A Journal of the ASEAN Committee on Science Technology Vol 31 No. 2 2014

V JSTdVol. 31, No. 2, 2014 AScJieonucrnea&l ofTethcehnAoSlEoAgyN Committee on ISSN 0217-5460TecDhenJvooeSulcloroienApgnmSaycEleAefnooN&ntr

ABOUT THE ASEAN JOURNAL ON SCIENCE AND TECHNOLOGY FOR DEVELOPMENTThe ASEAN Journal on Science and Technology for Development is a refereed Journal of the ASEAN Committeeon Science and Technology (ASEAN COST). It reports on science and technology policies and programmes, andresearch activities undertaken by COST in support of social and economic development of the ASEAN membercountries.The coverage is focused but not limited to, the main areas of activity of ASEAN COST, namely, Biotechnology,Non-Conventional Energy Research, Materials Science and Technology, Marine Sciences, Meteorology andGeophysics, Food Science and Technology, Microelectronics and Information Technology, Space Applications,and Science and Technology Policy, Infrastructure and Resources Development. ABOUT THE ASEAN COMMITTEE ON SCIENCE AND TECHNOLOGYThe ASEAN Committee on Science and Technology was established to strengthen and enhance the capability ofASEAN in science and technology so that it can promote economic development and help achieve a high qualityof life for its people. Its terms and reference are: ●● To generate and promote development of scientific and technological expertise and manpower in the ASEAN region; ●● To facilite and accelerate the transfer of scientific and technological development among ASEAN countries and from more advanced regions of the world to the ASEAN region; ●● To provide support and assistance in the development and application of research discoveries and technological practices of endogenous origin for the common good, and in the more effective use of natural resources available in the ASEAN region and in general; and ●● To provide scientific and technological support towards the implementation of existing and future ASEAN projects.Information on the activities of ASEAN COST can be obtained at its website http://www.asnet.org DISCLAIMERWhile every effort is made to see that no inaccurate or misleading data, opinion or statement appears in the Journal,articles and advertisements in the Journal are the sole responsibility of the contributor or advertiser concerned. Theydo not necessarily represent the views of the Editors, the Editorial Board nor the Editorial Advisory Committee.The Editors, the Editorial Board and the Editorial Advisory Committee and their respective employees, officersand agents accept no responsibility or liability whatsoever for the consequences of any inaccurate or misleadingdata, opinion or statement. © Copyright 2013: ASEAN Committee on Science and TechnologyNo part of this publication may be reproduced, stored in a retrieval system or transmitted in any form of by any means, without permission in writing from the copyright holder.

Editorial Board Editor-in-Chief Prof Emeritus Dr Md Ikram Mohd Said School of Chemical Sciences and Food Technology,Faculty of Science and Technology, Universiti Kebangsaan MalaysiaEditorial Board MembersMalaysia Assoc. Prof Tan Tin WeeDr Ahmad Ibrahim Department of Biochemistry,Academy of Sciences Malaysia National University of SingaporeProf Abdul Halim Shaari ThailandFaculty of Science, Universiti Putra Malaysia Prof Narongrit Sombatsompop School of Energy, Environment and Materials,Prof Thong Kwai Lin King Mongkut’s University of Technology,Institute of Biological Science, Faculty of ThonburiScience/UMBIO Cluster, Institute of GraduateStudies, University of Malaya Prof Prida Wibulswas President, Shinawatra UniversityBrunei DarussalamRosita Abdullah CambodiaSenior Special Duties Officer, Pal DesMinistry of Development Vice-Rector, Royal University of Phnom PenhAssoc. Prof Zohrah Sulaiman IndonesiaDeputy Vice-Chancellor, Dr Warsito Purwo TarunoUniversiti Brunei Darussalam Minister, Special Advisor for Research and CooperationMyanmarDr Zaw Min Aung Lao PDRDirector General, Department of Technical and Kongsaysy PhommaxayVocational Education, Acting Director General,Ministry of Science and Technology Cabinet Office of the Ministry of Science and TechnologyPhilippinesDr Carol M. Yorobe Keonakhone SaysulianeUndersecretary for Regional Operations, Acting Director General,Department of Science and Technology Department of Information TechnologySingapore VietnamAssoc. Prof Ong Sim Heng Dr Mai HaDepartment of Electrical and Computer Director General,Engineering, National University of Singapore Ministry of Science and Technology

Editorial Advisory PanelBrunei Darussalam Lao PDREddie Sunny Dr Maydom ChanthanasinhDeputy Permanent Secretary, Deputy Minister, Ministry of Science andMinistry of Development Technology, National COST ChairmanMyanmar SingaporeDr Ko Ko Oo Prof Low Teck SengNational COST Chairman, National COST Chairman,Deputy Minister, Ministry of Science and Managing Director, Agency for Science,Technology Technology and ResearchCambodia ThailandDr Om Romny Assoc. Prof Weerapong PairsuwanDirector, Institute of Technology of Cambodia Deputy Permanent Secretary, Ministry of Science and TechnologyPhilippinesDr Graciano P. Yumul MalaysiaUndersecretary for R&D, Dr Noorul Ainur Mohd NurDepartment of Science and Technology National COST Chairman, Secretary General, Ministry of Science,Indonesia Technology and InnovationProf Syamsa Ardisasmita, DEADeputy Minister for Science and Technology VietnamNetwork, National COST Chairman Dr Le Dinh Tien Deputy Minister for Science and Technology, National COST ChairmanEditor/Technical Editor Kanesan SolomalaiEx-Academy of Sciences Malaysia Amirul Ikhzan Amin Zaki Academy of Sciences Malaysia Production Manager Kamariah Mohd Saidin Universiti Putra Malaysia Publisher Universiti Putra Malaysia Press

Contents ASEAN J. Sc. Technol. Dev. Volume 31(2), 2014Drivers and Barriers for Going Green: Perceptions from the Business 49Practitioners in Malaysia 62 83 S-P. Loke, K. Khalizani, S. Rohati and A. Sayaka 90 101Generalized Fuzzy Filters in Ordered Ternary Semigroups M. J. Khan, A. Khan and N. H. SarminEffects of Oil Palm (Elais guineensis) Fruit Extracts on Glucose Uptake Activityof Muscle, Adipose and Liver Cells S. Faez, H. Muhajir, I. Amin and A. ZainahBiodiesel Production from Castor Oil and Its Application in Diesel Engine S. Ismail, S. A. Abu, R. Rezaur and H. SininDevelopment of Suspended Particulate Matter Empirical Equation for TropicsEstuary from Landsat ETM+ Data Z. Razak, A. Zuhairi, S. Shahbudin and Y. Rosnan



ASEAN J. Sci. Technol. Dev.,  31(2): 49 – 61Drivers and Barriers for Going Green: Perceptions from the Business Practitioners in Malaysia S-P. LOKE1*, K. KHALIZANI2, S. ROHATI1 AND A. SAYAKA3The changes of global environmental conditions have placed great challenges to governments andsocieties. While it is not easy for the companies to go green, we need a renewed concern for ourenvironment in order to revive the nation’s economic growth, social cohesion and ecological balances.This article identifies the drivers and barriers for the business industry to adopt green practices. Atotal of 571 business companies from the Perak State participated in this study. Four variables: (1)Regulations (β=0.159, p<0.05); (2) Social responsibility (β=0.201, p<0.05); (3) Pro-environmentalorganizational culture (β=0.389, p<0.01); and (4) Organizational supports (β=0.369, p<0.01) werefound to significantly affect the company’s green initiatives. The results indicated that the maininternal barriers were: it lacked of financial resources (66.2%) and skilled staff (63.9%); whereasthe main external barriers were: the penalty imposed were not severe enough for making any extraefforts (64.8%) and the penalty was light for violation of environmental regulations (63.2%). Thisresearch had implications for the academics, practitioners and policy makers. It provided greaterinsights into the green practices in Malaysian firms. The research findings also urged the localgovernments to greatly enhance regulatory scrutiny on the production and manufacturing industries.Key words: Green practices; going green; environmental strategy; regulations; social responsibility;barriers; industriesToday, all nations — regardless of whether generally measures the climate protectionthey are developed economies or emerging performance of 61 countries aiming to enhanceeconomies — are challenged with highly visible transparency in international climate politics.ecological problems (Hart 2000). Pollution and Malaysia together with countries like Chinaclimate changes have impacted not only the and Singapore, appeared in the bottom-rankedphysical environment, but also the terrestrial group of newly industrialised countries forand marine ecosystem as well as the society at being one of the largest carbon dioxide emitterslarge. While rapid economic development and (Figure 1).population growth are some of the root causes,business organizations are often blamed mainly Although Malaysia has climbed from 55thfor these environmental problems. position in 2013 to 51st this year, among the ASEAN member countries including India, Malaysia has become a more polluted China, Japan and Korean Republic, it has scoredcountry as reported by the Climate Change the lowest position based on the score of CCPIPerformance Index (CCPI) 2014. This CCPI (Figure 2).1 Faculty of Business Management, Universiti Teknologi MARA, Perak Campus, Bandar Seri Iskandar, 36210 Bota, Perak2 Faculty of Business Management, Universiti Teknologi MARA Kedah, Kedah Campus, 08400 Merbok, Kedah3 Institute Darul Rizduan, B-1-9, Greentown Suria, Jalan Dato’ Seri Ahmad Said, 30450 Ipoh, Perak* Corresponding author (e-mail: [email protected]; [email protected])

ASEAN Journal on Science and Technology for Development, 31(2), 2014Figure 1. Climate Changes Performance Index for Newly Industrialized Countries (CCPI 2014).Figure 2. Climate Changes Performance Index for ASEAN countries including India, China, Japan and Korean (CCPI 2014). According to Perry and Singh (2001), dominant in electronics and chemicals hadthe environmental problems of Malaysia higher compliance rate under the respectiveare concentrated in the main centres of regulations (Perry & Singh 2001). Althougheconomic activity such as Kuala Lumpur, as early as in 1974, the regulation frameworkKlang Valley, Penang and Johor. A study was already in place to mitigate the industrialconducted almost two decades ago on 3889 pollution problems, the monitoring andMalaysian manufacturing industries revealed enforcement mechanisms were found to bethat industries with foreign investment limited. 50

S-P Loke et al.: Drivers and Barriers for Going Green: Perceptions from the Business Practitioners in Malaysia Chen, Shih, Shyur, and Wu (2012) opportunities to cut costs, reduce risk,argued that the increased public attention of drive revenues, and enhance intangiblesustainability and environmental issues and value. They build deeper connectionsthose regulations such as Waste Electrical with customers, employees, and otherElectronic Equipment and Eco-design stakeholders. Their strategies revealRequirement for Energy Using Product were a new kind of sustained competitiveestablished. Undoubtedly, the proliferation advantages that we call Eco-Advantageof research on renewable energy and (Esty & Winston 2009, p.14).environmental protection is largely due to theimpacts of climate change and declining fossil As highlighted by the authors of Green tofuel reserves. Hong, Roh and Rawski (2012) Gold, the eco-advantage mindset is a powerfuladded that there is an urgency for firms to motivator to help companies to face challengesbe responsive towards ecological or natural and find new ways to seize advantages. In fact,environment in order to sustain and preserve the Green Wave has swept across the businessthe wealth of natural resources for our next world forcing the companies to react and thesegenerations. trends and forces will continue to evolve. Being eco-efficient is one of the crucial determinants LITERATURE REVIEW to survive in a cost-conscious world. Such restructured landscape requires a new refinedBusinesses have increasingly embraced green business strategy. Some companies and sectorsconcept in their marketing efforts (Raska have responded faster than others. Companies& Shaw 2012). David (2012) added that must be creative to break out of the pack. This isconsumers today are attracted to businesses that because those that do not will struggle to remainpreserve nature’s ecological balance and foster competitive in the marketplace.a clean and healthy environment. Thus, anygreen initiative should be sufficiently visible The sustainability and sustainablefor gaining attention from the customers as development defined by the World Commissionthere is an increased demand in green practices on Environment and Development as “meetingfrom them (Andic, Yurt & Baltacioglu 2012). the needs of the present without compromisingFor example, ElTayeb, Zailani and Jayaraman the ability of future generations to meet(2010) found that customer pressure is one of their own needs” (Loucks, Martens & Cho,the drivers for green purchasing in Malaysia. 2010). Similarly, Kleindorfer, Singhal, and Wassenhove (2005) stated that sustainability The triple focus on green productivity — is the co-ordination of resources in meetingenvironment, quality, and profitability — people’s wants for a satisfying life, besides theis aimed to ensure long-term survival of necessity to respect the bottom line of threethe firms (Diabat & Govidan 2011). More “Ps”, which are planet, people and profit.interestingly, smart companies could actuallyuse environmental strategy to innovate, create Environmental sustainability is related tovalue, and build competitive advantages. the proper and efficient use of natural resourcesThe business world has created numerous over time whereas the firm sustainabilityopportunities of innovation which firms have refers to its ability to gain long-term returns.become the leading of sustainability movement These two concepts are closely related becausein many ways. environmental principles and guidelines can generate green innovations which are in Environmental leaders see their business fact, reducing cost, rising up the productivity through an environmental lens, finding and increasing the companies’ competitive 51

ASEAN Journal on Science and Technology for Development, 31(2), 2014capabilities. Therefore, many different theories four pillars — Energy, Environment, Economyand empirical research have been dedicated and Social. Green Technology is aimed to be theto explore on the implementation and effects key driver in accelerating the national economyof green practices such as eco-design, cleaner and promoting sustainable development inproduction practices and waste management, Malaysia. The Malaysia Green Technologyenvironmental purchasing, and green/ reverse Corporation or known as GreenTech Malaysialogistics. has been striking not only to develop green technology roadmap and standards, but also In Malaysia, new controls on hazardous to promote an environmental friendly livingwaste have been added to the Environmental culture at large.Quality (Amendment) Act 1996 (Perry & Singh(2001). Sani (1999) stated that this amendment We know anecdotally better environmenthas included substantial increases to penalties management strategy enhances thefor a range of environmental offenses as to exert competitiveness of a firm. The environmentalcompliance pressure on the industry. Similarly, mismanaging, however, can damage the brandthe Malaysian government has demonstrated an reputation, destroy its competitiveness andincreased willingness to accept outside influence sometimes can knock off the value of theon environmental performance. For example, company overnight. As such, a more positivethe international criticisms on domestic forestry attitude towards environmental issues, e.g.have resulted in the establishment of the the adoption of green manufacturing wouldNational Timber Certification Centre with the institutionalize the companies’ awareness onindustry partners in order to create the Malaysian environmental concerns which could bringCriteria and Indicators for Sustainable Forest indirect benefits through better quality of theirManagement. Perry and Singh (2001) added manufacturing operations.that the growing public awareness and mediacoverage have placed increased emphasis on Businesses face challenges inthese environmental issues in the Seventh implementation of environmental initiativesMalaysia Five-Year Plan. especially when striking the balance of profitability and corporate social responsibility. Indeed, we must stabilize and reduce the Pressures came from various sources so that theirenvironmental burdens in order to achieve products are environmental friendly (ElTayebsustainability. Under the New Economic Model, et al. 2010). According to Orsato (2006), thethe Malaysian government has embarked on difficult aspect of environmental initiativesthe green initiatives as one of the nation’s is the basic reasoning on environmentalnew economic drivers and transforming the protection because such a move is stronglycountry to become a high income nation by year known as a public good. Although literature2020. While driving Malaysia towards greater demonstrates that effective environmentaleconomic development, these initiatives provide management generate eco-advantages for thea valuable framework on conservation and companies, Esty and Winston (2009) argued thatprotection of the nation’s heritage and natural capturing these advantages require expertiseenvironment. and capabilities to master the whole range of related issues. Often, the company has struggled Meanwhile, the National Green Technology to push for green effort due to private costs ofPolicy was successfully launched by the Prime prevention and clean up which lead to higherMinister of Malaysia on 24 July 2009. The operational costs and thus reduces its industrialNational Green Technology Policy is built on competitiveness (Porter & Van Der Linde 2000). 52

S-P Loke et al.: Drivers and Barriers for Going Green: Perceptions from the Business Practitioners in Malaysia The environmental effort does not always venture into green initiatives. These variablesproduce superior results. They may fail because were identified through a comprehensiveof poor planning, an absence of commitment review of the relevant literature. The researchand not having the right people in the key instrument was adapted from previous studies.roles. Some of the business strategies did A focus group was conducted with six industrynot work well due to focusing on the wrong panels to validate the questionnaire before dataissues, marketplace is misunderstood, the collection.customer responses towards green productsare interpreted wrongly and therefore the In this study, managers and executivesimplementation on the environmental thinking from the manufacturing firms located withinin the business was not successful. the state of Perak were targeted. The firms were selected from the Federation of Malaysian Nevertheless, the consciousness of issues Manufacturers Directory 2013. Manufacturingon environmental sustainability and the need to firms were chosen because operations of thesecomply with the standards are critical to drive industrial companies are frequently and directlycompanies to embark on greater environmental related to the environment — from pollutioncommitment. However, there is a clear lack of control to the most innovative green initiatives.empirical research in emerging economies. Weneed a greater understanding on the awareness of Questionnaires were personally hand-business industry’s environmental management delivered to a sample of 1000 randomlystrategy particularly within the Malaysia selected companies located in the state of Perak.setting. Based on this notion, the objective of The researchers also contacted SME Corp.this study was two-fold: (1) To examine the (Perak office) and Federal of Manufacturingfactors influencing the business firms’ attitude Malaysia (FMM) (Perak Branch) to seektowards environmental commitment, and (2) for their members’ participation in thisTo identify the barriers that inhibit them to go study. Data collection was carried out fromgreen. Understanding the fundamental factors August to October 2013. Based on these 1000for business practitioners to go green is indeed questionnaires originally distributed, a total ofcrucial because these identified key factors 571 of them were found completed and usable,can serve as a springboard to better promote yielding a response rate of 57.1%.the firm’s commitment to go green. It is hopedthat the research findings on sustainable Profiling of the Participating Firmsdevelopment practices can shed lights for thenation especially for the State Governments to There were a total of 571 companiesbetter manage the balance between economic (FMM directory and SMEs in Perak) whichgrowth and ecological sustainability. participated in this research: 561 manufacturer and 10 services companies. Majority of METHODOLOGY the participating companies were from the electrical, machinery and apparatus industriesResearch Design and Sampling (31.2 per cent) followed by the food products and beverage industries (24.5 per cent)The study was designed to test a structural model (Figure 3).whether these variables namely Regulations,Social Resposbility, Customer Pressure, The participating companies mostlyPro-Environmental Organizational Culture concentrated their businesses on both localand Organizational Supports would lead to and international markets with 63.8%. Thea greater level of company’s commitment to 53

ASEAN Journal on Science and Technology for Development, 31(2), 2014 40 31.2 30Percent 20 24.5 10 0 4.0 6.5 4.6 7.4 5.1 4.2 7.7 2.8 0.7 1.2 0.2 Pharmaceuticals Recycling Furniture Motor vehicles, trailers and semi trailers Communication equipment and apparatus Electrical machinery and apparatus Basic metal, machinery and equipment Rubber and plastic products Chemical and chemical products Energy and fuel Paper and paper products Wood products, cork, plating material Food products and beverages Figure 3. Industrial category of the participating firms.participating companies were Malaysian-owned Data Analysisfirms (78.6%) whereas the foreign-owned firmsaccounted for approximately 20 percent, and Before conducting the analysis for structuraljoint ventures accounted for less than 1 percent. modelling, the validity and the relability ofAlso, we found that most of these companies the survey instrument were generated. Datafell into the category of having workers between used for final data analysis was 562 after21 to 100 employees and between 101 to 200 the data with outliners were eliminated. Asemployees with 28.7% and 27.8%, respectively. shown in Table 1, the results indicated thatIn addition, when the responding companies all values for the validity and reliability testswere asked to indicate their certification of ISO were within the acceptable range. Except forfor quality and environmental management, item RG6, all factor loadings for each indicatorit was found that adoption for quality was were >0.5 indicating a high convergentmuch greater as compared to environmental validity. All cronbach alpha values were >0.70management such as ISO 14001 (Environmental demostrating a high consistency of the itemsManagement System (Specifications with used to measure each variable (i.e. regulations,guidance for use). social responsibility, customer pressure, pro-environmental organizational cultures, 54

S-P Loke et al.: Drivers and Barriers for Going Green: Perceptions from the Business Practitioners in Malaysia Table 1. Results for validity and reliability test (n=562).Variables and items Indicators Factor loadings Total items Cronbach Alpha 0.807Regulations RG1 0.69 6 0.811 RG2 0.78 RG3 0.71 0.801 RG4 0.66 0.814 RG5 0.52 RG6 0.49 0.800Social responsibility SR1 0.61 7 0.782 SR2 0.63 SR3 0.64 SR4 0.62 SR5 0.70 SR6 0.59 SR7 0.53Customer Pressure CP1 0.63 6 CP2 0.61 CP3 0.70 CP4 0.67 CP5 0.61 CP6 0.58Pro-environmental PE1 0.52 7organizational cultures PE2 0.65 PE3 0.66 PE4 0.72 PE5 0.73 PE6 0.51 PE7 0.57Organizational support OS1 0.64 7 OS2 0.65 OS3 0.63 OS4 0.55 OS5 0.61 OS6 0.64 OS7 0.50Green responsive GRI1 0.57 5initiatives (GRI) GRI2 0.62 GRI3 0.59 GRI4 0.63 GRI5 0.59Note: GRI measures the company’s willingness and current efforts to go green. 55

ASEAN Journal on Science and Technology for Development, 31(2), 2014organizational supports and green responsive pro-environmental organizational culture,initiative. Thus, it was concluded that the organizational support and customer pressure onsurvey instrument for measuring the variables the company’s proactiveness in environmentalwere valid and reliable. commitment. In this study, multiple fit indices were used: (1) chi-square (χ2); statistics to theStructural Model Evaluation degree of freedom (df); (2) the ComparativeUsing the SPSS AMOS, the structural model Fit Index (CFI); and (3) RMSEA (Root Meanwas generated to examine these criticalfactors: regulations, social responsibility, Square error of approximation) as suggested by Hair et al. (2010). The goodness of fit index measures if the model was adequately fit.Figure 4. The path diagram for structural model (n=562). 56

S-P Loke et al.: Drivers and Barriers for Going Green: Perceptions from the Business Practitioners in MalaysiaThe χ2 statistics to df should be less than 3; (β=0.389, p<0.01); and (4) Organizationalgoodness of fit indexes such as GFI, CFI and support (β=0.369, p<0.01) had significantlyNFI should be as close to 1.00. The error impacted the company’s Green responsiveindexes such as RMSEA and RMR should be initiative (GRI). However, customer pressureas minimum as possible and values ranging was not the driver that motivated the business0.05 to 0.08 were deemed accepted. As shown industries in Perak to adopt go green initiatives.in Figure 4, the results of the structural modelanalysis was deemed to have a reasonable Barriers for Business to Go Greengood fit for the data collected [chi-square(χ2) = 1872.222; degree of freedom (df) = 6338; As illustrated in Table 2, a total of 18 barrierschi-square (χ2)/df = 2.934; CFI = 0.835; that determining the low commitments towardsRMSEA = 0.059]. the environmental protection were identified.Drivers for Business to Go Green These 18 factors were then divided into both internal and external barriers. The internalThe results have revealed that that only four barriers were grouped into three categories:variables: (1) Regulations (β=0.159, p<0.05); (1) Resources, (2) Implementation, and (3)(2) Social responsibility (β=0.201, p<0.05); Attitudes and company cultures. In this study,(3) Pro-environmental organizational culture both human and financial resources were Table 2. Internal and external barriers to go green.Internal Resources Implementation Attitudes andBarriers company culture ●● Excessive financial ●● Unclear leadership constraints. strategy and goals towards ●● Focuses on cost savings. environmental issues. ●● Lack of management ●● Prioritizes on commitment and/or ●● Unclear responsibility commercial needs above supports. regarding who is in charge environmental concerns. of environmental policy/ ●● Lack of engagement/ practice. ●● Complies with minimum commitment from staff. criteria set by the relevant ●● Lack of clarity authority in order to ●● Lack of time and among line managers lower the overall costs. resources to focus on regarding whether they environmental issues. are responsible for ●● Low awareness on environmental issues. environmental issues. ●● Insufficient training regarding the ●● All pro-environmental importance of pro- efforts were way too environmental expensive to carry out. behaviour. ●● Lack of organizational ●● Lack of availability of concern for environmental skilled staff. sustainability.External ●● Insufficient incentives in place to encourage environmental behavior.Barriers ●● Penalty for violation of government environmental legislations was light. ●● Penalty for violation of government environmental legislations was not severe enough for making any extra efforts. 57

ASEAN Journal on Science and Technology for Development, 31(2), 2014cited to be the major barriers for the company in green practices. The penalty for violationto go green. The financial constraint was of environmental legislations was said tothe frequent reason why the company was be light and did not warrant extra effortshaving unfavourable attitude towards greater from the management. Thus, the companyenvironmental management efforts. They often undertake the bare minimum to fulfilstrongly believed that the implementation of legislation requirements. While there werethese green practices did not only cut into their financial initiatives offered by the Malaysianprofits but also required higher maintenance government such as grants and corporatecosts. tax reductions to promote greater green initiatives, the respondents from the focus We found that there were companies that group had highlighted that these incentiveswere more open and willing to go green, but were considered to be a weak motivator.the lack of specialized and technical skills hadpulled them back. The belief of the management Figure 5 illustrates the top five barrierson the derived benefits from environmental that inhibit the company to go green whichpractices would ultimately determine the level were drawn from both the internal and externalof commitment towards green efforts. This barriers. The percentage was derived from theis because such commitment would create scores on “very significant” and “significant”a climate to either deprive or support the when the responding companies were askedenvironmental management e.g. consistency to indicate the extent of these challenges theyof these top management supports, revision of faced in initiating green efforts.company’s priority and allocation of resources.Thus, implementation process could be greatly DISCUSSION AND RECOMMEDATIONSinterrupted without an appropriate corporateculture and full support from the management. Based on the research findings, it was found that the business industries faced both internal The shortcomings in the governmental and external barriers when seeking to addressframework were also found to have hindered their environmental issues and to embark onthe company to have greater commitment green practices. While the results had showedFigure 5. Top five barriers for participating firms to go green (n=571). 58

S-P Loke et al.: Drivers and Barriers for Going Green: Perceptions from the Business Practitioners in Malaysiainternal barriers such as lack of financial agencies to carry out their new environmentalresources and skilled staff were of great functions. Indeed, there appeared to be ahindrance, the participating firms also indicated widely held scepticism in the enforcement ofthat the penalty for violating the environmental environmental laws in our country. We urge theregulations were not severe to justify greater local government to enhance regulatory scrutinyefforts and commitment. In fact, it was cited as on the production and manufacturing industries.the 3rd most significant barriers for their green However, we believe that the business ownersresponsive initiative. and management should be proactive in taking positive environmental actions. They should Previous findings e.g. ElTayeb et al. be more sensitive not only on the awareness(2010) showed that the customer pressure was of legislation but also on the benefits of goinga driver for green purchasing among Malaysian green, both in their business sustainability andcustomers. However, in this study, we found society at large.that such external pressure from the customersdid not motivate the business industries in Meanwhile, we also call to alleviate thePerak to go green. The study by ElTayeb et al. public concern near waste disposal facilities(2010) focused on the green activities in by adopting stricter emission standards,relation to the suppliers meanwhile this current improving monitoring of emissions and payingresearch went beyond the purchasing activity adequate compensation; reduce governmentand looked into the overall green practices of subsidisation of recycling by shifting greaterthe companies. As revealed in the path analysis responsibility to producers and creatingresults, the structural model had depicted adequate economic incentives to reduce wastethe following four important drivers for generation; and to extend environment impactgreen responsive initiative: requirement from assessment (EIA) procedures to better integrateregulations; the company’s social responsibility; environmental concerns in sectorial projectsthe pro-environmental organizational culture; and programme. Finally, we strongly believeand the organizational supports. Since the that wider implementation of programme such“pro-environmental organizational culture” as Eco-Labelling Scheme, MyHijau Labelwas found to have the highest value of beta and GreenTAG could further promote thecoefficient (β), it meant that this factor played environmental friendly living culture and thusthe most significant role in promoting the firms indirectly motivate greater commitment forto adopt green practices. Thus, new approaches businesses to go green.should be developed to involve major internalstakeholders in participating in strategic CONCLUSIONenvironmental planning and defining concretetargets and deadlines for green practices. This study aimed to identify the drivers and barriers for the business industry to implement Regulatory pressure was found to be green practices. It had implications for theanother key driver for green responsive academics, practitioners and policy makers.initiative. Therefore, regulatory programme Firstly, it adds to the body of knowledgeshould be set up to ensure the compliance on green practices particularly within theof environmental requirement and standard. Asian settings. Secondly, the results could beThere was a greater need to strengthen (1) valuable to the managers by providing greaterthe liability legislation in order to better insights into green practices in Malaysiancompensate for damages to the environment in firms. The regulations, social responsibility,line with the ‘polluter pays’ principle, and (2) pro-environmental organizational culture andthe enforcement capacity of local government organizational supports were found to have 59

ASEAN Journal on Science and Technology for Development, 31(2), 2014significant impacts on the company’s green chain management via an analytic networkinitiatives. The environmental management process’, Computers and Mathematics withstandards of different organizations might vary Applications, vol. 64, pp. 2544–2557.in details and they are often subjected to the keyelements of an environmental policy statement, CCIP 2014, Climate change performance index,objectives and targets, implementation <https://germanwatch.org/en/download/ 8599.procedures, internal monitoring, auditing pdf>.and reporting. However, the adoption ofISO 14001 Environmental Management David, FS 2011, Strategic management: conceptSystem (Specifications with guidance for and cases, 14th edn, Pearson Prentice Halluse) was found as one of the most obvious International.determinations on a firm’s commitment towardsadopting green efforts. Diabat, A & Govindan, K 2011, ‘An analysis of the drivers affecting the implementation of Thirdly, the results could also help the green supply chain management’, Resources,government to further plan and enhance current Conservation and Recycling, vol. 55, pp. 659–667.guidelines and policies. In this study, wehad highlighted the key internal and external ElTayeb, TK, Zailani, S & Jayaraman, K 2010, ‘Thebarriers for firms to adopt green initiatives. examination on the drivers for green purchasingThe regulatory pressure is said to be a major adoption among EMS 14001 certified companiesdriver for their environmental performance as in Malaysia’, Journal of Manufacturingit pushes the companies to respond and react. Technology Management, vol. 21, no. 2, pp.We believe that a more stringent monitoring 206–225.from the government on the firm’s complianceto the environmental regulations. Effective law Esty, DC & Winston, AS 2009, Green to gold: howenforcement is equally important to ensure the smart companies use environmental strategyadherence towards the environmental standards to innovate, create value and build competitiveby businesses. advantage, John Wiley & Sons Inc. Date of submission: August 2014 Hair Jr. JF, Black, WC, Babin, BJ & Anderson, RE 2010, Multivariate data analysis: a global Date of acceptance: December 2014 perspective, Pearson, London. ACKNOWLEDGEMENT Hart, SL 2000, ‘Beyond greening: strategies for a sustainable world’, Harvard Business Review,The authors would like to thank the Institute vol. 72, no. 3, pp. 46–52.Darul Ridzuan (IDR) for financially supportingthis research under KRA Perak Amanjaya Hong, P, Roh, JJ & Rawski, G 2012, ‘BenchmarkingProject. sustainability practices: evidence from manufacturing firms’, Benchmarking: REFERENCES An International Journal, vol. 19, nos. 4/5, pp. 634–648.Andic, E, Yurt, O & Baltacioglu, T 2012, ‘Green supply chain: efforts and potential applications Kleindorfer, PR, Singhal, K & Van Wassenhove, for the Turkish market’, Resources, Conservation LN 2005, ‘Sustainable operations management’, and Recycling, vol. 58, pp. 50–68. Production and Operations Management, vol. 14, pp. 482–492.Chen, CC, Shih, H-S, Shyur, H-J & Wu, K-S 2012, ‘A business Strategy selection of green supply Loucks, ES, Martens, ML & Cho, CH 2010 ‘Engaging small- and medium-sized businesses in sustainability’, Sustainability Accounting, Management and Policy Journal, vol. 1, no. 2, pp. 178–200. Orsato, RJ 2006, ‘Competitive environmental strategies: when does it pay to be green?’, California Management Review, vol. 48, no. 2, pp. 27–144. 60

S-P Loke et al.: Drivers and Barriers for Going Green: Perceptions from the Business Practitioners in MalaysiaPerry, M & Singh, S 2001, ‘Corporate environmental Raska, D & Shaw, D 2012, ‘When is going green responsibility in Singapore and Malaysia: the good for company image?’, Management potential and limits of voluntary initiatives’, Research Review, vol. 35, nos. 3/4, pp. 326–347. Technology, Business and Society, United Nations Research Institute for Social Development. Sani, S 1999, ‘Environmental Management Issues and Challenges in the Next MillenniumPorter & Van Der Linde 2000, ‘Green and in Malaysia’, Environmental Management competitiveness: ending the stalement’, Harvard Programme, Universiti Kebangsaan Malaysia, Business Review, vol. 72, no. 3, pp. 131–167. Bangi, Malaysia. 61

ASEAN J. Sci. Technol. Dev.,  31(2): 62 – 82Generalized Fuzzy Filters in Ordered Ternary Semigroups M. J. Khan1, A. Khan1* and n. H Sarmin2In this paper, the concept of ^a, bh-fuzzy generalized filters in an ordered ternary semigroup S isintroduced, where a, b ! \"!, q, ! 0 q, ! / q, with a !! / q . We discussed some fundamentalproperties of ^!, ! 0 qh-fuzzy filters and introduced ^!, ! 0 q h-fuzzy filters. Some relatedproperties of ^!, ! 0 q h-fuzzy filters were provided and the relation between ordinary fuzzy filters,^!, ! 0 qh-fuzzy filters and ^!, ! 0 q h-fuzzy filters were also investigated.Key words: Fuzzy filters; ^a, bh-fuzzy filters; ^!, ! 0 qh-fuzzy filters; ^!, ! 0 q h-fuzzy filtersLehmer (1932) introduced the concept of subgroup. Since then Kuroki introducedternary semigroup. J. Los (1955) studied some the notion of fuzzy bi-ideals in semigroups.properties of ternary semigroups and proved Jun and Song (2006) introduce the generalthat every ternary semigroup can be embedded forms of fuzzy interior ideals in semigroup.in a semigroup. Cayley and Sylsester along Generalization of fuzzy bi-ideals in termswith several other mathematicians, in the 19th of ^!,! 0 qh-fuzzy bi-ideals was given bycentury considered ternary algebraic structures Kazanchi and Yamak (2008). The generalizationand cubic relations. Let S be a non-empty set, a of fuzzy bi-ideals in semigroups was given byfuzzy set, by definition, is an arbitrary mapping Jun et al. (2009). He gave characterizationsf | S $ 60,1@ where 60,1@ is the usual interval of regular ordered semigroups in terms ofof real numbers. The important concept of fuzzy generalized fuzzy bi-ideals. The notionfuzzy set, introduced by Zadeh in 1965 (Yuen of fuzzy filter in an ordered semigroups waset al. 2003), has opened up keen insights and introduced by Kehaypulu and Tsingelis (2002).applications in wide range of scientific fields. An ordered semigroup is a partially orderedMordeson et al. (2003) gave an up-to-date set which is both left and right compatibleaccount of fuzzy sub-semigroups and fuzzy with the semigroup operation. Applications ofideals of a semigroup. It gives applications ordered semigroup are found in the theory ofof fuzzy sub-semigroups in the field of fuzzy sequential machines, computer arithemetics,coding, fuzzy languages and fuzzy finite formal languages, error correcting codes andstate machines. The notions of ^a,bh-fuzzy design of fast adders. An ordered ternarysubgroups was first introduced by Bhakat and semigroup is an ordered semigroup with theDas (1996). They use the “belong to” relation property of associativity of its elements with^!h and “quasi coincidence” relation (q) to respect to the ordered semigroup operation.introduce the concept of ^!,! 0 qh-fuzzy In 2008, Shabir and Khan studied fuzzysubgroup which is, in particular, an important filters in ordered semigroups (Shabir &and useful example of Rosenfeld’s (1971) Khan 2008. Davvaz and Khan (2013) gave1 Department of Mathematics, Abdul Wali Khan University, Mardan, KPK, Pakistan2 Department of Mathematical Sciences, Faculty of Science, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia* Corresponding author (e-mail: [email protected])

M. J. Khan et al.: Generalized Fuzzy Filters in Ordered Ternary Semigroupsthe concept of generalized fuzzy filters in ordered semigroup. They providedifferent characterizations of ordered semigroup by using ^a,bh -fuzzy filters.In this work we will generalize ^a,bh-fuzzy filters in ordered ternary semigroup. We willstudy ^!,! 0 qh-fuzzy filters,^!,! 0 qh-fuzzy left (right, bi) filters and will give differentcharacterizations of ordered ternary semigroup in terms of ^!,! 0 qh-fuzzy filters and^!,! 0 qh-fuzzy left (right, bi) filters.PreliminariesA non-empty set T is called a ternary semigroup if there exists a ternary operationT # T # T \"|^x1, x2, x3h 8 x1 x2 x3 satisfying the following property: For all x1, x2, x3 x4, x5 ! T , 66x1 x2 x3@x4 x5@ = 6x1 6x2 x3 x4@x5@ = 6x1 x2 6x3 x4 x5@@. It is clear thatevery semigroup can be reduced to a ternary semigroup. However, Banach, showed that a ternarysemigroup does not necessarily reduce to a semigroup by giving the following example.Example 2.1 (Chinram & Saelee 2010)T = \"-i, 0, i, is a ternary semigroup but T is not a semigroup under the usual multiplication overcomplex numbers. The next example also shows that T is a ternary semigroup but is not a semigroup.Example 2.2 (Chinram & Saelee 2010)Z- is a ternary semigroup but is not a semigroup under the multiplication over integers. Los(1955)showed that every ternary semigroup can be embedded into a semigroup. A partiallyordered semigroup T is called an ordered ternary semigroup if for all x1, x2, x3 x4 ! T ,x1 # x2 \" x1x3x4 # x2x3x4, x3x1x4 # x3x2x4, x3x4x1 # x3x4x2.Example 2.3 (Chinram & Saelee 2010)^Z-, $, #h is an ordered ternary semigroup. Throughout this paper T will denote orderedternary semigroup, unless otherwise specified. Let A be a non-empty subset of T, wedenote ^A@ = \"x ! T | x # a for some a ! A,. For non-empty subsets A,B,C, we denoteABC = \"abc | a ! A, b ! B, c ! C ,. A non-empty subset A of T is called a ternary subsemigroupT if ^A@ 3 A and AAA 3 A.Definition 2.4 (Chinram & Saelee 2010)A non-empty subset F of T is called a left filter of T if it satisfies: (1) xyz 3 F , for all x, y, z ! F . (2) for all x, y ! T , x # y and x ! F $ y ! F , (3) for all x, y, z ! T , xyz ! F $ z ! F .F is called a right filter of T if it satisfies conditions (1) and (2) of Definition 2.4 and (4) for all x, y, z ! T , xyz ! F $ y ! F . 63

ASEAN Journal on Science and Technology for Development, 31(2), 2014F is called lateral filter T, if it satisfies conditions (1) and (2) of Definition 2.4 and (5) for all x, y, z ! T , xyz ! F $ x ! FF is called a filter of T if it satisfies conditions (1) and (2) of Definition 2.4 and (6) for all x, y, z ! T , xyz ! F $ x, y, z ! F .Definition 2.5 (Chinram & Saelee 2010)A fuzzy subset μ of T is called a fuzzy left filter (resp. fuzzy right filter) of T if for all x, y, z ! Twe have: (1) x # y \" n^xh # n^ yh, (2) n^xyzh $ min \"n^xh, n^ yh, n^zh,, (3) n^xyzh # n^zh^resp. n^xyzh # n^xhh.Let F be a non-empty subset of T. Then the characteristic function |F of F is defined by |F | T \" 60, 1@ |x 7 |F ^xh = '10 if x ! F otherwisefor all x ! T . Obviously, a nonempty subset F of T is a filter if and only if the characteristic function |F ofF is a fuzzy filter of T.(α,β)-fuzzy FiltersLet T be an ordered ternary semigroup. A fuzzy subset μ of T defined by: n | T \" 60,1@, y 7 n^xh = '0t if y # x otherwiseis called an ordered fuzzy point with support x and value t and is denoted by xt . Pu and Liu (1980)gave the meaning of symbols xt an where a ! \"!, q ! 0 q, ! / q,. The symbol xt ! n (resp. xt qn )means that n^xh $ t (resp. n^xh + t 2 1) and say that the fuzzy subset xt belongs to (resp. quasico-incident with) μ. To say that xt an means that the relation xt an does not hold. Let μ be a fuzzysubset of T, then for t ! ^0,1@ the set U^n; th = \"x ! T | n^xh $ t, is called the level subset of μ. 64

M. J. Khan et al.: Generalized Fuzzy Filters in Ordered Ternary SemigroupsTheorem 3.1 (Los 1955)A fuzzy subset μ of T is a fuzzy filter of T if and only if the level subset U(μ;t) is a fuzzy filter of T. Proof. Let x, y ! T , x # y and x ! U^n; th for t ! ^0,1@. Then n^xh $ t . Since μ is fuzzyfilter of T, we have n^xh # n^ yh, n^ yh $ t which implies that y ! U^n;th. Let t ! ^0,1@,x,y,z ! U^n,th. Then n^xh $ t , n^ y $ th, n^zh $ t . Since n is fuzzy filter of T, we haven^xyzh $ min \"n^xh,n^ yh,n^zh, = t , xyz ! U^n; th . Let x,y,z ! T , xyz ! U^n; th for somet ! ^0,1@. Then n^xyzh $ t . Since n is a fuzzy filter of T we have n^xyzh # min \"n^xh,n^ yh,n^zh,,that is, n^xh $ t , n^ y $ th, n^zh $ t which implies that x,y,z ! U^n; th. Hence U^n; th is a fuzzyfilter of T. Conversely, assume that for a fuzzy subset n of T and t ! ^0,1@, the level subset U^n; th ofn is a filter of T. Let x, y, z ! T , x # y . If n^xh = 0 then n^ yh $ 0 = n^xh. If n^xh = t^! ohthen x ! U^n;th. Since U^n;th is a filter of T, we have y ! U^n;th, n^ yh $ t = n^xh. Letx, y, z ! T . Suppose that there exist t ! ^0,1@ such that n^xyzh $ t 2 min \"n^xh,n^ yh,n^zh,.Then xyz ! U^n; th but x, y, z g U^n; th which is a contradiction to the fact that U^n; th is a filterof T. Hence n^xyzh # min \"n^xh,n^ yh,n^zh,. Let x, y, z ! T . Let there exist t ! ^0,1@ such thatn^xyzh 1 t # min \"n^xh,n^ yh,n^zh,. Then x,y,z ! U^n; th but xyz g U^n;th which is again acontradiction. Hence n^xyzh $ min \"n^xh,n^ yh,n^zh,. Thus n is a fuzzy filter of T.Example 3.2 (Davvas & Khan 2012)Let T = \"a,b,c,d,e,f , be a set. Define a ternary operation ^ h on T as ^abch = a ) b ) c, wherethe binary operation \" ) \" and order relation \" # \" are defined as follows: *a b c d e f aa b b d e f bb b b b b b cb b b b b b dd b b d e f ee f f e e f ff f f f f fand#|= \"^a,ah,^b,bh,^c,ch,^d,d h,^e,eh,^ f,f h,^a,d h,^a,eh,^b,eh,^d,eh,^b,f h,^c,f h,^c,eh,^ f,eh,. 65

ASEAN Journal on Science and Technology for Development, 31(2), 2014 Then ^T,(), #h is an ordered ternary semigroup. The subset \"a,d,e, and T are filters of T.Define a fuzzy subset n of T as under: n^eh = 0.8, n^d h = 0.7, n^ah = 0.6, n^bh = 0.4, n^ch = 0.3, n^ f h = 0.5.Then, U^n; th = ]]]]]Z[]\]]\"a,Tzd,e,, if 0 1 t # 0.3 if 0.5 1 t # 0.6 if 0.8 1 t # 1Then by Theorem 3.1, n is a fuzzy filter of T.Theorem 3.3Let n be a fuzzy subset of T. Then U^n; th is a fuzzy filter of T for all t ! ^0,5,1@ if and only nsatisfies the following conditions: (1) ^6x,y ! T h^max \"n^ yh,0.5, $ n^xhwith x # yh, (2) ^6x,y,z ! T h^max \"n^xyzh,0.5, $ min \"n^xh,n^ yh,n^zh,h, (3) ^6x,y,z ! T h^max \"n^xh,n^ yh,n^zh,0.5, $ n^xyzhh. Proof. Suppose that U^n; th is a filter of T for all t ! ^0.5,1@. Let us assume that there existx,y ! T with x ≤ y and r ! ^0,5,1@ such that max \"n^ yh,0.5, 1 n^xh = r . Then x ! U^n;rhbut y g U^n, rh, a contradiction. Hence Condition (1) is true. Let there exist x, y, z ! T ands ! ^0,5,1@ such that max \"n^xyzh,0.5, 1 min \"n^xh,n^ yh,n^zh, = s. Then x,y,z ! U^n; sh butxyz g U^n; sh, a contradiction. Hence Condition (2) is true. Now let us suppose that there x, y, z ! Tand w ! ^0,5,1@ such that max \"n^xh,n^ yh,n^zh,0.5, 1 n^xyzh = w . Then xyz ! U^n; wh butx, y, z g U^n; wh which is a contradiction to the fact that U^n; th is a filter of T for all t ! ^0.5,[email protected] Condition (3) is true. Conversely, assume that n satisfies Conditions (1), (2) and (3). Letx,y,z ! U^n; th . T h e n n^xh $ t , n^ yh $ t a n d n^zh $ t . N o w f o r m C o n d i t i o n ( 2 ) ,max \"n^xyzh,0.5, $ min \"n^xh,n^ yh,n^zh, = t 2 0.5 w h i c h m e a n s t h a t n^xyzh $ t , s oxyz ! U^n; th . Let x, y ! T with x # y and x ! U^n; th then n^xh $ t . Therefore fromCondition (1) we have for some t ! ^0.5,1@, max \"n^ yh,0.5, $ n^xh $ t 2 0.5 which means thatn^ yh $ t , that is y ! U^n; th. Let x, y, z ! T such that xyz ! U . Then n^xyzh $ t , we have fromCondition (3), max \"n^xh,n^ yh,n^zh,0.5, $ n^xyzh $ t 2 0.5 so that n^xh $ t , n^ yh $ t , that is,x,y,z ! U^n; th. Let x, y, z ! T such that x,y,z ! U^n; th. Then n^xh $ t , n^xh $ t and n^xh $ t . 66

M. J. Khan et al.: Generalized Fuzzy Filters in Ordered Ternary SemigroupsNow from Condition (2), we have, max \"n^xyzh,0.5, $ min \"n^xh,n^ yh,n^zh, = t 2 0.5 whichimplies that n^xyzh $ t , that is, xyz ! U^n; th. Hence U^n; th is a filter of S.In view of Definition 2.2, we now give the following definition.Definition 3.4A fuzzy subset n of an ordered ternary semigroup is called an ^!,!h-fuzzy filter of T if it satisfies: (1) ^6x, y ! Sh^6t ! ^0,1@h^x # y,xt ! n \" yt ! nh, (2) ^6x, y, z ! Sh^6t, r, s ! ^0, 1@haxt, yr, zs ! n \" ^x, y, zhmim\"t,r,s, ! nk, (3) ^6x, y ! Sh^6t ! ^0,1@h^^xyzht ! n \" xt ! n, yt ! n, zt ! nh.In view of Definition 3.4, we have the following theorem.Theorem 3.5A fuzzy subset n of T is a fuzzy filter of T if and only if it satisfies the Conditions (1), (2) and (3)of Definition 3.4. Proof. Using Definition 2.5, the proof is straight forward. If n is a fuzzy subset of T defined by n^xh # 0.5 for all x ! T , then the set \"xt | xt ! / qn,is empty. So the case when a =! / q is excluded from the following definition.Definition 3.6A fuzzy subset n of T is called an ^a,bh-fuzzy filter of T where a !! / q, if it satisfies thefollowing conditions: (1) ^6x, y ! T h^6t ! ^0,1@h^x # y, xt an \" yt bnh, (2) ^6x, y, z ! T h^6t, r, s ! ^0, 1@haxt an, yr an, zs an \" ^xyzhmin\"t,r,s,bnk, (3) ^6x, y ! T h^6t ! ^0, 1@h^^xyzht ! n \" xt bn, yt bn, zt bnh.Theorem 3.7Let n be a non-zero ^a,bh-fuzzy filter of T. Then the set n0 = \"x ! T | n^xh 2 0, is a filter of S. Proof. Let x, y, z ! n0 . Then n^xh 2 0 , n^ yh 2 0 , n^ zh 2 0 . So xn^xh ! n , yn^yh ! nand zn^zh ! n . Since n is an ^a, bh -fuzzy filter of T, we have ^xyzhmin\"xn^xh, yn^yh , zn^zh,bnf o r e v e r y b ! \"!, q, ! 0 q, ! / q, . T h u s n^xyzh $ min \"xn^xh , yn^yh , zn^zh, 2 0 s o t h a t 67

ASEAN Journal on Science and Technology for Development, 31(2), 2014xyz ! n0 . Let x, y ! T such that x # y and x ! n0 . Then n^xh 2 0. Suppose that x^ yh = 0.Then xn^xhan but yn^yhbn for every a ! \"!, q, ! 0 q, and b ! \"!, q ! 0 q, ! / q, , acontradiction. Note that x1 qn but y1 bn for every b ! \"!, q, ! 0 q, ! / q, a contradiction.Hence n^ yh 2 0 , which implies that y ! n0 . Let x, y, z ! T such that x, y, z ! n0 . Thenn^xh 2 0, n^ yh 2 0, n^ zh 2 0. Suppose that n^xyzh = 0. Then xn^xhan, yn^yhan , zn^zhan but^xyzhmin\"xn^xh,yn^yh,zn^zh,bn for every a ! \"!, q, ! 0 q, and b ! \"!, q, ! 0 q, ! / q, a contradiction.Note that x1 q n, y1 q n, z1 q n but ^xyzhmin\"1,1,1, = ^xyzh\"1,bn for every b ! \"!, q, ! 0 q, acontradiction. Hence n^xyzh 2 0 so that xyz ! n0 . Let x, y, z ! T such that xyz ! n0 . Thenn^xyzh 2 0 . Suppose that n^xh = 0 or n^ yh = 0 or n^zh = 0 . Let a ! \"!, q, ! 0 q,, thenxyzn^xyzhan but xn^xhbn or zn^zhbn for every b ! \"!, q, ! 0 q, ! / q,. Note that ^xyzh1 q n butx1 bn or y1 bn or z1 bn, for every b ! \"!, q, ! 0 q, ! / q, a contradiction. Thus n^xh 2 0,n^ yh 2 0, n^ zh 2 0. Hence n0 is a filter of T.^!, ! 0 qh-fuzzy FiltersIn this section, we introduce !, ! 0 q-fuzzy filter in ordered ternary semigroup and will characterizeordered ternary semigroup in terms of !, ! 0 q -fuzzy filters.Definition 4.1A fuzzy subset n of T is called an !, ! 0 q -fuzzy filter of T if it satisfies the following conditions: (1) ^6x, y ! Sh^6t ! ^0, 1@h^x # y, xt ! n \" yt ! 0 q nh, (2) ^6x, y, z ! Sh^6t, r, s ! ^0, 1@haxt ! n, yr ! n, zs ! n \" ^xyzhmin\"t,r,s, ! 0 q nk, (3) ^6x, y, z ! Sh^6t ! ^0, 1@h^^xyzht ! n \" xt ! 0 q n, yt ! n 0 q, zt ! 0 q nh.Example 4.2Consider the ordered semigroup as given in Example 3.2, and define a fuzzy subset n by: n^eh = 0.8, n^d h = 0.7, n^ah = 0.6, n^ch = 0.4, n^bh = 0.3, n^ f h = 0.45 then n is an !, ! 0 q -fuzzy filter of T. (i) n is not ^!, !h-fuzzy filter of T, since d0.68 ! n, c0.38 ! n and f0.4 ! n but ^dcf hmin\"0.68, 0.38, 0.4, = b0.38 g n. (ii) n is not ^q, !h -fuzzy filter of T, since c0.7 q n , d0.68 q n and f0.78 qn but ^cdf hmin\"0.7, 0.68, 0.78, = b0.68 g n. (iii) n is not ^!, qh-fuzzy filter of T, since a0.58 ! n , b0.3 ! n and c0.38 ! n but ^abchmin\"0.58, 0.3, 0.38, = b0.3 q n. 68

M. J. Khan et al.: Generalized Fuzzy Filters in Ordered Ternary SemigroupsTheorem 4.3A fuzzy subset n of T is an ^!,! 0 qh-fuzzy filter of T if and only if it satisfies the followingconditions: (1) ^6x,y ! T h^x # y, n^ yh $ min \"n^xh,0.5,h, (2) ^6x,y,z ! T h^n^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5,h, (3) ^6x,y,z ! T h^min \"n^xh, n^ yh, n^zh, $ min \"n^xyzh,0.5,h. Proof. Suppose that n is an !, ! 0 q -fuzzy filter of T. Let x, y ! T such that x # y . Ifn^xh = 0, then n^ yh = n^xh. Let n^xh ! 0 and assume that there exists some t ! ^0,1@ suchthat n^ yh 1 t # min \"n^xh,0.5, . If n^xh 1 0.5 then n^ yh 1 t # n^xh which implies thatxt ! n but yt g n, a contradiction. If 0.5 # n^xh, then n^ yh 1 0.5 # n^xh, that is x0.5 ! nbut y0.5 g n, again a contradiction. Hence n^ yh $ min \"n^xh, 0.5,. Let x, y, z ! T . Let usassume that n^xh = 0 , or n^ yh = 0 , n^zh = 0 , then n^xyzh $ min \"n^xh, n^ yh, n^zh,0.5, .Suppose that n^xh ! 0 , n^yh ! 0 and n^zh ! 0 and assume on contrary that theree x i s t x,y,z ! T a n d s ! ^0,1@ s u c h t h a t n^xyzh 1 s # min \"n^xh, n^ yh, n^zh,0.5, . I fmin \"n^xh, n^ yh, n^zh 1 0.5 , t h e n n^xyzh 1 s # min \"n^xh, n^ yh, n^zh, w h i c h i m p l i e sthat xs, ys, zs ! n but ^xyzhs g n a contradiction. If 0.5 # min \"n^xh, n^ yh, n^ zh, thenn^xyzh 1 s # 0.5 # min \"n^xh, n^ yh, n^ zh, w h i c h i m p l i e s t h a t x0.5, y0.5, z0.5 ! n b u t^xyzh0.5 g n , again a contradiction. Hence n^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5, for allx,y,z ! T . Let x,y,z ! T . If n^xyzh = 0. Then min \"n^xh, n^ yh, n^zh,0.5, $ min \"n^xyzh,0.5,.L e t n^xyzh ! 0 a n d a s s u m e t h a t t h e r e e x i s t x, y, z ! T a n d r ! ^0,1@ s u c h t h a tmin \"n^xh, n^ yh, n^zh,0.5, 1 r # min \"n^xyzh, 0.5, . I f n^xyzh 1 0.5 , t h e n min \"n^xh,n^ yh, n^ zh, 0.5, 1 r # n^xyzh which implies that ^xyzhr ! n but xr, yr, zr g n, a contradiction.If 0.5 # n^xyzh , then min \"n^xh, n^ yh, n^zh,0.5, 1 r # 0.5 # n^xyzh which implies that^xyzh0.5 ! n b u t x0.5, y0.5, z0.5 g n a g a i n a c o n t r a d i c t i o n . H e n c e min \"n^xh, n^ yh,n^zh, $ min \"n^xyzh,0.5, for all x,y,z ! T . Conversely, assume that for a fuzzy subset n of T, Conditions (1), (2), and (3) hold.Let x, y ! T such that x # y and xt ! n for some t ! ^0,1@ . Then n^xh $ t . Then byCondition (1), we have, n^ yh $ n^xh, that is yt ! n. Hence yt ! 0 q n. Let x, y, z ! T andt, r, s ! ^0, 1@ such that xt, yr, zs ! n. Then n^xh $ t , n^ yh $ r and n^ zh $ s. By Condition(2) of the hypothesis, we have, n^xyzh $ min \"n^xh, n^ yh, n^zh,0.5, $ min \"t, r, s, 0.5, . Ifmin \"t, r, s, 0.5, # 0.5, then n^xyzh $ min \"t, r, s, then ^xyzhmin\"t,r,s, ! n. If min \"t, r, s, 2 0.5,t h e n n^xyzh + min \"t, r, s, 2 0.5 + 0.5 = 1 w h i c h i m p l i e s t h a t ^xyzhmin\"t,r,s, ! n . H e n c e^xyzhmin\"t,r,s, ! 0 q n. Let x, y, z ! T such that ^xyzht ! n for all t ! ^0, 1@. Then n^xyzh $ t . 69

ASEAN Journal on Science and Technology for Development, 31(2), 2014By condition (3), we have min \"n^xh, n^ yh, n^zh, $ min \"n^xyzh, 0.5, $ min \"t, 0.5, . Ift # 0.5 then min \"n^xh, n^ yh, n^ zh, $ t which implies that xt, yt, zt ! n . If t 2 0.5 , thenmin \"n^xh, n^ yh, n^zh, + t 2 0.5 + 0.5 = 1, that is, n^xh + t 2 1, n^ yh + 1, n^zh + t 2 1. Thusxt q n, yt q n, zt q n. Hence xt !0q n, yt !0q n, zt !0q n. Consequently n is an ^!, ! 0 qh-fuzzyfilter of T.Remark 4.4A fuzzy subset of an ordered ternary semigroup is an ^!,! 0 qh-fuzzy filter of T if and only if itsatisfies Conditions (1), (2) and (3) of Theorem 4.3.Remark 4.5By the above remark every fuzzy filter of T is an ^!,! 0 qh-fuzzy filter of T. However the converseis not true in general.Example 4.6Consider the ordered ternary semigroup as given in Example 3.2, and define a fuzzy subset n ofT as follows: n^eh = 0.8, n^d h = 0.7, n^ah = 0.6, n^ch = 0.4, n^bh = 0.3, n^ f h = 0.45, then n is an ^!,! 0 qh-fuzzy filter of S. But n is not an ^a, bh-fuzzy filter of S, where a ! \"!, q, ! 0 q, and b ! \"!,q, ! 0 q, ! / q, as shown in Example 4.2. Using Theorem 4.3, we have the following characterization of fuzzy filters of ordered ternarysemigroups. The proof of Theorem 4.7 is easy and is therefore omitted.Theorem 4.7Let T, $, # be an ordered ternary semigroup and z ! F 3 T . Then F is a filter of T if and only ifthe characteristic function |F of F is an ^!,! 0 qh-fuzzy filter of T.Theorem 4.8Let F be a filter of T and n a fuzzy subset of S such that n^xh |= '$0 0.5 if x ! F if x ! T \ FThen,(a) n is a ^q, ! 0 qh-fuzzy filter of T,(b) n is an ^!,! 0 qh-fuzzy filter of T. Proof. (a) Let x, y ! S , x # y and t ! ^0, 1@ such that xt qn, then x ! F . Since x # y ,we have, y ! F. If t # 0.5 then n^ yh = 0.5 $ t implies that yt ! n . If t 2 0.5 then 70

M. J. Khan et al.: Generalized Fuzzy Filters in Ordered Ternary Semigroupsn^ yh + t 2 0.5 + 0.5 = 1 implies that yt q n. Hence yt !0q n. Let x, y, z ! T and t, r, s ! ^0, 1@such that xt q n, yr q n and zs q n. Then x, y, z ! F implies that xyz ! F . If min \"t, r, s, # 0.5then n^xyzh $ 0.5 $ min \"t, r, s, implies that ^xyzhmin\"t,r,s, ! n . If min \"t, r, s, 2 0.5 thenn^xyzh + min \"t, r, s, 2 0.5 + 0.5 = 1 implies that ^xyzhmin\"t,r,s,q n. Hence ^xyzhmin\"t,r,s, ! 0 q n.Let x, y, z ! T such that ^xyzht q n where t ! ^0,1@, we have xyz ! F then x, y, z ! F . Ift # 0.5 then min \"n^xh, n^xh, n^xh, $ 0.5 $ t and we have n^xh $ t , n^ yh $ t , n^zh $ t .Hence xt, yt, zt ! n. If t # 0.5 then we have min \"n^xh, n^xh, n^xh, + t 2 0.5 + 0.5 = 1 thatis n^xh + t 2 1, n^ yh + t 2 1, and n^ zh + t 2 1, it follows that xt q t , yt q t and zt q t . Hencext !0qt , yt !0qt and zt !0qt . Proof. (b) Let x, y ! S , x # y and t ! ^0, 1@ such that xt ! n , then x ! F . Sincex # y , we have, y ! F . If t # 0.5 then n^ yh = 0.5 $ t implies yt ! n . If t # 0.5 thenn^ yh + t 2 0.5 + 0.5 = 1 implies yt q n . Hence yt !0q n . Let x, y, z ! T and t, r, s ! ^0, 1@such that xt ! n, yr ! n and zs ! n. Then x, y, z ! F implies xyz ! F . If min \"t, r, s, # 0.5t h e n n^xyzh $ 0.5 $ min \"t, r, s, i m p l i e s ^xyzhmin\"t,r,s, ! n . I f min \"t, r, s, 2 0.5 t h e nn^xyzh + min \"t, r, s, 2 0.5 + 0.5 = 1 implies ^xyzhmin\"t,r,s,q n. Hence ^xyzhmin\"t,r,s, ! 0 q n. Letx, y, z ! T such that ^xyzht ! n where t ! ^0,1@, we have xyz ! F then x, y, z ! F . If t # 0.5then min \"n^xh, n^xh, n^xh, $ 0.5 $ t and we have n^xh $ t , n^ yh $ t , n^zh $ t . Hencext, yt, zt ! n . If t 2 0.5 then we have min \"n^xh, n^xh, n^xh, + t 2 0.5 + 0.5 = 1 that isn^xh + t 2 1, n^yh + t 2 1 and n^zh + t 2 1, it follows that xtq t , ytq t and ztq t . Hencext qt , yt !0qt and zt !0qt . It is important to note that in Theorem 4.8 we impose a condition on the fuzzy subset n of T.Without the condition,n^xh |= '$0 0.5 if x ! F if x ! T \ Fn may not be a ^q, ! 0 qh-fuzzy filter of T as shown in Example 4.2(ii). In the following theorem we give the condition an ^!,! 0 qh-fuzzy filter to be an ^!, !h-fuzzyfilter of T.Theorem 4.9Let n be an ^!,! 0 qh-fuzzy filter of T such that n^xh 1 0.5 for all x ! T . Then n is an^!,!h-fuzzy filter of T. Proof. Assume that n is an ^!,! 0 qh-fuzzy filter of T such that n^xh 1 0.5 for all x ! T .Let x, y ! T with x # y and xt ! n for some t. Then n^xh $ t . Since n is an ^!,! 0 qh-fuzzyfilter of T, using Theorem 4.3, we have n^ yh $ min \"n^xh, 0.5, = n^xh $ t . Hence yt ! n. Let 71

ASEAN Journal on Science and Technology for Development, 31(2), 2014x, y, z ! T such that xt ! n, yr ! n, zr ! n for some t, r, s ! ^0, 1@. Then n^xh $ t , n^ yh $ rand n^xh $ s and so by hypothesis n^xyzh $ min \"n^xh, n^ yh, n^xh, 0.5, $ min \"t, r, s, .H e n c e ^xyzhmin\"t,r,s, ! n . L e t x, y, z ! T s u c h t h a t ^xyzht ! n t h e n n^xyzh $ t s omin \"n^xh, n^ yh, n^xh, $ min \"n^xyzh, 0.5, = n^xyzh $ t . H e n c e xt ! n , yt ! n , zt ! n .Consequently n is an ^!,!h-fuzzy filter of T.Theorem 4.10A fuzzy subset n of T is an ^!,! 0 qh -fuzzy filter of T if and only if the setU^n; th |= \"x ! T n^xh $ t, is a filter of T for all t ! ^0,0.5@. Proof. Suppose that n is an ^!,! 0 qh-fuzzy filter of T. Let x, y ! U^n; th with x # yand x ! U^n; th where t ! ^0,0.5@. Then n^xh $ t and it follows from Theorem 4.3(1)t h a t n^ yh $ min \"n^xh, 0.5, $ min \"t, 0.5, = t a n d s o y ! U^n;th . L e t x, y, z ! U^n; thfor some t ! ^0,0.5@. Then n^xh $ t , n^ yh $ t , n^zh $ t . It follows from Theorem 4.3(2)that n^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5, $ min \"t, 0.5, = t and so xyz ! U^n; th . Now letxyz ! U^n; th for some t ! ^0,0.5@. Then n^xyzh $ t and it follows from Theorem 4.3(3)that min \"n^xh, n^ yh, n^zh, $ min \"n^xyzh, 0.5, $ min \"t, 0.5, = t and so n^xh $ t , n^ yh $ t ,n^zh $ t . Hence x, y, z ! U^n; th. Conversely, assume that for a fuzzy subset n of T and for all t ! ^0,0.5@ the setU^n; th |= \"x ! T n^xh $ t, is a filter of T. Let there exist x, y ! T with x # y such thatn^ yh 1 min \"n^xh, 0.5,, then we can choose t ! ^0,0.5@ such that n^ yh 1 t # min \"n^xh, 0.5,,then xt ! n but yt g n, a contradiction. Hence n^ yh $ min \"n^xh, 0.5, for all x, y ! T withx # y . If there exist x, y, z ! T such that n^xyzh 1 min \"n^xh, n^ yh, n^zh, 0.5,, then we canchoose r ! ^0,0.5@ such that n^xyzh 1 r # min \"n^xh,n^ yh,n^zh,0.5,. Then x, y, z ! U^n; rhbut xyz g U^n; rh, a contradiction. Hence n^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5,. Finally let thereexist x, y, z ! T such that min \"n^xh, n^ yh, n^zh, 1 s # min \"n^xyzh, 0.5,, then we can chooses ! ^0,0.5@ such that min \"n^xh, n^ yh, n^zh, 1 s # min \"n^xyzh, 0.5,. Then xyz ! U^n; sh butx, y, z g U^n;sh, again a contradiction. Hence min \"n^xh, n^ yh, n^zh, $ min \"n^xyzh, 0.5, forall x, y, z ! T and t ! ^0,0.5@. Consequently n is an ^!,! 0 qh-fuzzy filter of T.For any fuzzy set n of T and for any t ! ^0,0.5@, we denote Q6n; t@ = \"x ! T xt q n, and 6n@t = \"x ! T xt ! 0 q n,.It is obvious that 6nt@ = U^n; th , q6n; t@. 72

M. J. Khan et al.: Generalized Fuzzy Filters in Ordered Ternary SemigroupsTheorem 4.11A fuzzy subset n of T is an ^!,! 0 qh-fuzzy filter of T if and only if for all t ! ^0,1@, 6n@t ^! 0his a filter of T. Proof. Assume that n is an ^!,! 0 qh-fuzzy filter of T. Let t ! ^0,1@ be such that 6n@t ^! 0hand x, y ! T , x # y, x ! 6n@t . Then n^xh $ t or n^xh + t 2 1. Since n is an ^!,! 0 qh-fuzzy filterof T and x # y, we have n^ yh $ min \"n^xh, 0.5,. We consider the following cases: (1) n^xh $ t , (2) n^xh + t 2 1.(1) If t 2 0.5, then n^ yh $ min \"n^xh, 0.5, $ min \"t, 0.5, = 0.5 and n^ yh + t 2 0.5 + 0.5 = 1implies yt q n. Hence y ! q^n; th 3 6n@t . If t # 0.5 then n^ yh $ min \"n^xh, 0.5, $ min \"t, 0.5, = timplies yt ! n, i.e., y ! U^n; th 3 6n@t .( 2 ) I f t 2 0.5 , t h e n n^ yh $ min \"n^xh, 0.5, $ min \"1 - t, 0.5, = 1 - t , t h a t i s ,n^ yh + t 2 1 a n d s o yt q n , i . e . , y ! q^n; th 3 6n@t , h e n c e y ! 6n@t . I f t # 0.5 , t h e nn^ yh $ min \"n^xh, 0.5, $ min \"1 - t, 0.5, = 0.5 $ t , that is, yt ! n and so y ! U^n; th 3 6n@t ,hence y ! 6n@t . Thus in both cases we have yt ! n . Let x, y, z ! 6n@t . Then n^xh $ t ,or n^xh + t 2 1, n^yh $ t or n^yh +2 1, n^zh $ t or n^zh + t 2 1. We consider thefollowing cases: (1) n^xh $ t , n^ yh $ t and n^zh $ t , (2) n^xh $ t , n^ yh $ t and n^zh + t 2 1, (3) n^xh $ t , n^ yh + t 2 1, and n^zh $ t , (4) n^xh + t 2 1, n^ yh $ t and n^zh $ t , (5) n^xh + t 2 1, n^ yh + t 2 1 and n^zh $ t , (6) n^xh + t 2 1, n^ yh + t 2 1 and n^zh + t 2 1.(1) If n^xh $ t , n^ yh $ t and n^zh $ t . Then since n is an !,! 0 q n-fuzzy filter of T, we havefrom Theorem 4.3(2),n^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5, $ min \"t, 0.5, = 't0.5 if t 2 0.5 if t # 0.5and so ^xyzht q n or ^xyzht ! n i.e., xyz ! q^n; th , U^n; th = 6n@t .73

ASEAN Journal on Science and Technology for Development, 31(2), 2014(2) If n^xh $ t , n^ yh $ t and n^zh $ t . Then from Theorem 4.3(2), if t 2 0.5 n^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5, 2 min \"t, t,1 - t, 0.5, = '1t - t if t # 0.5and so ^xyzht q n or ^xyzht ! n i.e., xyz ! q^n; th , U^n; th = 6n@t .(3) If n^xh $ t , n^ yh + t 2 1 and n^zh $ t . Then using Theorem 4.3(2), we haven^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5, 2 min \"t,1 - t, t, 0.5, = '1t - t if t 2 0.5 if t # 0.5and so ^xyzht q n or ^xyzht ! n i.e., xyz ! q^n; th , U^n; th = 6n@t .(4) n^xh + t 2 1, n^ yh $ t and n^zh $ t . Then we haven^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5, 2 min \"1 - t, t, t, 0.5, = '1t - t if t 2 0.5 if t # 0.5Hence ^xyzht q n or ^xyzht ! n i.e., xyz ! q^n; th , U^n; th = 6n@t .(5) If n^xh + t 2 1, n^ yh + t 2 1 and n^zh $ t then we have from Theorem 4.3(2),n^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5, 2 min \"1 - t,1 - t, t, 0.5, = '1t - t if t 2 0.5 if t # 0.5and we have ^xyzht q n or ^xyzht ! n i.e., xyz ! q^n; th , U^n; th = 6n@t .(6) If n^xh + t 2 1, n^ yh + t 2 1 and n^zh + t 2 1. Thenn^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5, 2 min \"1 - t,1 - t,1 - t, 0.5, = '01.-5 t if t 2 0.5 if t # 0.5and thus we have n^xyzh 2 1 - t or n^xyzh 2 0.5 $ t , i.e. ^xyzht q n or ^xyzht ! n and hencexyz ! q^n; th , U^n; th = 6n@t . Let x, y, z ! 6n@t then n^xyzh $ t or n^xyzh + t 2 1. Assume that n^xyzh $ t then since n isan ^!,! 0 qh-fuzzy filter of T, we have by Theorem 4.3(3),min \"n^xh, n^ yh, n^zh, $ min \"n^xyzh, 0.5, $ min \"t, 0.5, = '0t .5 2 1 - t if t # 0.5 if t 2 0.574

M. J. Khan et al.: Generalized Fuzzy Filters in Ordered Ternary Semigroupsso that x, y, z ! U^n; th , q^n; th = 6n@t . Suppose that n^xyzh + t 2 1, then we havemin \"n^xh, n^ yh, n^zh, $ min \"n^xyzh, 0.5, $ min \"1 - t, 0.5, = '01.-5 t if t 1 0.5 if t $ 0.5and thus xyz ! q^n; th , U^n; th = 6n@t . Consequently 6n@t is a filter of T. Conversely, assume that for all t ! ^0,1@, the set 6n@t ^! 0h is a filter of T. If thereexist x0, y0 ! T with x0 # y0 such that n^ y0h 1 min \"n^x0h, 0.5, then we can chooset0 ! ^0, 0.5@ such that n^ y0h 1 t0 # min \"n^x0, 0.5h, . It follows that x0 ! U^n; t0h 3 6n@t0and so x0 ! 6n@t0 . Thus n^x0h $ t0 or n^x0h + t0 2 1. Since x0 # y0 , x0 ! 6n@t0 and 6n@t0is a filter of T, it follows that y0 ! 6n@t0 , i.e. n^ y0h $ t0 or n^ y0h + t0 2 1, a contradiction.Hence n^ yh $ min \"n^xh, 0.5, for all x, y ! T with x # y . If There exist a, b, c ! T such thatn^abch 1 min \"n^ah, n^bh, n^ch, 0.5, t h e n n^abch 1 t1 # min \"n^ah, n^bh, n^ch, 0.5, f o rsome t1 ! ^0, 0.5@. Hence a, b, c ! U^n; t1h 3 6n@t1 and so abc ! 6n@t1 . Thus n^abch $ t1 orn^abch + t1 2 1, a contradiction. Hence n^xyzh $ min \"n^xh, n^ yh, n^zh, 0.5, for all x, y, z ! T .Now if there exist d, e, f ! T such that min \"n^d h, n^eh, n^ f h, # min \"n^def h, 0.5, , then\"n^d h, n^eh, n^ f h, 1 t2 # min \"n^def h, 0.5, for some t2 ! ^0, 0.5@. Then def ! U^n; t2h 3 6n@t2and so d, e, f ! 6n@t2 . Thus n^d h $ t2 or n^d h + t2 2 1, n^eh $ t2 or n^eh + t2 2 1 andn^ f h $ t2 or n^eh + t2 2 1, a contradiction. Hence min \"n^xh, n^ yh, n^zh, $ min \"n^xyzh, 0.5,for all x, y, z ! T . Consequently n is a fuzzy filter of T. Let n be a fuzzy subset of T and J = \"t t ! ^0,1@ and nt ^! 0h is a filter of T ,. WhenJ = ^0, 1@, n is an ordinary filter of T (Theorem 3.1). When J = ^0, 0.5@, n is an ^!,! 0 qh-fuzzyfilter of T (Theorem 4.10). Consider J = \"t t ! ^0,1@ and nt ^! 0h is a filter of T ,. In the nexttopic we will answer the following questions: (1) If J = ^0, 0.5@, what kind of fuzzy filters of T will be n? (2) If J^r,s@, ^r,s ! ^0,1@h will n be a kind of fuzzy filter of T or not?Definition 4.12A fuzzy subset n of T is called an ^!, !0 q h-fuzzy filter of T if it satisfies the followingconditions: (F1) ^6x, y ! T h^6t ! ^0,1@h^x # y, xt ! n $ yt !0 q nh, (F2) ^6x, y, z ! T h^6t, r, s ! ^0, 1@ha^xyzhmin\"t,r,s, ! n $ xt !0 q n, yr ! 0 q n, zs ! 0 q nk, (F3) ^6x, y, z ! T h^6t ! ^0, 1@h^xt ! n, yt ! n, zt ! n $ ^xyzht !0 q nh.75

ASEAN Journal on Science and Technology for Development, 31(2), 2014Example 4.13 (Davvaz & Khan 2012)Consider the ordered ternary semigroup T as given in Example 3.2 and define the fuzzy subsetn of T as follows: n^eh = 0.6, n^d h = 0.5, n^ah = 0.4, n^bh = 0.2, n^ch = 0.3, n^ f h = 0.0.35Then, U^n; th = \[]]]]]]]Z]\"zTa,d,e, if 0 1 t # 0.2 if 0.35 1 t # 0.4 if 0.6 1 tThen n is an ^!, !0 q h-fuzzy filter of T.Theorem 4.14A fuzzy subset n of T is an ^!, !0 q h-fuzzy filter of T if and only if it satisfies the followingconditions: (F4) ^6x,y ! T h^max \"n^ yh, 0.5, $ n^xhh, (F5) ^6x, y, z ! T h^max \"n^xyzh, 0.5, $ min \"n^xh, n^ yh, n^zh,h, (F6) ^6x, y, z ! T h^min \"n^xh, n^ yh, n^zh, 0.5, $ n^xyzhh. Proof: (F1) $ (F4). Let x, y ! T , x # y such that max \"n^ yh, 0.5, 1 n^xh = t . Thent ! ^0.5, 1@, yt ! n but xt ! n. By (F1), we have xt q n. Then t # n^xh and n^xh + t # 1 whichimplies that t # 0.5, a contradiction. Hence (F4) is true.(F4) $ (F1). Let x, y ! T , t ! ^0,1@ such that yt ! n then n^ yh 1 t . (i) If n^ yh $ n^xh thenn^xh 1 t and so xt ! n. Hence xt !0 q n (ii) If n^ yh 1 n^xh then by (F4) we have n^xh # 0.5.If xt ! n then t # n^xh # 0.5 which implies that xt q n. Hence, xt !0 q n.(F2) $ (F5). If there exist x, y, z ! T such that max \"n^xyzh, 0.5, 1 min \"n^xh, n^ yh, n^zh, = s.Then s ! ^0.5, 1@, ^xyzhs ! n but xs ! n, ys ! n, zs ! n. By (F2), we have xs q n, ys q n,zsq n. Then ^s # n^xh and n^xh + s # 1h, ^s # n^yh and n^yh + s # 1h and ^s # n^xh andn^xh + s # 1h which implies that s # 0.5, a contradiction. Hence (F5) is valid.( F 5 ) $ ( F 2 ) . L e t ^xyzhmin\"t,r,s, ! n t h e n n^xyzh 1 min \"t, r, s, . ( i ) I fn^xyzh $ min \"n^xh, n^ yh, n^zh, then min \"n^xh, n^ yh, n^zh, 1 min \"t, r, s, and so n^xh 1 t ,n^ yh 1 r , n^zh 1 s. It follows that xt ! n, yr ! n, zs ! n. Thus xt !0 q n, yr !0 q n. (ii)If n^xyzh 1 min \"n^xh, n^ yh, n^zh, , then by (F5), we have 0.5 $ min \"n^xh, n^ yh, n^zh, .I f xt ! n o r yr ! n o r zs ! n , t h e n t # n^xh # 0.5 , r # n^ yh # 0.5 , s # n^xh # 0.5which implies that xt q n, yr q n, zs q n. Thus xt !0 q n, yr !0 q n, zs !0 q n. 76

M. J. Khan et al.: Generalized Fuzzy Filters in Ordered Ternary Semigroups(F3) $ (F6). If there exist x, y, z ! T such that min \"n^xh, n^ yh, n^zh, 0.5, 1 n^xyzh = rthen r ! ^0.5, 1@, xr ! n , yr ! n and zr ! n but ^xyzhr ! n . By (F3), we have ^xyzhr q n .Then r # n^xyzh and n^xyzh + r # 1. Which implies that r # 0.5, a contradiction. Hence, (F6)is valid.(F6) $ (F3). Let xt ! n, yt ! n and zt ! n. Then n^xh 1 t , n^ yh 1 t and n^zh 1 t , that ismin \"n^xh, n^ yh, n^zh, 1 t . (i) If min \"n^xh, n^ yh, n^zh, $ n^xyzh, then n^xyzh 1 t and so^xyzht ! n. It follows that ^xyzht !0 q n (ii) If min \"n^xh, n^ yh, n^zh, 1 n^xyzh then we havefrom (F6), 0.5 $ n^xyzh. If ^xyzht ! n, then t # n^xyzh # 0.5 which implies that ^xyzht q n. Thus^xyzht !0 q n.Lemma 4.15Let n be a fuzzy subset of S. Then U^n; th^! 0h is a filter of T if and only if it satisfies the conditions(F4) and (F6) of Theorem 4.14.Theorem 4.16A fuzzy subset n of T is an ^!, !0 q h-fuzzy filter of T if and only if U^n; th^! 0h is a filter of Tfor all t ! ^0.5,1@. Proof. Follows from Theorem 4.14 and Lemma 4.15.Based on Shabir and khan (2008), in next definition we give the concept of fuzzy filters with thetresholds. Yuan, Zhang and Ren gave the definition of a fuzzy subgroup with tresholds in Shabirand khan (2008), which is a generalization of Rosenfeld’s (1971) subgroup, and Bahakat and Das’sfuzzy subgroup.Definition 4.17Let r, s ! ^0,1@ and r # s. Let n be a fuzzy subset of an ordered ternary semigroup T. Then n iscalled a fuzzy filter with threshold (r,s) of T. If it satisfies the following conditions: (1) ^6x, y ! T h^x # y, max \"n^ y, rh, $ min \"n^xh, s,h, (2) ^6x, y, z ! T h^max \"n^xyzh, r, $ min \"n^xh, n^ yh, n^zh, s,h, (3) ^6x, y, z ! T h^max \"n^xh, n^ yh, n^zh, r, $ min \"n^xyzh, s,h. If n is a fuzzy filter with thresholds (r,s) of T, then n is an ordinary fuzzy filter of T ifr = 0, s = 1 and n is an ^!,! 0 qh-fuzzy filter of T if r = 0 and s = 0.5. In the next theorem,we characterize fuzzy filters with thresholds ^r,s@ of T, by their level filters. 77

ASEAN Journal on Science and Technology for Development, 31(2), 2014Theorem 4.18A fuzzy subset n of an ordered ternary semigroup T is a fuzzy filter with thresholds ^r,sh of T ifand only if nt ^! 0h is a filter of T for all t ! ^r, s@. Proof. Suppose that n be a fuzzy filter with thresholds ^r,sh of T. Let x, y ! Twith x # y and x ! nt then n^xh $ t and from (1) of Definition 4.17, we havemax \"n^ yh, r, $ min \"n^xh, s, $ min \"t, s, $ t 2 r s o t h a t n^ yh $ t a n d h e n c e y ! nt .Let x, y, z ! nt , then n^xh $ t , n^ yh $ t , n^zh $ t then by (2) of Definition 2.17(2), weh a v e , max \"n^xyzh, r, $ min \"n^xh, n^ yh, n^zh, s, $ min \"t, s, = t 2 r a n d s o n^xyzh $ timplies that xyz ! nt . Let xyz ! nt . Then n^xyzh $ t and we have from Definition 4.17(3),min \"n^xh, n^ yh, n^zh, r, $ min \"n^xyzh, s, $ \"t, s, = t 2 r w h i c h i m p l i e s t h a t n^xh $ t ,n^ yh $ t , n^ zh $ t , so that x, y, z ! nt . Conversely, assume that n be a fuzzy subset of T such that nt ^! 0h is a filter of T for allt ! ^r,s@. Let x,y ! T with x # y such that max \"n^ yh, r, 1 min \"n^xh, s, = t then t ! ^r,s@,n^ yh 1 t and x ! nt . Since x # y and nt is a filter of T, we have y ! nt implies that n^ yh $ t ,a contradiction. Hence \"n^ yh, r, $ min \"n^xh, s, for all x, y ! T with x # y . Let thereexist x, y, z ! T such that max \"n^xyzh, r, 1 min \"n^xh, n^ yh, n^zh, s, = t then t ! ^r,s@ ,n^xyzh 1 t , x ! nt , y ! nt , z ! nt . Since nt is a filter of T, we have xyz ! nt whichimplies that n^xyzh $ t , a contradiction. Hence max \"n^xyzh, r, $ min \"n^xh, n^ yh, n^zh, s,f o r a l l x, y, z ! T . F i n a l l y l e t u s s u p p o s e t h a t t h e r e e x i s t x, y, z ! T s u c h t h a tmax \"n^xh, n^ yh, n^zh, r, 1 min \"n^xyzh, s, = t then t ! ^r,s@, n^xh 1 t , n^ yh 1 t , n^zh 1 tand xyz ! nt . Then since nt is a filter of T, we have, x, y, z ! nt which implies that n^xh $ t ,n^ yh $ t , n^zh $ t which is a contradiction. Hence max \"n^xh, n^ yh, n^zh, r, $ min \"n^xyzh, s,for all x, y, z ! T . Consequently n is a fuzzy filter of T.Remark 4.19(1) By Definition 4.17, we have the following conclusion: If n is a fuzzy filter of T with thresholds^r,s@, then we have: (i) n is an ordinary fuzzy filter when r = 0 and s = 1, (ii) n is an ^!,! 0 qh-fuzzy filter when r = 0 and s = 0.5, (iii) n is an ^!, !0 q h-fuzzy filter when r = 0 and s = 1.(2) By Definition 4.17, we can define other types of fuzzy filters of an ordered ternarysemigroup T such as fuzzy filter with thresholds ^0.5, 0.6@ of T, fuzzy filter with thresholds ^0.4, 0.8@of T, etc. 78

M. J. Khan et al.: Generalized Fuzzy Filters in Ordered Ternary Semigroups(3) However, the fuzzy filter with thresholds of T may not be an ordinary fuzzy filter, may not bean ^!,! 0 qh-fuzzy filter, and may not be an ^!, !0 q h-fuzzy filter. This is shown in the followingexample.Example 4.20 (Davvaz & Khan 2012)Consider the ordered semigroup as given in Example 3.2 and define a fuzzy subset n of T asfollows: n^eh = 0.8, n^d h = 0.7, n^ah = 0.6, n^bh = 0.4, n^ch = 0.3, n^ f h = 0.5Then, T if 0 1 t # 0 \" a, d, e , if 0.5 1 t # 0.6 U^n;th = \"a,d,e,f , if 0.4 1 t # 0.5 \" a, d, e, f, b , if 0.3 1 t # 0.4 z if 0.8 1 tThus, n is a fuzzy filter with thresholds ^0.5, 0.6@ of T. But neither n is a fuzzy filter of T and^!,! 0 qh-fuzzy filter of T nor an ^!, !0 q h-fuzzy filter of T.^!, ! 0 qh-fuzzy left(right) filtersDefinition 5.1A fuzzy subset n of T is called an ^!,! 0 qh-fuzzy left (resp. fuzzy right) filter of T if it satisfies: (1) ^6x, y ! T h^6t ! ^0,1@h^x # y, xt ! n $ yt ! nh, (2) ^6x, y, z ! T h^6t, r, s ! ^0, 1@haxt ! n, yr ! n, zs ! n $ ^xyzhmin\"t,r,s, ! nk (3) ^6x, y, z ! T h^6t ! ^0,1@h^xt ! n $ ^xyzht ! n^resp. ^zyxht ! nhhExample 5.2 (Davvaz & Khan 2012)Let T = \"a,b,c,d,e,f , be a set. Define a ternary operation ^ h on T as ^abch = a ) b ) c, wherethe binary operation \" ) \" and order relation \" # \" are defined as follows: *abcde f abcdddd bcddddd cdddddd ddddddd eeeeeee fffffff 79

ASEAN Journal on Science and Technology for Development, 31(2), 2014and, #|= \"^a,ah, ^b,bh, ^c,ch, ^d,d h, ^e,eh, ^ f,f h, ^a,d h, ^b,d h, ^c,d h, ^d,eh, ^d,f h, ^a,eh, ^b,eh, ^c,eh, ^a,f h, ^b,f h, ^c,f h,, Then, ^T, $,#h is an ordered ternary semigroup. Left filter of T are \"e, f , and T. Define a fuzzysubset n of T as follows: n^ah = n^bh = n^ch = n^d h = 0.3, n^eh0.7 =, n^ f h = 0.5Then, U^n; th = ]]]]]Z[\]]]\"zTe,f , if 0 1 t # 0.3 if 0.3 1 t # 0.5 if 0.7 1 t # 1Then, n is an ^!, !0 q h-fuzzy filter of T.Theorem 5.3For a fuzzy subset μ of T the Conditions (1) – (3) respectively, of Definition 5.1, are equivalentto the following conditions: (i) ^6x,y ! T h^n^ yh $ min \"n^xh,0.5,h, (ii) ^6x,y,z ! T h^n^xyzh $ min \"n^xh,n^zh,0.5,h, (iii) ^6x,y,z ! T h^n^xyzh $ min \"n^xh,0.5,^resp. n^xyzh $ min \"n^zh,0.5,hh Proof. (1) & (i) Suppose that (1) is true. Let there exists a, b ! T , t0 ! ^0, 1@, sucht h a t n^bh 1 t0 # min \"n^ah, 0.5, , t h e n at0 ! n b u t bt0 g n , a c o n t r a d i c t i o n . H e n c en^ yh $ min \"n^xh, 0.5, for all x,y ! T .(i) & (1) Assume that (i) is true. Let a, b ! T , a # b, t ! ^0,1@ such that xt ! n, then n^xh $ t ,by hypothesis, we have n^ yh $ min \"n^xh, 0.5, $ min \"t, 0.5,. If t # 0.5, then n^ yh $ t impliesyt ! n. If t 2 0.5, then n^ yh $ 0.5 so that n^ yh + t 2 0.5 + 0.5 = 1 implies that yt q n. Henceyt !0q n.(2) & (ii) Suppose that (2) is true. If n^xh = n^ yh = n^zh = 0, then n^xyzh $ min \"n^xh, n^ yh,n^zh, 0.5, . Let n^xh ! 0 , n^ yh ! 0 , n^xh ! 0 and on contrary suppose that there existsa, b, c ! T , t1 ! ^0,1@ such that n^abch 1 t1 # min \"n^ah, n^bh, n^ch, 0.5, which implies thatat1 ! n, bt1 ! n, ct1 ! n but ^abcht1 g n, a contradiction. Hence n^xyzh $ min \"n^xh, n^ yh,n^zh, 0.5, for all z, y, z ! T . 80

M. J. Khan et al.: Generalized Fuzzy Filters in Ordered Ternary Semigroups(ii) & (2) Assume that (i) is true and let x, y, z ! T and t, r, s ! ^0, 1@, such that xt, yr, zs ! n.Then n^xh $ t , n^ yh $ r , n^zh $ s . By hypothesis, we have n^xyzh $ min \"t, r, s, 0.5, .I f min \"t, r, s, 2 0.5 , t h e n n^xyzh $ min \"t, r, s, i m p l i e s t h a t ^xyzhmin\"t,r,s, ! n . I fmin \"t, r, s, 2 0.5. Then n^xyzh $ 0.5, so that n^xyzh + min \"t, r, s, 2 0.5 + 0.5 = 1 implies^xyzhmin\"t,r,s,q n. Hence ^xyzhmin\"t,r,s, ! 0 q n.(3) & (iii) Suppose that (3) is true. If there exists a, b, c ! n and t2 ! ^0, 1@ such thatn^abch 1 t2 # min \"n^ah, 0.5, . T h e n at2 ! n b u t ^abcht2 g n a c o n t r a d i c t i o n . S on^xyzh $ min \"n^xh, 0.5, for all x, y, z ! n.(iii) & (3) Assume that (3) is true. Let x, y, z ! T and t ! ^0,1@ such that xt ! n. Then n^xh $ t .By hypothesis we have, n^xyzh $ min \"n^xh, 0.5, $ min \"t, 0.5,. If t # 0.5, then n^xyzh $ timplies that ^xyzht ! n. If t 2 0.5. Then n^xyzh $ 0.5. So that n^xyzh + t 2 0.5 + 0.5 = 1 implies^xyzht q n. Hence ^xyzht !0q n.Definition 5.4Let r,s ! ^0,1@ and r 1 s. Let m be a fuzzy subset of an ordered ternary semigroup T. Then n iscalled a fuzzy left (resp. right) filter with thresholds (r,s) of T if it satisfies the following conditions: (1) ^6x,y ! T h^x # y, max \"n^ yh,r, $ min \"n^xhs,h (2) ^6x,y ! T h^max \"n^xyzh,r,h $ min \"n^xh,n^ yh,n^zh,s, (3) ^6x,y ! T h^max \"n^xyzh,r,h $ min \"n^xh,s, ^resp. max \"n^xyzh,r, $ min \"n^zh,s,h. If n is a fuzzy filter with thresholds ^r, sh of T. Then m is an ^!,!0qh-fuzzy left (resp. right)filter of T if r = 0 and S = 0.5. Now we characterize fuzzy left (right) filters with thresholds ^r,shof T by their level left (right) filters.Theorem 5.5A fuzzy subset n of an ordered ternary semigroup T is a fuzzy left (resp. right) filter with thresholds^r, sh of T if and only if nt ^r, 0h is a left (right) filter of T for all t ! ^r,s@. Proof. The proof follows from Theorem 4.18. Date of submission: September 2014 Date of acceptance: December 2014 81

ASEAN Journal on Science and Technology for Development, 31(2), 2014 References Kazanci, O & Yamak, S 2008, ‘Generalized fuzzy bi-ideals of semigroup’, Soft Computing, vol. 12,Bhakat, SK & Das, P 1996, ‘ ^!, ! 0 qh -fuzzy pp. 1119–1124. subgroups’, Fuzzy Sets and Systems, vol. 80, pp. 359–368. Kazanci, O & Davvaz, B 2009, ‘Fuzzy n-ary polygroups related to fuzzy points’, ComputersBhakat, SK & Das, P 1996, ‘Fuzzy subrings and & Mathematics with Applications, vol. 58, pp. ideals redefined’, Fuzzy Sets and Systems, vol. 1466–1474. 81, pp. 383–393. Kehayopulu, N & Tsingelis, M 2002, ‘Fuzzy setsDavvaz, B 2006, ‘^!, ! 0 qh-fuzzy subnear-rings and in ordered groupoids’, Semigroup Forum, vol. ideals’, Soft Computing, vol. 10, pp. 206–211. 65, 128–132.Davvaz, B & Corsini, P 2008, ‘On ^a,bh-fuzzy Lehmer, DH 1932, ‘A ternary analoue of abelian Hy -ideals of Hy -rings’, Iran. J. Fuzzy Syst., vol. groups’, Amer. J. Math., vol. 59, pp. 329–338. 5, pp. 35–47. Los, J 1955, ‘On the extending of model’,Davvaz, B, Zhan, J & Shum, KP 2008, ‘Generalized Fundamenta Mathematicae, vol. 42, pp. 38–54. fuzzy polygroups endowed with interval valued membership functions’, Journal of Intelligent and Liu, L & Li, K 2005, ‘Fuzzy implicative and Boolean Fuzzy Systems, vol. 19, pp. 181–188. filters of R0-algebras’, Information Sciences, vol. 171, pp. 61–71.Davvaz, B 2008, ‘Fuzzy R-subgroups with thresholds of near-rings and implication Lee, SK & Park, KY 2003, ‘On right (left) duo po- operators’, Soft Computing, vol. 12, pp. 875–879. semigroups’, Kangweon-Kyungki Math. Jour., vol. 11, pp. 147–153.Davvaz, B, Kazanci, O & Yamak, S 2009, ‘Generalized fuzzy n-ary subpolygroups endowed Mordeson, JN, Malik, DS & Kuroki, N 2003, with interval valued membership functions’, ‘Fuzzy semigroups, studies in fuzziness and soft Journal Intelligent and Fuzzy Systems, vol. 20, computing’, vol. 131, Berlin, Springer-Verlag. pp. 159–168. Ma, X, Zhan, J & Jun, YB 2009, ‘On ^!, ! 0 qh-fuzzyDavvaz, B & Mozafar, Z 2009, ‘^!, ! 0 qh-fuzzy Lie filters of R0-algebras’, Math. Log. Quart., vol. 55, subalgebra and ideals’, International Journal of pp. 493–508. Fuzzy Systems’, vol. 11, no. 2, pp. 123–129. Pu, PM & Liu, YM 1980, ‘Fuzzy topology I,Davvaz, B & Khan, A 2012, ‘Generalized fuzzy neighborhood structure of a fuzzy point and filters in ordered semigroups’, Iranian Journal of Moore-Smith convergence’, J. Math. Anal. Appl., Science & Technology, vol. A1, pp. 77–86. vol. 76, pp. 571–599.Davvaz, B, Khan, A, Sarmin, N & Hidayatulla, K, Rosenfeld, A 1971, ‘Fuzzy groups’, J. Math. Anal. 2013, ‘More general forms of interval valued Appl., vol. 35, pp. 512–517. fuzzy filters of ordered semigroup’, Iranian Journal of Fuzzy System, vol. 15, no. 2, pp. Shabir, M & Khan, A 2008, ‘Characterizations of 110–126. ordered semigroups by the properties of their fuzzy generalized bi-ideals’, New MathematicsJun, YB & Song, SZ 2006, ‘Generalized fuzzy and Natural Computation, vol. 4, pp. 237–250. interior ideals in semigroups’, Information Sciences, vol. 176, pp. 3079–3093. Shabir, M & Khan, A 2008, ‘Fuzzy filters in ordered semigroups’, Lobachevskii J. Math., vol. 29, pp.Jun, YB, Khan, A & Shabir, M 2009, ‘Ordered 82–89. semigroups characterized by their ^!, ! 0 qh -fuzzy biideals’, Bull. Malays. Math. Sci. Soc., Chinram, R & Saelee, S 2010, ‘Fuzzy ideals and vol. 32, no. 2, pp. 391–408. fuzzy filters of ordered ternary semigroups’, Journal of Mathematics Research, vol. 2, pp.Kuroki, N 1981, ‘On fuzzy ideals and fuzzy biideals 93–97. in semigroups’, Fuzzy Sets and Systems, vol. 5, pp. 203–215. Yuan, X, Zhang, C & Ren, Y 2003, ‘Generalized fuzzy groups and many-valued implications’,Khan, A & Shabir, M 2009, ‘^a, bh-fuzzy interior Fuzzy Sets and Systems, vol. 138, pp. 205–211. ideals in ordered semigroups’, Lobachevskii Journal of Mathematics, vol. 30, pp. 30–39. Zadeh, LA 1965, ‘Fuzzy sets’, Inf. Cont., vol. 8, pp. 338–353. 82

ASEAN J. Sci. Technol. Dev.,  31(2): 83 – 89 Effects of Oil Palm (Elais guineensis) Fruit Extracts on Glucose Uptake Activity of Muscle, Adipose and Liver Cells S. Faez1*, H. Muhajir2, I. Amin3 AND A. Zainah4The effect of oil palm (Elaeis guineensis) fruit aqueous extract (OPF) on glucose uptake activity ofthree different cell lines was investigated. The cell lines were incubated with different concentrationsof OPF to evaluate the stimulatory effect of OPF towards glucose uptake activity of L6 myotubes,3T3F442A adipocytes and Chang liver cell line. The glucose uptake activities of all tested cellswere enhanced in the presence of OPF extract (basal condition). Nevertheless in combination ofOPF extract and 100 nM insulin, the glucose uptake activity was only significantly enhanced in L6myotubes and 3T3F442A adipocytes cell lines. The extracts enhanced the glucose uptake into cellsthrough either insulin-mimetic or insulin-sensitizing property or combination of these two properties.It can be suggested that the OPF extract exerts its antihyperglycemic action partly by mediatedglucose uptake into the glucose-responsive disposal cells, muscle, adipose and liver.Hyperglycaemia in diabetic patients is reduced weight gain (Arner 2003; Zangeneh et al.in a number of ways. One known mechanism 2003). Thus these limitations have fuelled theto reduce hyperglycaemia is by glucose uptake search for alternative therapeutic agents for theinto peripheral cells (muscle, adipocytes treatment of diabetes mellitus.and liver) (Patel & Mishra 2008). Glucose istaken into the cells through glucose transpoter Several plants have been identified to(GLUTS). In the liver glucose, it is taken up improve glucose uptake into peripheral cells.through GLUT2 while GLUT4 play its role to The antihyperglycemic properties of plants weretake up glucose in adipocyte and muscle cells reported to be associated with the polyphenolic(Troy & Glenn 2009). Insulin sensitizers like and flavonoid content which can be foundthiazolidinediones and metformin have been in various plants. Plants like Amomi semenreported to improve insulin-mediated glucose (Kang & Kim 2004) and Cortex Phellodendriuptake into peripheral cells (Zangeneh et al. (Ko et al. 2005) have been reported to enhance2003). They are oral antidiabetic drugs that insulin-mediated glucose uptake activity intousually prescribed to diabetic patients. However adipocytes cells. On the other hand Pterocarpusthe drugs exert some adverse effect where those marsupium had been reported to enhancedrugs were reported to cause gastrointestinal glucose uptake into L6 myotubes (Anandharajanproblems, peripheral edema and can cause et al. 2006). Those plants are believed to1 Department of Biotechnology, Kulliyah of Sciences, International Islamic University Malaysia, 25200 Indera Mahkota, Kuantan, Pahang, Malaysia2 Department of Microbiology, Faculty of Biotechnology and Biomolecular Sciences, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia3 Department of Nutrition and Dietetics, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, 43400 Serdang, Selangor Malaysia4 Medical Technology Division, Malaysian Nuclear Agency, Bangi 43000 Kajang, Selangor, Malaysia* Corresponding author (e-mail: [email protected])

ASEAN Journal on Science and Technology for Development, 31(2), 2014enhance glucose uptake into peripheral cells (Merck, Germany) was added into the mixturethrough either mimicking or sensitizing insulin prior to reflux. After refluxing for 2 h, theaction (Jung et al. 2007). Plants with high extract was cooled, filtered, and standardized tophenolic antioxidant compounds exert high 50 ml with 60% ethanol. Then the solvent waspotential as supplements for improving blood removed using a rotary evaporator. Finally theglucose control and preventing long-term OPF was preserved by freeze drying.complications among diabetics (Gallegheret al. 2003). Cell Line Maintenance The oil palm fruit is planted abundantly In the evaluation of glucose uptake activity,throughout Malaysia. The fruit has been three cell lines were used (L6 myotubes,extensively researched for its health and 3T3F442A adipocytes and Chang liver cells)nutritional properties, including antioxidant as the model of the glucose uptake system.activities, cholesterol lowering, anticancer L6 myotubes and 3T3F442A adipocytes wereeffects and protection against cardiovascular maintained in Dulbecco’s Modified Eaglediseases (Sundram et al. 2003). The fruits are Medium (DMEM) while Chang liver cells werealso reported to contain a significant degree of maintained in medium Roswell Park Memorialantioxidant polyphenolic compounds (Sundram Institute (RPMI). The complete culture mediumet al. 2003; Neo et al. 2010). Nevertheless was done by addition of 10% (v/v) fetal bovinethe antidiabetic effects of oil palm fruits on serum (FBS) and 1% (v/v) antibiotic solution.stimulating glucose uptake activity are yet The cell lines were incubated and humidifiedto be known. Therefore in the present study with 5% CO2 at 37°C condition. Followingthe ability of oil palm fruit extracts (OPF) sub-confluences (70%–80%) cultures wereto stimulate glucose uptake into peripheral split using 0.25% Trypsin to 1:3. After that(muscle, adipose and liver) cells were evaluated. the cells were centrifuged for 5 min at 10The L6 myotubes, 3T3F442A adipocytes and 000 r.p.m. and the pellets were suspendedChang liver cell line was used as the model of again into culture medium until reachingthe system. confluency. The L6 myoblasts were induced to differentiate into myotubes by reducing the Materials and methods FBS in the complete culture medium from 10% to 2% (Ziyou et al. 2009). The cells wereSampling and Sample Preparation maintained with this medium for 4–6 days post-confluence. Along the period the cells wereThe oil palm fruit, E. guineensis, was collected observed and the extent of differentiation wasfrom the Universiti Putra Malaysia Agriculture established by observing multinucleation ofPark. Ripe oil palm fruits were harvested from cells. The 3T3F442A fibroblast-like cells weretrees and used as the fresh sample. The fresh spontaneously differentiated into adipocytesoil palm fruits were scraped into thin flakes upon reaching confluency. The presence ofand dried in oven overnight at 40°C. The dried viscous media in each well confirmed thefruits then were ground into small particles extent of differentiation as free fatty acidsand the oil was removed with hexane (Merck, were produced by the cells and secreted intoGermany) by using the Soxhlet method (45°C, the media.8 h). Following removal of oil, the OPF wasextracted according to the method described Evaluation of Glucose Uptake Activity ofby Wang and Halliwell (2001). Briefly 1g of OPF Extractdried de-oiled oil palm mesocarp was mixedwith 40 ml of 60% aqueous ethanol (Merck, The cells were seeded into 12-well plate at theGermany). Subsequently, 5 ml of 6 M HCl concentration of 2 × 105 cells per well. The cells 84

S. Faez et al.: Effects of Oil Palm (Elais guineensis) Fruit Extracts on Glucose Uptake Activitythen were left overnight to allow attachment. the present study, three cell lines were usedAfter the cells were attached on the following to study the effects of OPF extract on glucoseday, they were washed with serum-free medium uptake activity. The cell lines used include L6thrice. Then the cells were incubated with myotubes, 3T3F442A adipocytes and Changthe same medium for two hours. After the liver cells. A previous report has found that L6starvation period, the cells were then washed myotube was the best-characterized cellularwith Kreb’s-Ringer bicarbonate buffer (KRB) model of skeletal muscle to study the glucosethrice. Then the cells were pre-incubated for 30 uptake activity by GLUT4 translocation (Patelmin with various concentration of OPF extract. & Mishra 2008). The 3T3F442A adipocyte onMetformin and rosiglitazone maleate were used the other hand has been widely used to studyas positive control. After 30 min of incubation, the effects of an insulinotropic agent on glucose500 µl of 2-deoxy-[3H]-glucose 1 µCi/ml uptake activity into adipocytes (Sakurai et al.,diluted 0.1 mM glucose was added to each well 2004). Meanwhile the Chang liver cell is oneexcept the blanks to initiate the glucose uptake of the models used widely to study the glucosereaction. The reaction was allowed to occur for transports in the liver besides HepG2 and H4IIE60 min. Subsequently the cells were washed (Rengarajan et al. 2007). All models of celltrice with ice-cold KRB buffer. Then 0.1% lines used in this study showed the ability tosodium dodecyl sulphate dissolved in 0.1M enhance glucose uptake activity when treatedphosphate buffer pH 7.4 was added to solubilize with OPF extract at particular concentrationsthe cells. Then the mixture of each well was (Figures 1–3). This observation indicated thattransferred into the scintillated cocktail and 15 there was a possibility that the antidiabeticml of scintillated cocktail, Ultima GoldTM was compounds were present in the OPF extract. Theadded. Finally the radioactivities incorporated compounds might have the potential to regulateinto the L6 myotubes which indicated the hyperglycaemia through the enhancementglucose uptake activity were measured using of glucose disposal into muscle, adiposeLiquid Scintillation Counter (Hewlett Packard, and liver cells. However purification of theUSA). compounds was not conducted in the present study. Therefore further experimentationsStatistical Analysis are needed to be carried out to evaluate the exact compounds which may be responsible toData were expressed as mean ± standard regulate the mechanism by which this disposaldeviation. One-way ANOVA (GraphPad Prism is mediated.5) were used for analysis and groups wereconsidered significantly different at the 5% Plant extracts may potentiate the glucosesignificance level (p<0.05). Dunnet post-hoc uptake into cells through insulin-mimetic ortest was done if a significant value was obtained insulin sensitizing properties. These propertiesfor ANOVA. of plants were reported to be associated with the flavonoids content which can be found Results and Discussions in various plants. For example, a flavonoids found in grapefruit namely naringenin, hasCell lines are homogenous culture of cells. been shown to increase basal glucose uptake inDue to their homogeneity, the cell lines L6 myotubes which is comparable to 100 nMbecome the best choice to study the effects insulin (Zygmunt et al. 2010). On the otherof an agent on insulin activity compared to hand, the grape seed contains procyanidinsisolated cells or tissues which are mixed in which is reported to have insulin-mimetic effectnature. Furthermore, cell lines are more stable during stimulating glucose uptake into 3T3L1and have longer lifespan compared to the adipocytes cells (Pinent et al. 2004). Glucose isisolated cells (Parthasarathy et al. 2009). In 85

ASEAN Journal on Science and Technology for Development, 31(2), 2014taken into the cells through glucose transpoter and Chang liver cells. Furthermore according(GLUTS). In the liver glucose it is taken up to the previous report, the 100 nM of insulinthrough GLUT2 while GLUT4 plays its role to concentration has also been widely used totake up glucose in adipocyte and muscle cells mediate the glucose disposals into cells (Sakurai(Troy & Glenn 2009). The GLUT4 is located et al. 2004). Nevertheless in the treatment usinginside the cells and its translocation to cell combination of OPF and 100 nM insulin, onlymembrane to facilitate glucose transport into the the L6 myotubes and 3T3F4424A adipocytescells are sensitize with the presence of insulin. In showed a significant enhancement in the glucosecontrast GLUT2 is located in the cell membrane uptake activity. The glucose uptake activity ofand facilitate glucose transport into cells Chang liver cell was not significantly enhancedwithout the need of insulin. GLUT2 can sensor in the presence of both OPF and 100 nM insulinthe presence of glucose independently since (Figure 3). The Chang liver cell mediated thethey have high capacity and low affinity (high glucose uptake through GLUT2 where insulinKm value, 15 mM–20 mM) for glucose. The high was not the core factors in their mechanism toKm value allows for glucose sensing where the mediate glucose uptake compared to GLUT4rate of glucose uptake is proportional to blood which required insulin to promote glucoseglucose level (Li et al. 2007). Nevertheless transportation into the muscle and adipose cells.according to the previous report the presence Therefore the difference of functional GLUTof insulin can also enhance the glucose uptake protein involved in the translocation of glucoseactivity in Chang liver cells (Satake et al. 2002). into the liver cells might explain the resultsOur present data are in accordance with the obtained in the previous study (Li et al. 2007).previous reports where insulin 100 nM alonesignificantly enhanced glucose uptake activity The mechanisms underlie the insulin-in all types of cells evaluated (Figures 1–3). mimetic and insulin-sensitizing property ofThe same concentration of insulin was used the OPF extract was not elucidated in theto mediate the glucose uptake activity of OPF presence study. However the insulin-like orextract in L6 myotubes, 3T3F442A adipocytes insulin-mimetic activity of a plant has beenPercentage of 2-deoxy [1-3H] glucose 500 uptake relative to control Control (dpm/105 cells/60 min) Insulin 100 nM 400 50 μg/ml 100 μg/ml 300 500 μg/ml 1000 μg/ml 200 100 0 With insulin 100 nM BasalFigure 1. Effect of OPF extract on basal and insulin-mediated glucose uptake activity of L6 myotubes. Values represent the means ±S.D. *p<0.05, **p<0.01 and ***p<0.001 compared with control. ●●p<0.01 compared to 100 nM insulin alone. 86

S. Faez et al.: Effects of Oil Palm (Elais guineensis) Fruit Extracts on Glucose Uptake Activityevaluated previously in several plants. An This activation subsequently enhanced tyrosineexample is the fanugreek seed. The fenugreek phosphorylation of insulin receptor substrate-1seed extract has been reported to mediate (IRS-1) and p85 subunit of phosphatidylinositol-the glucose uptake activity into liver and 3-kinase (PI3-kinase) which leads to glucoseadipocytes through the activation of tyrosine uptake by these cells (Vijayakumar et al.phosphorylation of ß-subunit of insulin receptor. 2005). The insulin sensitizing activity hasPercentage of 2-deoxy [1-3H] glucose 250 uptake relative to control Control (dpm/105 cells/60 min) Insulin 100 nM 200 50 μg/ml 100 μg/ml 150 500 μg/ml 1000 μg/ml 100 50 0 Basal With insulin 100 nMFigure 2. Effect of OPF extract on basal and insulin-mediated glucose uptake activity of 3T3F442A adipocytes. Values represent the means ±S.D. **p<0.01 and ***p<0.001 compared with control.Percentage of 2-deoxy [1-3H] glucose 250 uptake relative to control Control (dpm/105 cells/60 min) Insulin 100 nM 200 50 μg/ml 100 μg/ml 150 500 μg/ml 1000 μg/ml 100 50 0 Basal With insulin 100 nMFigure 3. Effect of OPF extract on basal and insulin-mediated glucose uptake activity of Chang livercells. Values represent the means ±S.D. *p<0.05, **p<0.01 and ***p<0.001 compared with control. ●p<0.05 and ●●p<0.01 compared to 100 nM insulin alone. 87

ASEAN Journal on Science and Technology for Development, 31(2), 2014been previously reported in several plants like mouse embryo is essential for neural tubeCampsis grandiflora (Jung et al. 2007) and defects associated with diabetic embryopathy’,miltiorrihiza Bunge. They were reported to Diabetologia, vol. 503, pp. 682−689.enhance the tyrosine phosphorylation of insulinreceptor and activation of downstream kinase. Neo, YP, Aziz, A, Tan, C P & Tan, YA 2010, ‘PhenolicTherefore there are possibilities that OPF extract acid analysis and antioxidant activity assessmentexert the insulin-mimetic and insulin-sensitizing of oil palm E. Guineensis fruit extracts’, Foodproperties through the mentioned mechanisms. Chemistry, vol. 122, pp. 353−359.However further evaluation are needed to becarried out to confirm this suggestion. Parthasarathy, C, Renuka, VN & Balasubramaniam, K 2009, ‘Sex steroids enhance insulin receptors Date of submission: September 2014 and glucose oxidation in Chang liver cells’, Clinica Chimica Acta, vol. 399, vols. 1−2, pp. Date of acceptance: November 2014 49−53. REFERENCES Patel, MB & Mishra, SH 2008, ‘Cell line in diabetic research: a review’, Pharmacognosy Review, vol.Anandharajan, R, Pathmanathana, K, 2, no. 4, pp. 188−205. Shankernarayanan, NP, Vishwakarmab, RA & Balakrishna, A 2005, ‘Upregulation of GLUT4 Pinent, M, Blay, M, Blade, MC, Salvado, MJ, and PPARγ by an isoflavone from Pterocarpus Arola, L & Ardevol, A 2004, ‘Grape seed derived marsupium on L6myotubes:Apossible mechanism procyanidins have an antihyperglycemic effect of action’, Journal of Ethnopharmacology, vol. in streptozotocin-induced diabetic rats and 97, pp. 253−260. insulinomimetic activity in insulin sensitive cell lines’, Endocrinology, vol. 145, no. 11, pp.Arner, P 2003, ‘The adipocyte in insulin 4985−4990. resistance: key molecules and the impact of the Thiazolidinediones’, Trends in Endocrinology and Rengarajan, S, Parthasarathy, C, Anitha, M & Metabolism, vol. 14, no. 3, pp. 137−144. Balasubramaniam, K 2007, ‘Diethyl phthalate impairs insulin binding and glucose oxidation inGallagher, AM, Flatt, PR, Duffy, G & Abdel-Wahab, Chang liver cells’, Toxicology in Vitro, vol. 21, YHA 2003, ‘The effects of tradisional antidiabetic pp. 99−102. plants in vitro glucose diffusion’, Nutrition Research, vol. 23, pp. 413−424. Sakurai, K, Kawazuma, M, Adachi, T, Harigaya, T, Saito, Y, Hashimoto, N & Mori, C 2004,Jung, SH, Ha, YJ, Shim, E K, Choi, SY, Jin, JL, ‘Bisphenol A affects glucose transport in mouse Yun-Choi, HS & Lee, JR 2007, ‘ Insulin mimetic 3T3F442A adipocytes’, British Journal of and insulin sensitizing activities of tanshinones, Pharmacology, vol. 141, no. 2, pp. 209−214. diterpine compounds of the root of Salvia miltiorrhiza Bunge, Phytomedicine, vol. 16, pp. Satake, S, Moore, MC, Igawa, K, Converse, M, 327−335. Farmer, B, Neal, DW & Cherrington, AD 2002, ‘Direct and indirect effects of insulin on glucoseKang, Y & Kim, YH 2004, ‘Glucose uptake- uptake and storage by the liver’, Diabetes, vol. stimulatory activity of Amomi semen in 3T3-L1 51, no. 6, pp. 1663−1671. adipocytes’, Journal of Ethnopharmacology, vol. 92, pp. 103−105. Sundram, K, Sambarthamurti, R & Tan, YA 2003, ‘Palm fruit chemistry and nutrition’, Asia PacificKo, BS, Choi, SB, Park, SK, Jang, JS, Kim, J. Clin. Nutr., vol. 12, no. 3, pp. 355−362. YE & Park, S 2005, ‘Insulin-sensitizing and insulinotropic action of Berberin from Cortidis Troy, LM & Glenn, KM 2009, ‘Skeletal muscle Rizhoma, Biological & Pharmaceutical Bulletin, glucose uptake during excercise: a focus vol. 28, no. 8, pp. 1431−1437. on reactive oxygen species and nitric oxide signalling’, Life, vol. 615, pp. 479-484.Li, R, Thorens, B & Loeken, MR 2007, ‘ Expression on the gene encoding the high Km glucose Vijayakumar, MV, Singh, S, Chhipa, RR & Bhat, MK transporter 2 by the early post-implantation 2005, ‘The hypoglycaemic activity of fenugreek seed extract is mediated through the stimulation of an insulin signalling pathway’, British Journal of Pharmacology, vol. 146, pp. 41−48. 88

S. Faez et al.: Effects of Oil Palm (Elais guineensis) Fruit Extracts on Glucose Uptake ActivityWang, H & Helliwell, K 2001 Determination of H, John, RY & Fuquan, Y 2009 ‘Preliminary flavonols in green and black tea leaves and quantitative profile of differential expression green tea infusion by high performance liquid between rat L6 myoblast and myotubes by stable chromatography’, Food Research International, isotope labelling by amino acids in cell culture’, vol. 34, pp. 223–227. Proteomics, vol. 95, pp. 1274−1292.Zangeneh, F, Kudva, YC & Basu, A 2003, ‘Insulin Zygmunt, K, Faubert, B, MacNeil, J & Tsiani, E 2010, sensitizers’, Mayo Clinic Proceedings, vol. 784, ‘Naringenin, a citrus flavonoid, increases muscle pp. 471−479. cell glucose uptake via AMPK, ‘Biochemical and Biophysical Research Communications, vol. 398,Ziyou, C, Xiulan, C, Bingwen, L, Sung, KP, Tao, no. 2, pp. 178−183. X, Zhensheng, X, Peng, X, Junjie, H, Haiying, 89

ASEAN J. Sci. Technol. Dev.,  31(2): 90 – 100Biodiesel Production from Castor Oil and Its Application in Diesel Engine S. Ismail1, S. A. Abu*, R. Rezaur3 and H. Sinin4In this study, the optimum biodiesel conversion from crude castor oil to castor biodiesel (CB)through transesterification method was investigated. The base catalyzed transesterification underdifferent reactant proportion such as the molar ratio of alcohol to oil and mass ratio of catalyst to oilwas studied for optimum production of castor biodiesel. The optimum condition for base catalyzedtransesterification of castor oil was determined to be 1:4.5 of oil to methanol ratio and 0.005:1of potassium hydroxide to oil ratio. The fuel properties of the produced CB such as the calorificvalue, flash point and density were analyzed and compared to conventional diesel. Diesel engineperformance and emission test on different CB blends proved that CB was suitable to be used asdiesel blends. CB was also proved to have lower emission compared to conventional diesel.Key words: biodiesel, transesterification, castor oil, diesel engine, emission; conversionPetroleum fuels play a very important role Biodiesel has become an interesting alternativein the development of various industries, fuel over conventional diesel for decades.transportations, agriculture sector and to meet Biodiesel is suitable to be used in diesel enginemany other basic human needs in modern due to the similar properties to conventionalcivilization. These fuels are limited and diesel in terms of power and torque and nonedepleting day by day as the consumption or very minor engine modification is requiredincrease very rapidly. Moreover, the use of (Mushtaq et al. 2011). Moreover, biodieselpetroleum fuel has caused a lot of environmental is biodegradable which will results in lessproblems by the high emission of harmful environmental impact upon accidental releasegases. A global movement towards generation to the environment (Janaun & Ellis 2010).of environmentally friendly yet renewablefuel is therefore under way to help meet the Biodiesel has many important technicalincreased energy demands. Biofuel had become advantages over conventional diesel such asone of the most promising alternatives for inherent lubricity, low toxicity, derivation frompetroleum fuels. a renewable and domestic feedstock, superior flash point, negligible sulphur content and lower Biodiesel is the potential biofuel that can exhaust emissions (Moser 2009). Biodiesel hadeasily being produced from vegetable oil. been used widely as a blend with diesel. The1 Department of Chemical Engineering, Faculty of Engineering, Universiti Malaysia Sarawak, 94300 Kota Samarahan, Sarawak, Malaysia2 Department of Mechanical Engineering, Faculty of Engineering, Universiti Malaysia Sarawak, 94300 Kota Samarahan, Sarawak, Malaysia.3 Department of Chemical Engineering, Faculty of Engineering, Universiti Malaysia Sarawak, 94300 Kota Samarahan, Sarawak, Malaysia.4 Pusat Pengajian Pra-Universiti, Universiti Malaysia Sarawak, 94300 Kota Samarahan,Sarawak. Malaysia* Corresponding author (e-mail: [email protected])

S. Ismail et al.: Biodiesel Production from Castor Oil and Its Application in Diesel Engineuse of biodiesel as diesel blends will promote to determine the optimum reaction conditioncleaner emission with less soot particles for the production of castor biodiesel. Then, theand whiter smoke. It also helps in reducing fuel properties such as density, flash point andengine wear by lubrication and produces calorific value was analyzed and compared toless sulphur emission. A biodiesel lifecycle conventional diesel. Engine performance andstudy in 1998 which was jointly sponsored by emission of castor biodiesel was also testedthe U.S. Department of Energy and the U.S. using various biodiesel blends and comparedDepartment of Agriculture concluded biodiesel to the conventional diesel.reduces carbon dioxide emissions by 78 percentcompared to petroleum diesel. The CO2 released Methodology and materialsinto the atmosphere when biodiesel burned isrecycled by plants, which produce more oxygen In this study, crude castor oil was extracted(Petracek 2014). from castor bean by using mechanical and solvent extraction. The castor beans used was Among the common vegetable oils used obtained from a local company. The acid valueas feedstock for the production of biodiesel are of the crude castor oil was determined bysoybean, rapeseed, castor, jatropha and palm oil. titrimetry. The castor oil was converted intoCastor oil is one of the promising feedstock for biodiesel by using two-step transesterificationbiodiesel production. Castor oil is produced by processes. In this process, the first step is acid-means of extraction from castor bean. Castor catalyzed esterification used to convert freeoil is distinguished by its high content (over fatty acids (FFA) in castor oil to methyl ester,85%) of ricinoleic acid. No other vegetable followed by base-catalyzed transesterificationoil contains so high a proportion of fatty using potassium hydroxide as a catalyst withhydroxyacids. Castor oils have high molecular methanol.weight (298), low melting point (5˚C) and verylow solidification point (–12˚C to –18˚C) that In the second step, potassium hydroxidemake it industrially useful, most of all it is has was dissolved in methanol and the mixture isthe highest and most stable viscosity of any then heated up to 60°C to react accordingly tovegetable oil (Shrirame et al. 2011). form methoxide. On the other hand, the pre- treated oil in step 1 was then heated up to 60°C. The chemical structure of castor oil is of The heated oil was mixed with the methoxidegreat interest because of the wide range of and the solution was shaken at 250 r.p.m. forreactions it affords to the oleochemical industry 2 h by using orbital shaker. The volume ratio ofand the unique chemicals that can be derived methanol to oil used was 1:4.0, 1:4.5 and 1:5.0from it. These derivatives are considerably while the volume ratio of potassium hydroxidesuperior to petrochemical products since they catalyst to oil used was 0.0025:1, 0.0050:1 andare from renewable sources, bio-degradable 0.0075:1. The volume ratio of alcohol to oiland eco-friendly (Nielsen et al. 2011). Recent was kept constant when the catalyst amountsresearch had concerns in the using of castor were being manipulated. The volume ratiooil as a feedstock for biodiesel production. As of catalyst to oil was kept constant when thecastor oil is non-edible, there is no issue of amount of alcohol was being manipulated.competition with the food market and it can be After completing the process, the mixture wasthe promising source of feedstock for biodiesel allowed to settle for 8 h and then the mixtureproduction. was poured into separatory funnels.In this study, the acid-based catalyzed The lower layer of glycerol, extra methanol,transesterification of castor oil was carried out catalyst and other byproducts were removed. 91

ASEAN Journal on Science and Technology for Development, 31(2), 2014The upper layer of methyl ester or biodiesel was transesterification for biodiesel production.washed several times with de-sterilized water From the titration method, the acid valueuntil the washing water become neutral. The of FFA in crude castor oil was determinedbiodiesel layer was filtered to remove impurities to be between 20% to 23%, which is higherand then the biodiesel was heated up to 100°C to than 4%. The best conversion method for oilremove any remaining water. The biodiesel was with free fatty acid higher than 4% was two-the tightly sealed and kept for storage. steps transesterification where FFA value is reduced at the first step (acid esterification) Biodiesel testing was carried out to before proceeding to the second step (basecompare the properties and performances transesterification) .of castor biodiesel and conventional diesel.The density, flash point and calorific value Optimization of Biodiesel Productionare measured respectively using density by Manipulation of Catalyst andmeter, multi-flash flash point tester and bomb Alcohol Amountcalorimeter. Emission analyses were carried outusing Flue Gas Analyzer. Castor biodiesel and For the first set of experiment, the amount ofconventional diesel were tested using FT-IR catalyst was set as the manipulated variableShimadzu Iraffinity-1 Spectrophotometer for while the amount of methanol was set ascomponent analysis. the constant variable. From Table 1, it is observed that the highest yield of biodiesel was Diesel engine test was performed using achieved with potassium to oil ratio of 0.0050:1.Techno-mate, TNM-TDE-700 machine. The However, the biodiesel yield before and afterdiesel engine testing was done three times the optimal amount of catalyst was noted towith each blend of biodiesel. The blending be lower. In the case of the catalyst shortagepercentage of biodiesel with diesel was set (0.0025:1 ratio), the biodiesel yield percentageto 0%, 10%, 20%, 30%, 40%, 50% and they was 60% as catalyst was exhausted before allare mentioned as B10, B20, B30, B40, B50. the crude oil was converted to biodiesel while inImportant values such as motor speed, output the case of excess catalyst (0.0075:1 ratio), thevoltage, output current and time for 20 ml fuel yield percentage was at 55% as excess catalystflow were recorded. The brake load for the attributed to soap formation which decreaseddiesel engine testing was fixed at 120 N and the the production of biodiesel.radius of brake arm was set to 0.5 m . For the second set of experiment, the RESULTS AND DISCUSSIONS amount of methanol was set as the manipulated variable while the amount of catalyst was setMeasurement of Free Fatty Acid (FFA) in as the constant variable. From Table 2, it isCrude Castor Oil observed that the highest yield of biodiesel was achieved with the 1:4.5 of oil-to-methanolMeasurement of FFA in crude castor oil is ratio. The biodiesel yield was also affectedessential for the decision of the method of by the amount of methanol used. The shortageTable 1. Biodiesel Yield Percentage for Different Amount of KOH (catalyst).KOH to oil ratio (v/v) Castor biodiesel produced (ml) Biodiesel yield (%) 0.0025:1 12 60 0.005:1 13 65 0.0075:1 11 55 92