Compositional Analysis on Ancient Bricks from Candi Sungai Mas ( Site 32 / 34 )

Problem statement: The significance of Kampung Sungai Mas as an impor tance archaeology site in Lembah Bujang was revealed in t he year 1979 when a cannel was built along Kampung Sungai Mas area. During the construction st udy of this cannel, thousands of ceramic pieces from India, Middle East, China, Southeast Asia and Europe was discovered, pieces of glasses from Middle East and Europe, stone and glasses beads fro m Middle East, India and local also pieces and blocks of construction substance from brick and mud stone or slate stone. An inscription also was discovered by one of the villagers when the cannel construction was in progress and this inscription sized estimated 46 ×23 cm produced from mud stone and relative estimati on based on Pallawa script on the stone and can be dated around 56th century AD. Scientific study done onto Candi Sungai Mas (Site32/34) ancient brick was prominent for archaeo logy data because since the very beginning, archeologist who was doing research in Kampung Sung ai Mas questioned whether the brick discovered was produced by the local society or bro ught from other area. Approach: Technique used was X-Ray Fluorescence Spectroscopy (XRF) and X-Ray Diffraction (XRD) technique. X-ray Fluorescence Spectroscopy (XRF) can determine the m ajor and trace elements of the bricks whilst XRay Diffraction (XRD) can determine the mineral con tent of the bricks. Results: Compositional analysis on Candi Sungai Mas (Site 32/34) showed that open firing technique was used to burnt the bricks and that there were two different raw source s used to produce the brick at Kampung Sungai Mas. The raw material used for making the bricks we re taken from Bujang River and Muda River basin. Conclusion: It can be concluded here that the protohistory soc iety at Kampung Sungai Mas especially Malay society has succeeded to conquer k nowledge and technology on making the brick and showed that we have developed long ago.


INTRODUCTION
The discovery of Sungai Mas Incription in Kampung Sungai Mas (Rahman and Yatim, 1990) and revealed thousands of artifacts from the cannel construction project has attracted many scholars and historians. The first excavation attempt in Kampung Kampung Sungai Mas (Site 32/34) is a protohistory site that keeps the country various priceless treasures. Sungai Mas was founded earlier from the more north areas that is Pengkalan Bujang. The claim was forwarded by Nik Hassan Shuhaimi Nik Abdul Rahman who has divided it into two phase those are Buddha phase which started from the 5th century till 9th AD which focused around Sungai Mas and also southern of Sungai Muda. The second phase from the 9th century to the 14th AD which focused to Pengkalan Bujang which later expanded into trading centre (Shuhaimi, 1984). By the 10th century, the Siwa teaching cult has expanded again around Pengkalan Bujang especially at the middle part and upper part of Sungai Bujang and along Merbok Kechil River (Shuhaimi, 1984).
Sungai Mas area was overlooked by the previous west researchers who had studied a lot the areas at Sungai Bujang, Sungai Batu, Sungai Merbok Kechil, Bukit Batu Pahat, Bukit Meriam, Bukit Penjara, Tikam Batu and area at the south those are Kota Aur and Seberang Prai. The findings of Buddha inscriptions and Buddha head which was chiseled onto greenstone which shows that Sungai Mas (Site 32) has developed since 5th or 6th century (Rahman and Zakaria, 1993) till 13th or 14th century and continues till the era of Kedah Malay Sultanate where this area is still active as trading harbor.
Scientific study onto ancient artifacts found in Bujang Valley have been carried out before this (Ramli et al., 2009;2011a). Many scientific studies were carried out onto clay pot artifacts where morphology and composition study was carried out to trace source used that is the local clay or not (Chia, 1998;Ramli et al., 2007;2011b). The clay morphology and composition study at Lembah Bujang also has been carried out and used as referential data for artifact and clay composition comparison in Lembah Bujang. Scientific study is hoped to prove our hypothesis that the brick used as the construction materials at Sungai Mas was produced locally. This can be done by data manipulation with the clay analysis data around the study area.

MATERIALS AND METHODS
Fractions of twenty six brick samples, taken from different excavation plot was recorded. Samples are taken to lab and treatment onto the samples was carried out. For the characterization of the shards and clay samples, analytical instruments used included X-Ray Diffraction SIEMENS D5000 Diffractometer and XRF Spectrometer Philips Model PW1480. The important of XRD and XRF techniques are shown in various field of research (Ridha et al., 2009;Woon et al., 2009;Abdullah et al., 2011;Hamzah et al., 2011). Samples for XRF analysis then be grinded into very fine powder form. A mixture of 0.4 g of sample in powder form (heated at 105°C) and 4.0 g of flux powder (Johnson Matthey Spectroflux 110) homogeneously been mixed together. The mixture was fused in an electric furnace at 1100°C, being set for one hour to make a glass.
Homogeneous molten sample than was casted into container and let to be cooled in stagesto become fuse glass with diameter of 32and 2mm thickness. Fused glass samples are prepared for major elements analysis such asSi, Na, K, Ca, Fe, Al, Ti, Mn, Mg and P 2 O 3 . Press pallet samples then be prepared for trace elements analysis such as As, Ba, La, Co, Cr, Cu, Ga, La, Nb, Ni, Pb, Rb, Hf, V, Zn and Zr. Press pallet samples was prepared by mixing 1.0 g of sample and 6.0 g of powder boric acid in a sample container and the be pressed to 20psi pressure by hydrolyte pressure instrument. Previous study on the fuss glass and press pallets samples then involved XRF Spectrometer Philips Model PW1480.

RESULTS AND DISCUSSION
Mineral content analysis was carried out onto 26 brick samples and the analysis results show there is minerals such as quartz, muscovite, microcline, kaolinite, illite, montmorillonite and minor mineral such as chloritiod and tremolite (Table 1). Quartz mineral is seldom found separated from other mineral. Quartz mineral is found in all research samples and the test result also shows that d14-1 sample and s12-2 sample only observed the reading of quartz mineral. This does not mean these samples do not contain other minerals but mineral from mica and feldspar groups has unwind due to high burning. Both samples have been burnt at a higher temperature from other brick samples. The estimated temperature achieved for both samples is as high as 800°C. The brick samples which contain quartz mineral and muscovite are such as d14-1, i7-2, i7-2, s6-1, s6-2, s6-3, s20-1, s20-2, t18-2, u12-1, u14-1 and u15-1. Brick samples containing mineral such as quartz, muscovite and orthoclase are samples d14-2, i10, s12-1, t18-3 and u15-2. Muscovite mineral is mineral from mica group whereas orthoclase mineral is mineral from feldspar group. Other brick samples containing additional mineral such as sample s12-3 contain quartz mineral, muscovite and chloritiod, sample t16-1 contain minerals such as quartz, muscovite, orthoclase and kaolinite, sample t16-2 contains minerals such as quartz, muscovite and illinite, sample t18-1 contains minerals such as quartz, illite and orthoclase and sample u12-2 contains mineral such as quartz, muscovite, tremolite and montmorillonite.
What is interesting to discuss about these brick samples is due to the existence of minerals such as kaolinite, illite and montmorillonite. The presence of these mineral showed that these samples are burnt below 550°C. The indifferently or uneven brick burning temperature showed that the bricks has been using open burning technique. Bricks created and dried are then arranged and gathered in an area and then burnt using woodpile placed on top of the bricks. If the bricks arrangement is a lot, it is not impossible brick with the lowest and farthest position from the fire would experience low burning rate as compared to the ones outside or nearer to the fire. This is different when the bricks are burnt in the furnace where the high temperature in the furnace and the way bricks are arranged are more orderly would ensure the bricks to achieve a standard temperature. Analysis of major elements done onto the brick samples at Candi Sungai Mas (Site 32/34) are shown in Table 2. Dry weight percentage of silica in between 66.72-81.06%. Dry weight percentage for titanium is between 0.51 and 0.94%. Iron contains dry weight percentage between 1.97 and 5.42%. Dry weight percentage for aluminum is between 12.11 and 20.31%. Mangan contains dry weight percentage between 0.01 and 0.06% whereas calcium contains dry weight percentage dry weight percentage between 0.08 and 0.37%. Dry weight percentage for magnesium and sodium are between 0.01, 1.07, 0.50 and 0.70%. Potassium and phosphorus contain dry weight percentage between 0.33, 2.94, 0.31 and 2.87%.
Elements like silica, aluminum and ferum are elements containing a high dry weight percentage for brick samples at Candi Sungai Mas (Site 32/34). Scatter plot of K 2 O and CaO dry weight percentage (Raedt et al., 2000) is plotted to see the distribution of bricks substances whereby the purpose of this graph being plotted is because to differentiate every sample based on its main substance (Fig. 1). From observation towards the major elements content in the brick of Sungai Mas has shown that there are two sources has been used to make the brick at this site. The source containing high potassium substance is of samples like d14-2, i10, i7-2, q14-1, q14-2, s6-1, s6-2, s6-3, s12-1, t16-1, t16-2, t18-1, t18-3, u14-1, u15-1 and u15-2.
Another source found at this site and contained low potassium substance is of sample like d14-1, d14-3, i7-1, i7-3, s12-2, s12-3, s20-1, s20-2, t18-2, u12-1 and u12-2. This clearly show the brick maker at this site has used two or more different sources to make their construction bricks. The question now is whether these different source bricks is used to make the same construction structure or there is many construction structures in this area. This is because this site is a well known ancient port site in the 5-12th century.

CONCLUSION
The brick burning technology used in this site is open burning. This has caused some brick to get insufficient temperature and fragile. This can be seen at the excavation of X phase where there is brick which is brittle and fragile. The burning temperature is between 550-800°C. The presence of minerals such as kaolinite, illite and montmorillonite in t16-1, t16-2 and u12-2 samples show that there are bricks which experience burning at the temperature below 550°C and bricks which are brittle and breakable. Based on the content analysis of minerals and compression strength test is obvious that the burning at a low temperature will caused a low compression strength of the brick.
After analyzing the data gathered from the analysis it is found that there is two main groups used as clay sources in at the study area. The first group is made up of d14-1, d14-3, i7-1, i7-3, s12-2, s12-3, s20-1, s20-2, t18-2, u12-1 and u12-2 samples. These samples contain main substance content and trace substance which is nearly the same and it is suggested that these samples is using the same source. The mineral content in these samples are nearly the same whereby the observed mineral obtained are such as quartz and muscovite. Only s12-3 and s12-3 samples contain different mineral such as chloritiod mineral for s12-3 and tremolite and montmorillonite for u12-2 sample. After the data from these samples are matched with clay data studied before this around the research site, it is found that these brick samples do not show any similarity in terms of chemistry composition content with clay chemistry composition around Historic Land Site of Kampung Sungai Mas, Kota Kuala Muda, Kedah. This shows that this brick sample does not use clay source found around Kampung Sungai Mas.
The pottery makers normally take the clay source at the distance of 7-10 km from their industry area and since the area around Kampung Sungai Mas has a lot of clay source, pottery makers in this area may not has used sources from outside or far from their areas. These bricks might be possibly brought from outside of Kampung Sungai Mas to build up specific structures such as temple, port structure since the site standard which function as an ancient port.
Different from brick samples like d14-2, i10, i7-2, q14-1, q14-2, s6-1, s6-2, s6-3, s12-1, t16-1, t16-2, t18-1, t18-3, u14-1, u15-1 and u15-2 which has the same chemistry composition content with the clay chemistry composition around Historic Land Site of Kampung Sungai Mas area. These bricks are similar to the clay taken from around Kota District area, Sungai Terus and Sungai Muda. This shows that this area once has skillful brick makers whereby the bricks used are produced on their own without bringing in brick from outside. Findings from the study indicates that there are more than one clay source used to make the bricks of Candi Sungai Mas (Site 32/34). One of the sources has been taken from Muda River valley and the other sources came from Bujang River valley.