The samples required for analysis using the optimized and validated method were collected at the Timbilil estate fields (0˚22'S, 35˚21'E) with an altitude of 2180 meters (mean above sea level) of Kericho County in Kenya. Two leaves and a bud were hand plucked fresh for the cultivars of TRFK895/22, Ejulu, TRFK895/7, AHP S15/10, TRFK914/39, TN14-3, D99/10, TRFK6/8, ST832, ST831, ST830, ST543, ST536, TRFK306/4, TRFK306/3, TRFK306/2, TRFK306/1, TRFK91/1, TRFK660/1, TRFK597/4, TRI306/FLD/GB/2, TRI306/FLD/GB/3, TRI306/FLD/ GB/4, TRI306/FLD/GB/5, TRI306/FLD/GB/6 and TRI306/FLD/GB/3. The different orthodox tea grades of BO-OP1, GO-FP, BO-FOF, BO-FP, GO-OP1, GO-FOP, GO-OPA, GO-FOF, GOFBOE, BO-FBOB and BO-FOP were sourced from selected tea processing factories. The cut, tear and curl (CTC) tea grades of PD, D1, F1, PF1 and BP1 were collected from Kipkebe and Matara tea factories. The samples were kept in well labeled aluminium lined khaki bags for preservation before determination.

The DPPH (2,2-Diphenyl-1-picrylhydrazyl) and HPLC grade methanol were purchased from Sigma-Aldrich Chemie GmbH, Germany through Kobian local suppliers. Distilled water was generated in the laboratory. Spectrophotometric measurements were done using a Shimadzu 1800 series spectrophotometer with 1 cm quertz cuvettes.

The hand plucked fresh two leaves and a bud were sorted for high quality material in each of the cultivars above. The leaf was then steamed for 1 min before being microwaved. The microwaved leaf and the different tea grade samples were finely ground using an electric coffee miller (AR40 Molineux) and packed in aluminium lined brown paper sachets ready for use in antioxidant activity analytical determinations.

2.0 ± 0.01 g of the sample was weighed into aluminium dishes and heated in an oven at 103˚C ± 2.0˚C for 4 hours to obtain constant weight. The samples were weighed again and moisture content was then determined by subtracting the final weight from the initial weight, and expressed as a percent. The moisture content obtained was used to determine the sample dry matter content necessary for standardizing the sample weights during analysis.

0.0024 g of DPPH was accurately weighed and transferred into a 100 mL volumetric flask. 80% methanol (v/v) in distilled water was added to above half full mark. The mixture was stirred on a magnetic stirrer until all the DPPH dissolved and then toped to the mark with 80% methanol. The DPPH is sensitive to light and hence must be prepared in light controlled environment. Prepare fresh every day.

Optimization in analytical chemistry involves designating a set of experiments in order to establish the proper conditions for carrying out a method, to achieve the best possible responses and ensure ideal analytical characteristics. For each parameters considered, the focus is to identify the factors for optimization, measure performance under the set conditions, establish trends and visualize improvement. The optimized conditions in this study were, sample weight, extraction solvent concentration, extraction temperature, extraction time, ideal sample extract concentration for DPPH reaction and sample extract-DPPH mixture incubation time.

The precision of the method was investigated by preparing one sample solution containing the target level of analyte. Ten replicates of this sample solution were analyzed with the mean, standard deviation and relative standard deviation (RSD %) determined. From the measured standard deviation (SD) and mean value, precision as relative standard deviation (% RSD) was calculated.

$\%\text{RSD}\left(\text{CV}\right)=\frac{\text{SD}}{\text{Mean}}\ast 100$ (1)

Six different sample concentrations were prepared in triplicate and determined for antioxidant activity using the conditions of the method. The mean anti-oxidant activity in each of the five concentrates was used to generate equations for the regression line and correlation coefficient (R²) in order to establish linear relationship of the method.

Accuracy of the analytical method was determined by preparing quality control (QC) materials of gallic acid and a non-aerated tea sample of TRI 6/8 cultivar in 10 mL of 80% (v/v) methanol in distilled water. The anti-oxidant activity (AA) of the QC materials was accurately determined against the DPPH solution. The QC material was used as the standard to spike the tea sample in order to determine the recovery percent of the method.

$\%\text{Recovery}=\frac{\text{AA sample spiked}-\text{AA sample}}{\text{AA gallic acid}}\times 100$ (2)

LOD and LOQ were obtained from the standard deviation (σ) of the blank response (n = 6) and slope (S) of calibration curve using the formula 3.3 σ/S and 10 σ/S, respectively.

All the samples were evaluated for anti-oxidant activity using the optimized and validated conditions of the method.

Statistical analysis was carried out using SAS^® V 9.1 for windows statistical software. ANOVA was used to determine the means, coefficient of variation (CV) and Least Significance Difference (LSD). The probability limit was set at p ≤ 0.05 significant levels. Results of the parameter determined were expressed as a mean of the triplicate determination.

The influences of sample weight and solvent concentration for the method were tested. The sample weight was optimized on the basis of a 0.06 mM DPPH concentration. From literature search, DPPH concentration of 0.06 mM is extensively used in analytical methods where DPPH is the scavenged free radical [37] . The tea sample weight was evaluated in the range of 0.001 - 1.000 g. It was found out that a 0.0100 ± 0.0001 g sample weight contained in a 10mL extracting solvent and diluted fivefold was optimal for the method. Plank et al., 2012 while working on a collaborative study on food and beverage reaction with DPPH, suggested that for antioxidant property of any sample, the DPPH color loss cannot always be taken to completion; therefore, the weight of sample is adjusted so that approximately 50% of DPPH color is lost [22] . A medium total catechins cultivar of ST 543 was selected as reference material for moderating the DPPH depletion by sample antioxidants to obtain a consistent end point of approximately 50 ± 2AA%. Methanol was used as the extraction solvent due to its high extraction efficiency in extracting both lipophilic and hydrophilic molecules [38] . The solvent concentration range was optimized at 50% - 100% methanol: water (v/v) and 80% methanol was found ideal since it gave a good yield of the antioxidants. A study by Chigayo et al., 2016 concluded that 80% methanol was the best solvent to use for extraction of phytochemicals and antioxidant properties of different solvent extracts of Kirkia wilmsii tubers [39] . Furthermore, elevated temperature of 70^◦C for the reaction mixture gave improved antioxidant activity in a short extraction time of 10 min. Finally, a spectrophotometric scan of DPPH colour change in a DPPH-sample matrix, was done over a period of 90 min ( Figure 2 ) in order to establish the reaction time end point. It was observed that after 30 min, recorded change in absorbance levels had little or no change.

The proposed method was validated as per the Association of Official Analytical Chemists (AOAC) guidelines [40] . Tests for precision, linearity, accuracy, LOD and LOQ for the method were carried out according to earlier adopted procedures in Section 2.7 above.

The precision of the validated method was expressed as percent relative standard deviation %RSD (CV). The results showed an RSD of less than 2% ( Table 1 ). This demonstrates that the present method is sufficiently precise for samples tested and can be used to get good predictive antioxidant activity (AA) results in tea.

Well ground microwaved samples of cultivar ST 543 was accurately weighed to generate concentrations of 0.025, 0.038, 0.050, 0.083, 0.125 and 0.250 mg/g sample on dry weight basis in 10 mL of 80% methanol in distilled water. Antioxidant activity against DPPH was determined using the conditions of the optimized and validated method before a calibration curve of percent antioxidant activity against concentration was generated ( Table 2 ). The linear regression data for the calibration curve showed good linear relationship over the concentration range of 0.025 - 0.250 mg/g sample dry.wt. Additionally, the linear equation was found to be Y = 200.7X + 5.25 (r² = 0.997). This showed that the method met the requirements of validation based on this parameter and hence fit for application during analysis.

Among known polyphenols, gallic acid (GA), a naturally occurring low molecular weight tri-phenolic compound has emerged as a strong antioxidant. Therefore, it was used as the QC material for spiking known concentrations of the tea sample to evaluate recovery of the method. The mean antioxidant activity (n = 3) for pre-analyzed sample, GA standard and GA spiked sample were 48.40, 18.25 and 65.58%respectively ( Table 3 ). The recovery for the method was found to be 94%a demonstration that the method is satisfactory for the intended purpose and adequate for routine analysis.

Limit of detection (LOD) is the lowest amount of analyte in a sample which can be detected but not necessarily quantitated as an exact value whereas the quantitation limit (LOQ) of an individual analytical procedure is the lowest amount of analyte in a sample which can be quantitatively determined with suitable precision and accuracy. The LOD found was 0.0017 mg/mL whereas LOQ was 0.0053 mg/mL. The low levels indicate that the method is capable of detecting significantly low antioxidant activity in tea matrices and tea based products. Moreover, the conditions of the validated method can be used in making decisions in tea related product formulations which would otherwise require very low levels of tea lacing to enhance flavor and strength.

Finely ground tea test samples were weighed (0.0100 ± 0.0001) g into extraction tubes using analytical balance (BL-3200 HL, Shimadzu, Japan). 5.0 mL of hot methanol/water (8:2 v/v) extraction mixture was dispensed into the extraction tubes containing the sample before being stoppered. Mixing was done with the help of a vortex mixer (MV-1000, Taiwan Region) for 1 min and the mixture was placed in a water bath set at 70˚C. The heating continued for 10 min, mixing on the vortex mixer at the intervals of 5 min and 10 min. The extraction tubes from the water bath were removed, allowed to cool to room temperature and the stoppers removed. The tubes were placed in a centrifuge at 3500 rpm for 10 min after which the supernatant was carefully decanted into graduated tubes. The extraction steps were repeated resulting in two extracts. The two extracts were combined and made to 10 mL with cold methanol/water (8:2 v/v). The extracted sample was subjected to DPPH assay.

The sample extract at room temperature (20˚C - 25˚C) was diluted fivefold withmethanol/water (8:2 v/v) and vortexed. A 0.1 mL of the diluted sample aliquot was well mixed with 3.9 mL of 0.06 mM DPPH. The working solution (Ws) was a mixture of 0.1 mL methanol/water (8:2 v/v) and 3.9 mL of 0.06 mM DPPH. Both working solution and sample/DPPH mixture were vortexed for 0.5 min before being incubated at 25˚C for 30 min in the dark. Absorbance (ABS) was then recorded at 517 nm using a Shimadzu 1800 series spectrophotometer. Methanol/water (8:2 v/v) mixturewas used as the blank.

The calculation of the sample antioxidant activity was done using the equation:

$\%\text{AA}=\frac{\text{ABS}\left(\text{Ws}\right)-\text{ABS}\left(\text{sample}\right)}{\text{ABS}\left(\text{Ws}\right)}\times 100$ (2)

With the development of functional foods having beneficial effects on human health, the interest of scientists, consumers and industrialists in raw materials rich in antioxidants has increased considerably over the last few years. Moreover, consumers require more accurate information on the composition of the food which they eat. Natural antioxidants such as phenolic compounds have been reported to possess beneficial bioactivities due to their capacity to act as antioxidants. In tea varietal type, processing and storage are factors that have considerably affected the antioxidant activity in the products.

The procedures of the optimized and validated DPPH method for determination of antioxidant activity in aerated, semi-aerated and non-aerated tea products were used in twenty six (26) selected cultivars of KALRO-TRI germplasm in Kenya. The antioxidant activity of the cultivars is as shown in Table 4 .

Means with the same letters are not significantly different at p < 0.005.

Cultivars TRI 306/1, 306/2, 306/3, 306/4, 91/1, 660/1 and 597/4 are purple pigmented. Furthermore, in addition to the individual phenolic biochemicals of the green pigmented tea cultivars, they are rich in anthocyanins and GHG a special polyphenol. These attributes are replicated in the cultivars of 306/FLD/GB/3, 4, 5, 6 and 13 second generation cultivars of the TRI 306 series. The remaining cultivars are all green pigmented with the cultivars ST536, 543, 830, 831, and 832 having their origin from China. The antioxidant activity ranged from 44.91% - 68.61%. TRFK597/4 had the highest antioxidant activity while 306/FLD/GB/6 gave the lowest levels. Different Kenyan tea cultivars have different polyphenolic composition which impacts on their unique biochemical qualities [41] . Antioxidant activity of any tea product is a function of its total polyphenol content. Thus, the higher the total polyphenol content in a product, the higher the antioxidant activity [42] . Additionally, a study carried out by Cleverdon et al. while comparing total polyphenol content in infusions of black, green, red rooibos, chamomile and peppermint over different steep times gave an observation that the absolute polyphenol content and prediction of total antioxidant capacity was dependent on the type of tea product used [43] . The optimized and validated method unequivocally distinguishes the different tea cultivars in terms of their antioxidant levels and hence can be used in routine analyses.

Differently processed teas were sourced from selected factories across the tea growing regions in Kenya. The orthodox tea grades of BO-OP1, GO-FP, BO-FOF, BO-FP, GO-OP1, GO-FOP, GO-OPA, GO-FOF, GOFBOE, BO-FBOB and BO-FOP were processed as aerated, semi aerated and unaerated according to the optimized manufacture conditions in each of the different factories. Additionally, the cut, tear and curl (CTC) grades of PD, D1, F1, PF1 and BP1 were fully aerated before drying and grading accordingly. All the teas were subjected to the procedures of the optimized and validated DPPH method in determining their antioxidant capacity. The antioxidant activity of the teas is as shown in Table 5 .

Means with the same letters are not significantly different at p < 0.005.

Aerated CTC teas in both Kipkebe and Matara tea factories ranged from 30.37% - 42.18%. The lowest antioxidant activity in black orthodox teas was 37.00% in Itumbe BO-OPA while the highest was 46.22% in Kangaita BO-OP1. Green orthodox teas had generally high antioxidant activity (55.83% - 66.71%) with Kangaita GO-FP being the highest.

Studies have shown that green tea has higher levels of polyphenols when compared to black tea [44] . However, the type and amount of flavonoids they contain differ [45] . For example, green tea contains a much higher amount of epigallocatechin-3-gallate (EGCG), whereas black tea is a rich source of theaflavins [46] . The findings on variation of antioxidant activity with different tea grades in the present study congruent well with work by Erol and Velioglu, 2010 on antioxidant activity of different grades of Turkish black tea [47] . Furthermore, differences in the total antioxidant potential obtained in green, black and earl grey teas, have demonstrated that not only the production method (fermentation or its absence), but also individual technological processes utilized by different manufacturers, as well as the origin of the tea itself, can impact the antioxidant properties of the final product which reaches the consumer [48] .