard coffee products.

Rheological Behaviour of Samples

[0185]

TABLE-US-00005 1 Alta Rica 2 Nescafe Gold Blend 3 Kenco Really Rich 4 Milicano 5 Percol 6 Kenco Really Rich 7 NGC Colombian 8 NGC Central 9 NGC Robusta 10 NGC Brazil

[0186] Samples were prepared with 10 g of coffee dissolved in 40 g of water at 85° C. Full dissolution was achieved with 2 minutes stirring with a 25 mm stirring bar at 150 rpm.

[0187] These samples were tested with simple shear sweeps between shear rates of 0.01-1000 s.sup.−1 using a Discovery HR-2 Rheometer, sample volume 8 ml, with the circulation bath set to minus 4° C. The samples were studied at temperatures of 20 and 65° C. and at concentrations of 1.5 and 20 wt %.

[0188] The data has then been fitted to the Quemada model which develops insights into fluid rheology based on the theory of internal structural units (SUs) suspensions.

[0189] Within concentrated systems the single particles and small flocs may form increasingly larger groups the size of which will be dependent on the shear rates applied.

[0190] Therefore since viscosity (q) is a function of structure (η=f(s)). And the structure is dependent on the levels of shear applied (as increased shear rates will merely act to disperse the macro and meso-structures of the flocs into the individual sub-units), the viscosity can be expressed in terms of the packing fraction/compactness since the more compact the SU's the higher the packing and therefore the more structure (viscosity) there will be.

[0191] This is because the compactness of the SU's will contribute to the level of structure;

[00001]${\eta }_{\infty }={{\eta }_F\left(1-\frac{\varphi }{{\varphi }_{\infty }}\right)}^{-2}$${\eta }_0={{\eta }_F\left(1-\frac{\varphi }{{\varphi }_0}\right)}^{-2}$

[0192] Where η is viscosity and ϕ is the measure of compactness.

[0193] FIG. 2 shows the results from this measurement. In this plot the important information is provided by the intercepts of the plots with the y-axis, representing the initial structure of the test samples. The lines, from top to bottom, are samples 9, 8, 10, 4, 7, 3, 2, 1, 6, 5.

[0194] We can conclude that at 20 wt % (i.e. concentrated samples) at 65° C. (close to temperature of consumption) that samples 4 (Milicano) and 7-10 have significantly higher η.sub.0. This means that from a microstructural perspective at lower shear rates (1 s.sup.−1) which are representative of those during mastication and are reflective of mouthfeel, these samples have more structure relative to the other samples. This implies that at these lower shear rates the compactness of their structural units is higher i.e. better packing of the structural units.

[0195] Tribology of the samples was also observed. “Tribology is the science and engineering of interacting surfaces in relative motion. It includes the study and application of the principles of friction, lubrication and wear.” Therefore the parameter to pay attention to is the μ.sub.max which represents the maximum friction observed for each sample. Since lubrication is indicative of mouthfeel here and a higher μ.sub.max indicates lower rates of lubrication which should translate to lower mouthfeel.

[0196] It was observed that at 65° C. (consumption temp) samples 7, 8 and 10 have significantly lower values for μ.sub.max indicating lower friction and hence higher mouthfeel. The exception is sample 9 (Robusta blend) with lower oil content.

[0197] Unless otherwise stated, all percentages herein are by weight.

[0198] Although preferred embodiments of the invention have been described herein in detail, it will be understood by those skilled in the art that variations may be made thereto without departing from the scope of the invention or of the appended claims.