Tengger Herbal (concocted) coffee formulation with fennel seed (Foeniculum vulgare) and ginger (Zingiber officinale) extract

Robusta Coffee Coffee is a drink suitable for consumption in cold areas. Bromo tourist area is a producer of fennel plants that can be used as a drink to warm the body. This study used Robusta coffee, fennel seed and ginger extract as the basic ingredients for making herbal coffee. The study aimed to obtain a herbal coffee formulation based on fennel seed and ginger extract to enhance and sustain food diversity in Bromo. This study employed the Response Surface Methodology (RSM) method with two factors (i.e. the proportion of Robusta coffee and the proportion of fennel seed and ginger extract) and three responses (i.e. pH, TSS, color L). Selection of the best treatment was carried out by using the effectiveness index method. The best treatment results and optimal solutions for RSM were then tested for antioxidant activity by using DPPH method. The results showed that optimal formulation of coffee (based on RSM) in a proportion of Robusta coffee of 12.50 g (proportion of fennel seed and ginger extract of 10.00 g) has a pH of 5.93, total soluble solids of 12.00 °Brix and L value of26.56 with IC50 value of 75.50 ppm. Whereas, the best organoleptic result was found from the first treatment, in the proportion of 14.00 g Robusta coffee (with fennel seed and ginger extracts of 10.00 g with a pH value of 5.73, total soluble solids of 13.00°Brix, L * value of 25.06 and IC50 of 69.50 ppm).


Introduction
Coffee is very popular to both local and international people due to its distinctive aroma. Indonesia is listed as the fourth largest coffee producer after Brazil, Vietnam and Colombia (Center for Data and Agricultural Information Systems, 2016). In 2013, the International Coffee Organization (ICO) estimated that the world coffee powder needs around 8.77 million tons (ICO, 2015). Based on the Agriculture Data and Information System Center (2016), Indonesian coffee production in 2014 was 643,857 tonnes, with 473,672 tonnes (73.57%) of Robusta coffee.
The Bromo Tengger Semeru National Park (TNBTS) covers a massive area of 800 square km in the center of East Java. This is the largest volcanic region in the province. The most suitable product to be developed in the Bromo area is coffee. Coffee is commonly served hot as a very suitable consumption in highland areas with cold temperatures such as in Bromo Tengger Semeru National Park. Beside coffee, wild plants such as fennel (Foeniculum vulgare) plant is also found in Bromo area, which is traded internationally. Fennel grows wild and massive in the TNBTS area at an altitude of 2100-2200 meters above sea level (Kurniawati and Wahyuningsih, 2014). According to FAO data (2013), the export of fennel seed along with coriander and ginger from Indonesia in 2009 was 612 tonnes and it dropped to 62 tonnes in 2011. Fennel is cultivated as a spice plant and medicinal plant in Indonesia (Sastrawan et al., 2013). Although fennel seed is rich of its antioxidant attribute, the utilisation of fennel remains low. Thus, food product development is required especially for Bromo typical products which using raw materials from fennel plant. Another ingredient which is often used as an enhancer of flavor in a beverage is ginger, due to its distinctive spicy taste from the gingerols. Indeed, ginger offers good benefits for health due to its antioxidant content (Uhl, 2000).
Exploration and formulation of coffee drinks based on fennel seed and ginger extract becomes an effort to utilise and increase food diversity in Indonesia, especially in Bromo

ORIGINAL RESEARCH Open Access
region. This study aimed to obtain a herbal coffee formulation based on fennel seed and ginger extract to enhance and sustain food diversity in Bromo.

Materials
The main material used for this study was Robusta coffee originating from Dampit Malang. Fennel seeds were obtained from fennel plants that grow around the Ngadas Village which is located in the Mount Bromo area. The ginger used was the type of empirical ginger and obtained from the big market in Malang. Materials for the analysis of antioxidants include aquadest, methanol and DPPH.

Methods
This study uses Response Surface Methodology (RSM) with the design of Central Composit Design (CCD) by using Design Expert 7.1.5 Trial Version software. The design of herbal coffee formulation is depicted in Table 1. Factors used in this study were as follows: X1 = proportion of Robusta coffee (g) X2 = proportion of fennel seed and ginger extract (g) Y1 = pH Y2 = TSS Y3 = L* The process of herbal coffee making is accomplished through several stages, started from making: Robusta coffee powder, fennel seed powder, fennel seed and ginger extract, and is then mixed to obtain herbal coffee powder.

Coffee Powder Making Process
The process of making Robusta coffee was as follows: coffee beans were sorted to obtain uniform seeds and to remove impurities or foreign objects contained in the seeds. Coffee beans were washed with clean water to remove dirt attached to the seeds, then dried in the sun to remove water from the washing. Coffee beans were weighed then roasted at 190 o C for 20 minutes, resulted a brownish color coffee bean. Coffee beans were cooled for 30 minutes, then mashed to coffee powder. Coffee powder was sifted using a 60 mesh sieve.

Fennel Seed Extract Making Process
The process of making fennel seed powder was as follows: fennel is obtained from the area of Bromo Tengger Semeru National Park in the form of old fennel seeds. Old seeds are grayish green. Fennel seeds were sorted to remove impurities and obtained uniform fennel seeds. Fennel seeds were washed with clean water, then dried to remove the water. Fennel seeds were weighed and mashed until the fennel powder was obtained

The Process of Making Ginger Extract
The process of making ginger powder is as follows: Ginger is obtained from the big market in Malang. Ginger is washed with clean water until there is no dirt left. The ginger root is peeled clean and then weighed 200 g. The rhizome of ginger is sliced small with a size of 1 cm, then the ginger slices are smoothed using a blender. 300 ml of water and 100 g of fennel powder are added in a blender to be mixed with ginger that has been mashed to become mush. Ginger porridge and fennel are filtered with a filter cloth to separate the pulp, then deposited for 60 minutes to precipitate starch. The ginger juice and fennel obtained are cooked at 100 o C for 20 minutes. In the cooking process 300 g of sugar is added with continuous stirring. After crystallization the cooking is stopped and still stirred until powder forms. Ginger and fennel powder are sifted on 60 mesh and powder that does not pass the sieve is mashed using a blender, then sifting again so that the ginger and fennel extract powder is obtained. The process of herbal coffee making is presented in Fig. 1.

Analysis
Tests carried out on coffee powder, include: pH testing by using a pH meter (Chintya and Fithri, 2015), testing of total dissolved solids (TSS) by using a refractometer (Muchtadi et al., 1989), and L color testing by using color reader (Fardias, 1992). The selection of the best treatment was conducted by organoleptic test (color, aroma, taste) by using index effectiveness method (De Garmo et al., 1984). Analysis of pH, TSS and L color data was accomplished by using Software Design Expert 7.1.5 Trial Version. Meanwhile, the organoleptic test data analysis wass processed with SPSS Stastistics 17.0 software. The results of RSM optimal solution and best selection were tested by antioxidant activity by using the DPPH method.

pH Response
The quadratic model was selected to test pH response. Table 2 presents that pH value of herbal coffee was in the range of 5.730 -6.270. The result of ANOVA analysis on herbal coffee pH variety is illustrated in Table 3. The model used in Table 3 is significant with p value of <0.0001. This result explains that it is less than 0.01% the likelihood of interference with the model. Based on the result of variance analysis, the proportion of Robusta coffee mixed with fennel seed and ginger extract significantly influenced pH value. The acids found in coffee have an effect on pH and contribute to the acidity in coffee (Wayan et al., 2015). An increase in the addition of fennel seed and ginger extract may reduce the pH value of herbal coffee, making it more acidic. This was possibly due to organic acid component (malate, oxalate, lauric acid) in fennel seed and ginger extract (Mujim, 2010). The equation obtained from the model is as follows: Where, Y = pH X 1 = Robusta coffee powder X 2 = fennel seed and ginger extract The results of the study are illustrated in the form of contour plots marking the interaction between the proportion of Robusta coffee and the proportion of fennel seed and ginger extract to pH value of herbal coffee as shown in Fig. 2.

Figure 2. Plot Contour of pH response
The outermost curved line on the contour presents the lowest pH value and the deeper the line shows the highest pH value. Likewise, the blue to red changes indicate that pH value gets higher. In contrast, if the red color goes to blue, pH value gets lower. The color change in Fig.2 illustrates that the addition of the proportion of Robusta coffee has a significant effect on the pH value of coffee. While the addition of fennel seed and ginger extract also shows a significant effect on pH value of coffee. The addition of similar proportion of fennel seed and ginger extract in Robusta coffee will bring significant color changes.

Total Response of Solid Solution
The model chosen for the TSS response is the quadratic model. The results of the analysis of the variance in TSS ANOVA can be seen in Table 4. The analysis of the total dissolved solids of herbal coffee ranges from 8.000 to 13.000 0 Brix. Based on Table 4, it is apparent that the proportion of Robusta coffee mixed with the proportion of fennel seed and ginger extract has a significant effect. Robusta roasted coffee beans have sugar content consisting of galactose, mannose and pentose which have an effect on the value of TSS in coffee. When added sugar in the making of ginger and fennel extract, it will increase the total value of dissolved solids in ginger and fennel extract. This is in accordance with the statement of Osundahusi et al. (2007) pointing out that the addition of sugar can have an effect on higher dissolved solids components.
Ginger has a sugar component consisting of glucose, fructose, galactose and sucrose. Whereas for fennel seed extract, it is thought that it only gives a small contribution to TSS coffee due to other dominant components in fennel seeds such as crude fiber, protein, fat and carbohydrate content (Ibrahim and Ayman, 2013). The equation obtained from the model is as follows: Where, Y = TSS X 1 = Robusta coffee powder X 2 = fennel seed and ginger extract The results of the TSS response are described in the form of contour plots from the response surface that show the relationship between the two factors to the response. The contour plot of the interaction between the proportion of Robusta coffee and the proportion of fennel seed extract and ginger to TSS coffee is shown in Fig. 3. The addition of the proportion of Robusta coffee has a significant effect on TSS coffee. Furthermore, fennel seed and ginger extract also have a significant effect on TSS coffee formulations. Graphical color changes in fennel seed and ginger extract with similar proportion of Robusta coffee, shows significant color changes.

L Color Response
The analysis of the L value in Table 2 shows that the L value of herbal coffee ranges from 25.060 to 26.980. The model chosen for the L color response is the quadratic model. Variance analysis result of ANOVA in L color herbal coffee can be seen in Table 5. The ANOVA result in Table 5 shows that the proportion of Robusta coffee has a significant effect. While the proportion of fennel seed and ginger extract do not significantly influence the response of L color value. The process of color formation in coffee is caused by Maillard reaction when roasting coffee beans. Primadia (2009) explains that maillard reactions involve carbonyl compounds (reducing sugars) and amino groups (amino acids). This is in accordance with the explanation of Fauzi et al. (2016) that the process of roasting the hydrolysed hydrocarbon compounds into carbon elements is followed by the formation of non-volatile melanoidin compounds due to the polymerisation of sugars and amino acids which plays a role in giving color to roasted coffee.
The herbal coffee color is dominated by Robusta coffee powder due to its black color; therefore, concoction of fennel seed and ginger extract does not have a significant effect on coffee color. Fennel seed and ginger extract has a yellowish brown color which comes from the color of ginger and fennel seeds. Ginger has a yellow color and fennel seeds have a green color when they are unripe and turn to grayish green when they are ripe (Aspan, 2008). In the cooking process, fennel seed and ginger extract are oxidised, causing the color of the extract to become yellowish brown. The equation obtained from the model is as follows: Y = 8.610 + 0.892X 1 + 1.144X 2 -0.03X 1 X 2 -0.048X 1 2 -0.068X 2 2 Where, Y = L color X 1 = Robusta coffee powder X 2 = fennel seed and ginger extract The results of the L color response are depicted in the form of contour plots from the response surface marking the relationship between the proportion of Robusta coffee and the proportion of fennel seed and ginger extract to L color of coffee as shown in Fig. 4.

Figure 4. Plot Contour of L Color Response
The outermost curved line on the contour is the lowest L color value and the innermost line shows the highest L color value. The color change from the contour represents the change in L value as shown in Fig. 4. The blue to red color indicates that the L value is higher, whereas if the red one goes to blue, the L value is lower. A low L value indicates that the coffee color is getting darker.
The L color change in the contour plot graph appears when adding Robusta coffee with similar proportion of fennel seed and ginger extract showing a significant change. Thus, the addition of Robusta coffee proportion brings a significant effect on the L color of coffee. Again, fennel seed and ginger extract do not have a significant effect on L color of coffee. This result is based on non-significant color changes in fennel seed and ginger extract with the same proportion of Robusta coffee.

Verification of Optimisation Result
Response optimisation aims to obtain the most optimal results based on the desired limitation of factors as limits of response shown in Table 6. Optimal treatment is obtained in the equal proportion of Robusta coffee and proportion of fennel seed and ginger extracts of 10.000 g (with prediction results of pH response of 5.860; total dissolved solid of 11.940 and L value of 26.100). The choice of the optimal solution is based on the highest desirability value or which is close to 1 (one) and is equal to 0.789. The lowest and highest prediction results produced by the program are used as a limitation of verification result through the conducted research. The results of the verification and the predicted response values can be seen in Table 7.
The results of the optimal solution suggested by the program as in Table 7, are re-verified to ensure the predictions of the computational results which are in accordance with the actual response of the study. The value of the verification results for each response is still in the range of the lowest and highest predictive values; therefore, predictive results can be received with a confidence level of 95%.

Organoleptic Test Result
Organoleptic testing is defined as a test using the human senses as an instrument. The hedonic test (reception) is one of the most preferable type of organoleptic tests. The average results of panelists' preference for herbal coffee are shown in Table 8.

Flavor of Herbal Coffee
According to SNI 01-3542-2004 good coffee quality standards have a normal flavor, namely the flavor of coffee. Panelists' favorite scores on the aroma of coffee using a preference scale of 1-7 (very disliked to very fond) obtained results of 4.67-6.00. The average panelist preference score can be seen in Table 8. Good coffee quality standards according to SNI 01-3542-2004 mentions that it has a normal aroma, which is flavored with coffee. The panelist favorite score on the coffee aroma uses a preference scale of 1-7 (very dislike to very fond) and the results are 4.670-6.000 (neutral to like). According to Buffo and Freire (2004), the aroma of coffee comes from volatile compounds (such as aldehyde, ketone and alcohol) contained in coffee beans. Coffee aroma formation occurs during roasting, which causes physical and chemical changes in coffee beans as well as the formation of a distinctive aroma in coffee.
The distinctive aroma produced by herbal coffee is derived from fennel seed and ginger extract. The zingiberen and zingiberol components of ginger when brewed with coffee will give a distinctive aroma of ginger and coffee. While the distinctive aroma of fennel seeds comes from fenchone compounds. Kurniawati and Wahyuningsih (2014) explained that fenchone is a type of ketone having a distinctive camphor-like odor.

Color of Herbal Coffee
The sense of sight becomes the main tool for qualifying the quality of coffee through the appearance of the color and texture of the liquid surface (Tarigan et al., 2015). The Friedman test results on the color of herbal coffee indicate that there are differences in the coffee color of each treatment.
The main factor influencing the coffee color is derived from the Robusta coffee powder. Fennel seed and ginger extract have no effect on the color of herbal coffee, because the color of coffee powder is already thicker (black); therefore, the black coffee color dominates the color of fennel seed and ginger extract. According to Somporn et al. (2011), the level of roasting of coffee beans has an influence on the coffee bean color as well as the amount and type of volatile compounds which will be produced later. The color change of herbal coffee is related to the protein content in coffee. Melanoidin is a protein derivative compound (the result of the Maillard reaction) which is responsible for the color characteristics of coffee (Tarigan et al., 2015).

Taste of Herbal Coffee
Taste attributes play an important role in determining the quality of a product by using taste buds. The average panelist preference score for coffee flavor obtained is from 4.330 to 5.670. The average panelist preference level for the taste of herbal coffee can be seen in Table 8. Based on the results of the Friedman test conducted on the taste of herbal coffee, it shows that there are real differences in the taste of coffee for each treatment.
The panelists stated that the flavor of herbal coffee in each treatment had different flavor arising from the fennel seed and ginger extract. When brewing volatile ginger compounds, it gives a spicy sensation when consumed. The spicy characteristic and aromatic taste is due to the added content of oleoresin in the ginger extract. The main components which contribute spicy flavor to ginger are gingerol and shogaol. Gingerol content in ginger ranges from 14% to 25% and shogaol is between 2.8% to 7.0%. In addition, fenchone compounds of 13.99% in fennel seed extract give a bit of bitter taste to the coffee produced (Kurniawati and Wahyuningsih, 2014).
The taste of coffee is formed during the roasting process especially from sucrose, amino acids and chlorogenic acid contained in coffee beans. The taste of coffee produced comes from non-volatile organic compounds such as caffeine, chlorogenic acid, trigonelin and dissolved solids content (Fauzi et al., 2016). Panelists state that there is little sour taste in coffee due to acidic compounds in cofee bean such as: formic acid, acetic acid, oxalic acid, citric acid, lactic acid, malic acid (Wayan et al., 2015).

The Best Treatment
Determination of the best treatment is performed by comparing the product value (NP) from each treatment using the effectiveness index. De Garmo et al. (1984), explained that deciding the best treatment was determined by the highest average value of the product. The results of the best treatment in coffee formula is the formula 1 treatment having the highest product value. The value analysis results of the best product selection for herbal coffee can be seen in Table 9.
Based on the organoleptic test result of the 13 treatments, the coffee with the highest product value was coffee with treatment 1 with a product value of 0.920. While the lowest product value is found in treatment 6 with a product value of 0.000. The higher the value of the product shows that herbal coffee is more preferred by the panelists. In contrast, lower result value shows panelists tendency to dislike the herbal coffee.

IC 50 Antioxidant
Antioxidants are antidotes to free radicals which play a role in inhibiting the oxidation process. The antioxidant activity test was carried out on the results of the selection of the best treatment and the optimal solution results using RSM. The results of testing the antioxidant activity of racic coffee can be seen in Table 10.
Based on Table 10. The highest antioxidant activity is found in treatment 1, which is IC 50 value of 69,500 ppm. While the lowest antioxidant activity is found in the RSM optimization treatment which is IC 50 value of 75,500 ppm. A higher addition of the proportion of Robusta coffee powder may produce better IC 50 value of herbal coffee Antioxidant activity of treatment 1 and optimization of RSM are classified as strong antioxidant. This is consistent with the explanation of Zuorro and Roberto (2013) that Robusta coffee has high antioxidants. Antioxidants in coffee beans come from polyphenol compounds such as chlorogenic acid and cafeic acid. Antioxidants in herbal coffee formulation are also influenced by the content found in fennel seed and ginger extract. Ghasemzadeh et al. (2010) stated that the antioxidants in ginger have a relationship with the total phenolic content, as higher total phenols and flavonoids show higher antioxidant activity in ginger. In contrast, the antioxidant content in fennel seeds is classified as weak as a study conducted by Shahat et al. (2011) explained that fennel seeds from vulgare varieties had significantly small levels of antioxidant activity.

Conclusions
The difference in the proportion of Robusta coffee mixed with fennel seed and ginger extract will significantly influence the pH parameters, but does not have a significant effect on the parameters of total dissolved solids and herbal coffee color. The optimal coffee solution obtained from RSM produces a proportion of Robusta coffee of 12.500 g with 10.000 g of fennel seed and ginger extract having a pH of 5.930; total dissolved solids of 12.000 and L values of 26.56 with IC 50 value of 75.500 ppm. While the selection of the best treatment based on organoleptic test in this study is herbal coffee which has the highest product value (NP) of 0.920, which is treatment 1 with a proportion of 14.000 g of Robusta coffee mixed with proportion of fennel seed and ginger extracts of 10.000 g. Herbal coffee treatment 1 has a pH value of 5.730, total dissolved solid of 13.000 and an L color value of 25.06 with an IC 50 value of 69.500 ppm.