Optimasi Hot Water Treatment Menggunakan Response Surface Methodology untuk Mempertahankan Mutu Buah Naga Merah

Abstract

Buah naga daging merah mengandung senyawa betasianin yang merupakan senyawa pigmen berwarna merah-ungu sekaligus sumber senyawa antioksidan. Buah naga tergolong buah nonklimaterik oleh karena itu sebaiknya dipanen saat matang fisiologis. Buah naga memiliki umur simpan pendek setelah dipanen, sehingga perlakuan hot water treatment (HWT) dapat diterapkan untuk mempertahankan mutu buah. Penelitian ini melakukan kombinasi perlakuan HWT dan indeks kematangan buah untuk memprediksi respon mutu menggunakan Response Surface Methodology (RSM). Tujuan penelitian adalah memperoleh model perubahan mutu buah naga sebagai fungsi dari suhu HWT, lama perendaman, dan indeks kematangan menggunakan Response Surface Methodology (RSM) dan menentukan kondisi optimum kombinasi suhu HWT, lama perendaman, dan indeks kematangan dengan menggunakan RSM terhadap mutu buah naga. Buah naga yang telah dipanen diberikan perlakuan HWT sesuai dengan rancangan percobaan, kemudian dikemas dengan menggunakan kotak karton bersekat. Buah selanjutnya disimpan pada suhu 10°C± 3°C selama 21 hari, dan dilakukan analisis mutu setiap tiga hari sekali. Metode penelitian yang digunakan adalah RSM dengan model perancangan Box-Behnken Design (BBD) yang dianalisis dengan menggunakan perangkat lunak Design Expert. Terdapat tiga faktor perlakuan yaitu suhu air panas (40°C, 45°C, dan 50°C), waktu perendaman (5 menit, 10 menit, dan 15 menit), dan indeks kematangan buah (indeks III, indeks IV, dan indeks V). Percobaan ini terdiri dari enam belas unit eksperimental, dengan variabel respon mutu yang diamati meliputi susut bobot, kekerasan, nilai kecerahan, total padatan terlarut (TPT), vitamin C, kandungan betasianin, dan nilai total plate count (TPC). Setiap variabel respon mutu dianalis satu persatu menggunakan analisis ragam ANOVA (Analysis of variance) untuk menentukan signifikansi model. Model matematis dipilih oleh perangkat lunak berdasarkan nilai koefisien determinasi tertinggi (R2). Model matematis yang menunjukkan signifikan dan lack of fit tidak signifikan dipilih untuk menganalisis masing-masing respon mutu. Respon mutu yang memiliki model signifikan selanjutnya dilakukan proses optimasi simultan untuk menentukan kondisi optimum kombinasi suhu HWT, lama perendaman, dan indeks kematangan. Optimasi simultan merupakan proses penentuan satu kombinasi faktor yang menyeimbangkan beberapa respon sekaligus sesuai dengan kriteria yang telah ditetapkan menggunakan fungsi desirability. Hasil analisis menunjukkan bahwa respon susut bobot dan kekerasan memiliki model yang tidak signifikan sehingga tidak dilibatkan dalam optimasi simultan. Adapun respon nilai kecerahan (L*), TPT, vitamin C, dan betasianin memiliki model yang signifikan dengan menggunakan regresi ordo kuadratik. Model matematis untuk memprediksi respon kecerahan (L*) adalah: +29,30-3,06A- 1,33B-1,02C-0,9625AB-0,3375AC+3,13BC+0,7675A2+3,91B2+0,9475C2. Grafik kontur dan respon permukaan menunjukkan bahwa nilai L* tertinggi, yaitu 38,17, diperoleh pada perlakuan suhu 40°C waktu perendaman 5 menit pada indeks kematangan IV. Model matematis respon TPT adalah: +12,80-0,4250A- 0,1625B+0,0625C+0,7750AB-0,0250AC-0,2000BC-0,6000A²+1,07B²+1,22C². Grafik kontur dan respon permukaan menunjukkan nilai TPT tertinggi, yaitu 14,7 °Brix, diperoleh pada perlakuan suhu 40°C waktu perendaman 5 menit pada indeks kematangan IV. Model matematis respon vitamin C adalah: +2,76+0,0450A+0,2675B- 0,5025C+0,3700AB+0,3100AC+0,2700BC-0,1850A2+0,2750B²+0,4550C². Grafik kontur dan respon permukaan menunjukkan nilai vitamin C tertinggi, yaitu 3,46 mg/100 gr, diperoleh pada perlakuan suhu 50°C waktu perendaman 15 menit pada indeks kematangan IV. Model matematis respon betasianin adalah: +51,91+7,64A+2,29B +1,44C+3,45AB+0,9625AC+4,13BC-2,41A2 -5,37B2 -9,55C2 . Grafik kontur dan respon permukaan menunjukkan nilai betasianin tertinggi, yaitu 56,44 mg/L, diperoleh pada perlakuan suhu 50°C waktu perendaman 15 menit pada indeks kematangan IV. Masing-masing respon mutu yang memiliki model signifikan selanjutnya dilakukan optimasi secara simultan dengan menggunakan fungsi desirability. Tujuan optimasi ditetapkan untuk variabel proses dan variabel respon. Variabel proses ditetapkan berada dalam kisaran nilai (in range), sedangkan tujuan optimasi untuk variabel respon ditentukan sebagai berikut: nilai kecerahan (L*) ditetapkan minimum, total padatan terlarut (TPT) ditetapkan dalam kisaran nilai (in range), serta kandungan betasianin dan vitamin C ditetapkan maksimum. Nilai bobot atau nilai kepentingan untuk respon L* dan betasianin ditetapkan paling tinggi, yaitu 5 (+++++), sedangkan respon vitamin C diberikan nilai kepentingan 4 (++++). Dari tahap optimasi diperoleh kombinasi optimum perlakuan HWT yaitu pada perlakuan suhu 50°C waktu perendaman 15 menit pada indeks kematangan IV dengan nilai desirability sebesar 0,855. Model memberikan nilai prediksi untuk dibandingkan dengan nilai aktual dari perlakuan optimum tersebut. Hasil nilai prediksi pada model menunjukkan nilai kecerahan sebesar 28,40, nilai total padatan terlarut (13,5 °Brix), nilai vitamin C (3,53 mg/100 g), dan nilai betasianin (56,44 mg/L). Nilai aktual melalui pengukuran pada perlakuan optimum tersebut menunjukkan TPT dan vitamin C nilainya mendekati dengan nilai prediksi model yaitu 13,1 °Brix dan 3,36 mg/100 g. Hal ini menunjukkan bahwa model cukup akurat memprediksi nilai TPT dan vitamin C, tetapi kurang akurat memprediksi nilai kecerahan dan betasianin. Hasil uji total plate count menunjukkan bahwa perlakuan HWT dengan kombinasi suhu 50°C menghasilkan jumlah total bakteri lebih rendah dibandingkan dengan kombinasi perlakuan suhu 40°C. Begitu pula pada perlakuan dengan menggunakan kombinasi indeks kematangan III dan IV memiliki nilai total bakteri lebih rendah dibandingkan dengan kombinasi perlakuan dengan indeks kematangan V. Kata kunci : box-behnken design, buah naga merah, hot water treatment, optimasi, response surface methodology

Red-fleshed dragon fruit contains betacyanin compounds, which are red– purple pigments and also serve as sources of antioxidants. Dragon fruit is classified as a non-climacteric fruit; therefore, it should be harvested at physiological maturity. Dragon fruit has a short shelf life after harvest; thus, the application of hot water treatment can be used to maintain fruit quality. This study combined hot water treatment and fruit maturity index to predict quality responses using Response Surface Methodology. The objectives of this research were to obtain a model describing changes in dragon fruit quality as a function of hot water treatment temperature, immersion time, and maturity index using Response Surface Methodology, and to determine the optimum combination of hot water treatment temperature, immersion time, and maturity index for maintaining dragon fruit quality. Harvested dragon fruits were subjected to hot water treatment according to the experimental design and subsequently packed in partitioned cardboard boxes. The fruits were then stored at a temperature of 10°C ± 3°C for 21 days, and quality analyses were conducted at three-day intervals. The research method employed was Response Surface Methodology using a Box-Behnken Design, which was analyzed with Design Expert software. Three treatment factors were evaluated: hot water temperature (40°C, 45°C, and 50°C), immersion time (5 minutes, 10 minutes, and 15 minutes), and fruit maturity index (index III, index IV, and index V). The experiment consisted of sixteen experimental units, with observed quality response variables including weight loss, firmness, lightness value, total soluble solids, vitamin C content, betacyanin content, and total plate count (TPC). Each quality response variable was analyzed individually using analysis of variance to determine model significance. The mathematical model was selected by the software based on the highest coefficient of determination value (R2). Mathematical models that showed significant analysis of variance results and non-significant lack-of-fit values were selected for further analysis of each quality response. Quality responses with significant models were subsequently included in the simultaneous optimization process to determine the optimum combination of hot water treatment temperature, immersion time, and fruit maturity index. Simultaneous optimization refers to the process of identifying a single combination of factors that balances multiple responses simultaneously according to predefined criteria using a desirability function. The results showed that the weight loss and firmness responses produced non- significant models and were therefore excluded from simultaneous optimization. In contrast, the responses for lightness value, total soluble solids, vitamin C content, and betacyanin content exhibited significant quadratic regression models. The mathematical model for predicting the lightness response was +29,30-3,06A-1,33B-1,02C-0,9625AB-0,3375AC+3,13BC+0,7675A2+3,91B2 +0,9475C2 . Contour and surface response plots indicated that the highest lightness value, 38,17, was obtained at a treatment temperature of 40°C, an immersion time of 5 minutes, and maturity index IV. The mathematical model for the total soluble solids was +12,80-0,4250A-0,1625B+0,0625C+0,7750AB-0,0250AC-0,2000BC – 0,6000A²+1,07B²+1,22C². Contour and surface response plots showed that the highest total soluble solids value, 14,7°Brix, was obtained at a temperature of 40°C, an immersion time of 5 minutes, and maturity index IV. The mathematical model for the vitamin C response was +2,76 + 0,0450A + 0,2675B – 0,5025C + 0,3700AB + 0,3100AC + 0,2700BC-0,1850A² + 0,2750B² + 0,4550C². Contour and surface response plots indicated that the highest vitamin C content, 3,46 mg/100 g, was obtained at a temperature of 50°C, an immersion time of 15 minutes, and maturity index IV. The mathematical model for the betacyanin response was +51,91 + 7,64A +2,29B+1,44C+3,45AB+0,9625AC + 4,13BC – 2,41A² - 5,37B² - 9,55C². Contour and surface response plots showed that the highest betacyanin content, 56,44 mg/L, was obtained at a temperature of 50°C, an immersion time of 15 minutes, and maturity index IV. Each quality response variable that exhibited a significant model was subsequently subjected to simultaneous optimization using the desirability function. The goals were set for both process variables and response variables. The process variables were constrained to remain within the experimental range, whereas the optimization goals for the response variables were defined as follows: the lightness value was set to be minimized, total soluble solids were constrained to remain within the experimental range, and betacyanin content and vitamin C content were set to be maximized. The weighting values, representing the level of importance assigned to each response, were set at the highest level of five (+++++) for the lightness value and betacyanin content, whereas vitamin C content was assigned an importance level of four (++++). The optimization results indicated that the optimum hot water treatment condition was a temperature of 50°C, an immersion time of 15 minutes, and maturity index IV, with a desirability value of 0,855. The model predicted values of 28,40 for lightness, 13,5 °Brix for total soluble solids, 3,53 mg/100 g for vitamin C, and 56,44 mg/L for betacyanin. Actual measurements under the optimum treatment condition showed that the total soluble solids and vitamin C values were close to the predicted values, namely 13,1°Brix and 3,36 mg/100 g, respectively. These results indicate that the model was sufficiently accurate in predicting total soluble solids and vitamin C content, but less accurate in predicting lightness and betacyanin content. Total plate count analysis showed that hot water treatment at a temperature of 50°C resulted in lower total bacterial counts compared with treatment at 40°C. Similarly, fruits treated at maturity indices III and IV exhibited lower total bacterial counts than fruits treated at maturity index V. Keywords : box-behnken design, dragon fruit, hot water treatment, optimization, response surface methodology