Rock magnetic properties of an An-ei Lava, in
Sakurajima Volcano : Application to
experimental study of geomagnetic
paleointensity
著者名(英)
Naoko UENO, Zhong ZHENG, Takaharu SATO
journal or
publication title
Journal of Toyo University. Natural science
number
49
page range
111-121
year
2005-03
URL
http://id.nii.ac.jp/1060/00002500/
Creative Commons : 表示 - 非営利 - 改変禁止
http://creativecommons.org/licenses/by-nc-nd/3.0/deed.ja
Rock magnetic properties of an An－ei Lava，
in Sakurajima Volcano
一Application to experimental study of geomagnetic paleointensity一
Naoko UENO＊， Zhong ZHENG*＊ and Takaharu SATO＊＊＊
Abstract Magnetic mineral of An－ei lava in Sakurajima volcano that shows self－reversalpTRM（partial thermo－remanent magnetization）in narrow range of temperature
was determined by X－ray diffraction method． It is cleared that magnetic mineralresponsible to NRM（natural remanent magnetization）is not hemoilmenite but
tltano－magnetite Fe（3－．〕Ti。040f which molecular fraction of ulvo－spinel x＝0．25～0．30． Anew version of the Thelliers’method of geomagnetic paleointensity（Zheng’s ver－ sion）was applied in which pTRM was directly measured． Result obtained from the standard Thelliers’method of the Coe’s version of the Thelliers’method was also shown． Experiments on thermal analysis of saturation magnetization and hystere－ sls were conducted from room temperature to liquid N， temperature， which suggested that the titano－magnetite of single domain was the main magnetic carrier． Key words：An－ei lava， XRD， geomagnetic paleointensity， Coe’s version of Thel－ liers’method， Zheng’s version of Thelliers’method．1， Introduction
An－ei lava（sample No． SF17）erupted ir11779 is one of the Sakurajima volcanic rocks reported by Ueno et aL（2004）which presented a partial self－reversal magneti－ zation that might be caused by interaction between the two phases of titano－mag－ netite at double ranges of 245℃～260℃and 330℃～340℃（Fig．1）． X－ray diffraction method（XRD）was carried to identify the magnetic mineral， The result documented the absent of hemo－ilmenite， a well reported carrier of self－ reverSal magnetization． Hysteresis parameter was determined by measurement of hysteresis loop between ホ上野直子：東洋大学文学部英語コミュニケーション学科 〒112－8606東京都文京区白山5－28－20 Department of English Communication， Toyo University，5－28－20， Hakusan， Bunkyo－ku， Tokyo，112－8606 JAPAN ＊＊ A 、重：綜合開発株式会社地球科学事業部 〒133－0057東京都江戸川区西小岩1－3（H6三幸ビル2号館 Sogo Kaihatu Co．， Sanko BuiL 1－30－16， Nishikoiwa， Edogawaku， Tokyo，133－0057 JAPAN ＊榊 ｲ藤’高晴：広島大学総合科学部 〒739－8521 広島県東広島市鏡山1－7－1 Faculty of Integrated Arts＆Scierlces， Hiroshima University，1－7－L Kagamiyama， Higashi－ Hiroshima，739－8521 JAPAN112 Naoko UENo et al． pTRM acquisition curve of Sakurajima lava sample：Anei Iava SF17 （ρ・E＼＜）ト⇔＼Σ匡」°O 2，0E－2 1．5E－2 1，0E－2 5．OE－3 0．OE＋0 一5．OE－3 一1．OE－2 ． ‘ 0 100 200 300 400 Temperature（t C） 500 600 700 1：rrLZ：：11）一：1 Fig． l pTRM acquisition curve of SF17（Ueno et aL 2004） liquid N2 and 100℃to see the temperature dependence of the effective domam slze of main magnetic mineral． Paleointensity of the geomagnetic field was determined by Thelliers’method of both versions of Coe’s and Zheng’s（Zheng et al．2004）to detect the effect of self－ reversal pTRM appeared in narrow range of temperature． Sample is collected in Furusato－machi Kagoshima City， Kyusyu， Japan． The sampling site was located at 31°33’00”N，130°39’45”E． The detailed characteristics of NRM of the same sample are shown in Ueno et a1。（2004）．
2．Magnetic properties of the sample at low temperature
Thermal analysis of saturation magnetization（Js）and hysteresis were perform－ ed by using a Vibrating Sample Magnetometer（VSM）made by Riken Denshi Co． at the magnetic field of 1000 mT（Fig．2）． The Js curve shows characteristic point of temperature at－120℃～－150℃， almost the same temperature as the Verwey point of magnetite（－143℃）． The hysteresis parameters are plotted in Day diagram（Fig．3）． Results from Izu－ Oshima basalt（1986 eruption）and Haruna dacitic lava（Hutatudake eruption）are also shown in the diagram． In the lower temperature， single domain becomes more effective as carrier of magnetism．Rock magDetic properties of an An－ei Lava， in Sakurajirna Vo］caDo 113 Sample： SF17 0．15 0．10 0．05
A
翌 ie．oe ξ ￥ －0．②5 一②．10 一②．15 Hc：122 mT Hcr：209 mT Mr：56．4E ’ Am’lkg Ms：96．8E．3 AmJ／kg 一1000 一500 e H（mT） 5②0 100② Sample：Sakurajima， SF17 （翌＼°り∈嶋山）㊤「 536．40 536．35 536．30 536．25 536．20 536．15 536．10 536．05 536．00 535．95 Weight Atmosphere Applied field Rate 284mg In air 1．OT 1°C／min 一250 －200 －150 Fig． 2 正lysteresis（－194℃） 283．9mg， Max fie工d 1000 mT） 一100 －50 0 50 Temperature（℃） and Js－T at low 100 150 4．OEイ03 3．5E－03 3．OE－03 2．5E－03 2．OE－03 1．5E－03 1．OE－03 5．OE－04 O．OE＋OO 一5．OE－04 ↑や＼ω「や temperature 〔Sample weight ● 3Result of XRD analysis
XRD analysis was performed at Hiroshima University to identify the magnetic mineraL The calibration of the X－ray diffraction rneter was done by Si powder． In contrast to clear pllesence of titano－magnetite peaks， the peak of the ilrnenite一114 Naoko UENO et a］． 0．7一 0．6
05
差o・4褒 Σ0．3 0．2 0．1 O．O D …●－194°C ．．．P』．．．’．…．．…』’．．．．．…．．．； …e ”’15°°ci ． ●－100°Ci ●－solC …PSD°15°・ SFI7 o）”‘ 1 2 3 4 5 6 Hcr／Hc 7 8 1．00 10q
ωΣ亙芝 0．Ol ぐ．ミ、 ＼1
，llk ＼＼、＼ ＼ ＼、、 、＼、s＼、さ
●Oshima1986 △SF17 ！ ◆Haruna17」3」 ・ら・m 1 H。掃。 Fig．3 Day Plot of SF17（the left figure） Day Plot of SF17 with Oshima 1986 and Haruna（the right figure） 100 Sample：SF17 6000 5000 4000a
εb3000
’曇 8 ε 2000 1000 000 1 1 」： 1 ー 1 Legend 氈@magne世e 氈@Ulvospinel ｣ Plagioclase Volt明e 40kV burrent lOO賦 sarget Cu ei l ter Gra坤i te monocrometer rlit 1．DS－0．15頑一1．SS PAS＿UODE CO“IM’OrS rcann三ng speed 2 degノ．min ra団pling（5tEP） 0，02 degS肺othing lO points ‘‘ 1 Data 「 1． ；： 1 1 ．1 ． ． ‘ll
‘ i e ：1 1‘ ：」 ： ． F 1 1 1… ‘ ll◆．φ 1 ‘1 ‘．li
1 ： ．i 1：． ． 氈D ‘！ ‘1 ｜ ‘ ▲ ヌ 一 ． 1 ‘ 三 10 15 20 25 30 Fig．4 35 40 45 2θ（CuKa） XRD Pattern 50 55 60 65 70 hematite series could not be detected， At the same time， it is cleared that magnetic mineral responsible to NRM was a titano－magnetite Fe、」．、：Ti、01， in which molecu－ lar ratio x of ulvo－spinel series was calculated to be x＝0．25～0．30 based on the linear assumption of x with lattice distance． Small x agrees with the result of Js－T analysis that the characteristic temperature is close to the Verwey point of magnetite． XRD data is shown in Fig．4．Rock magnctic properties（）f an An－ei Lava、 in Sakurajima Volcano u T x 115 M
圃
竃」°↑
Fig．5 Coe version of Thelliers’method（Arai diagram） F｜．、，1．＝40．0μT， F．、，、＝46．0μT4．Paleointensity by Coe’s version of Thelliers’method
Coe’s version（Coe，1967）of the standard Thelliers’method was carried out in the nitrogen atmosphere with pTRM test． Sample was cut into a disk shape of l inch in diameter and l cm in thickness． It took about l hour in one cycle of heating and cooling including 10 minutes of hold time at the highest temperature leve1． Heating and cooling were carried in zero magnetic field at the tenユperature so called demagnetization step， then magnetized in the laboratory field parallel to NR．M direction under the same temperature level to room temperature， These steps were carried out from I℃om temperature to the Curie temperature with interval of 20℃～50℃．Paleointensity estimated by Williamson’s method（Kono and Tanaka， 1984）was 46．0μT with data between room temperature and 204℃as shown in Arai diagram（Fig．5）． If data between room temperature and 240℃（self－reversal temperature）was used， the intensity was calculated to be 41μT．5．Paleointensity by Zheng’s version of Thelliers method
To reject data from the temperature block that shows inconsistency between NRM and TRM， Zheng’version of Thelliers method was applieded． In Zheng’version， after thermal demagnetized at the temperature level Ti， pTRM （T1， T，一ルwas acquired by heating f1℃m room temperature T‘．， up to Ti， cooling in artificial field perpendicular to the NRM direction to T－l and without field back to T‘1．In the present case， the artificial field applied was 50μT， Measurement of pTRM and NRM was carried after the 10 mT altemative current field demagneti－ zation（AFD）to reduce the contribution of lower coercivity grains． Effectuality of1］6 Naoko UEN｛・）et aL （1） 2．OE－2 1．6E－2 2 一 E 2 1 3 E O 8 （〕・巨≧）ト◎≧芦旦 4，0E・3 O．OE＋0 （2） 2，0E－2 Slocktng Tornppature ～pectra of pTRH （3） 2．OE・2 （加）blocking T卿rotupe Speパro of 側 40 80 1ZO 150 ZOO 240 Tenperatvre（・c） 280 320 1，6E・2 Z 3 ロ E E ヱ ロ （ひ≧）ト．ふξ 4．OE－3 0．OE＋0 TRM Test
1
Il｛｛
一●一δTRM！引「 －pTRMノδ丁 ■ Paleo偏eぱ（F＿an） TFtSl（2ee’C，晦D F．．lob＝59pT Plateau ● F．．ee，54t槌μT 100 80 （↑ユ）苫Φごo亘£ 0 0 6 4 20 1．6E’2 o ll・2三一Z≧ ピ 誌．OE．3 否 4．OE－3 0，0ε＋0 （Un）blocking Tenprature Spectra of馴＆TMI 100 80 （4） 2．OE－2 1．6巨一2 2 3 ロ ロ E E ど む （∪が≧）ト゜き゜ （⊆）Σ●ご8〆£ 60 40 20 4．OE－3 O．OE◆O 40 80 120 160 200 240 Tenperature（4C） 810cking T綱r武ure Spectrロ of pT閣 pTRM 280 15t Run（NRM） －Znd Run（TRM） 320 0 40 80 120 160 200 2④ 了enperoture（・（） 280 320 40 80 120 160 200 240 Temperatvre（°C） 280 ，1。 Fig．6 Black open circle Red lirle Blue solid circlc Red triangle Fig．6－（3） Apparent and corrected paleointensity Black solid circle Green line Green solid cil．’cle Big green soiid ’ Fig．6－（4） Difference o Fig．6 Zheng version of Thelliers’method Fig．6－（1） Difference between before and after 10mT AC demagnetization Blue……pTRM before AF demag， Red……pTRM after AF demag， 一（2）Apparent and corrected paleointensity of NRM Clrc e fpTRM between NRM and artificial TRMRed………pTRM of NRM
Green……pTRM of artificial TRM ………ﾂNRM／δT
・一・…垂sRM／δT
………`pParent paleointensity … …一・Corrected paleolntensity of artificial TRM ………ﾂTRM／δT
………oTRM／δT
……… `pPallent paleointensity l ……Plateau paleointensity used fol・correctionRock magnetic properties of an An－ei Lava， in Sakurajima Volcano 117 the attendance of NRM loss curve the thermal spectrum of pTRM should be coin－ esame field of 50μT． In Fig． In this case， 200℃could be NRM was seem Large difference between NRM and artifi－ cial TRM produced under 140℃， was caused by the large el・ror of measurement be－ cause of the small pTRM at low temperature as well as by the chemical change in thermal experiment． Plateau of the calculated artificial TRM intensity in 140℃～200℃noted with large green circle in Fig．6－（3）was divided by 50μT to get correction rate． Cor－ rected paleomagnetic field， shown with red triangle in Fig．6－（2）， can be obtained from divided plateau noted with Iarge blue solid circle divided by the correction rate． The average of the corrected paleomagnetic field is 44．3±0．6μT． The detailed procedure of the measurement by Zheng’version of Thelliers’method is illustrated in Fig．7． Fig．8is the photograph of observation by electron rnicroscope． Small magnetite of under 1μm could be the particle of single domain or pseudo－single domain tltanO－magnetlte． In this study， paleointensity of SF17 was calculated under 200℃by both versions of Coe’s and Zheng’s， that is， under the double partial self－reversal ranges of 245 ℃～260℃and 330℃～340℃． SF17 was not a suitable salnple to detect the effect of self－reversal pTRM appeared in narrow range of temperature on paleointensity studv． AF demagnetization was shown in Fig．6－（1）in which pTRM before with after demagnetization was cornpared． Paleointensity calculated in every temperature interva1（T、， T、一］）were shown with blue solid circle il〕Fig，6－（2）， The same specimen was used for TRM test as shown in Fig．6－（3）． TRM was acquired by cooling from the maximum demagnetization temperature level T、、 to room temperature Tu in an artificial field of 50μT parallel to the direction of NRM． Artificial TRM was then thermal demagnetized and applied pTRM with the same procedure and temperature intervals as the case of NRM． Intensity of artifi－ cial TRM， that should be 50μT， calculated in every temperature interva1（Trl， T1） were shown with solid green circle in Fig．6－（3）． If NRM was a thermal remanence originated， and that of TRM should be coincident to each other． And also if mineralogical alteration in experiment was not occurred， the from NRM（1st run）and those from artificial TRM（2nd run） cident to each other， because both were applied in th 6－（4），the spectrum of pTRM in both experilnents was compared． both the pTRM spectra and NRM and TRM losses from 140℃to regarded as coincident to each other． Between 140℃and 200℃， pure thermal magnetization originated．
118 Na⊂）k（：） し．』「・二Nl） （！t al． （ω @三⊂コ〉㌔」三em）①」コ言」Φ○∈①↑ S②0 Me（】suremen七 Pro⊂edure of Thellier－Zheng version 1 70② ．｝
600
5②0400
30②200
10② ② 「 1st cycle H⊥ H＝o H＝② ー．0 ＝ H ．ー 2nd cycle H⊥ H＝o H＝② ．ぽ丁1
H＝0 1 3nd cycle H⊥ H＝O H＝0 ． ．‘ 0 ＝I H Fig．7 040
P］・（：〕C／edui’P 〔：）1L S② 120 160 2②0 240 280 320 360 400 Time（arbitrary unites） Ineas／lrelnent of Zhen9㍉ version of 「Vhelliers’ nlethod Fig．8 b］lect］・on micr（っscopic〔〕bsel・vati（m of SF］7 ． 6Results and Conclusion
N／agneti（．・℃haracters⊂）ll An－ei lava in Sakurajinユa volcan⊂）thaしsh（・） ・s self－1’ever－ sal pTRNI in l／alTow l・al／ge of t．emperatm・e weFe studie⊂1． Ilemo－ilmenite（・⊂）uld not be f（．）しmd in XRD analysis． Maill magneti℃ 1．nineral was titano－rnagnetite Fe．・ TLC）． of mol 1’ati（：）ot』u］v（．）－spilユel x t⊂）be〔）．25．、．0．llう（．〕． The grain sizes are dolninated by pseud〔・）－single domain． Zheng’veFsi〔〕n of Thelliei・s’methc）（｝gave Feasonable pal－ e〔〕intensity（：）f．］4．3±0，6μT． In Zhen9’versio11， PTRN．r was dil’ectly acqしiii’ed． The er「ect of self－reve］・sal P∫1“RIXI could not be found in pa］ec）intensitS．『 study （・）f SF17，Rock magnetic properties of an An－ei Lava， in Sakurajima Volcano 119 because the best pa］eointensity of SF17 was calculated under 200℃， the temperature under the double partial self－reversal ranges， in both methods of Coe’s and Zheng’s．
Acknowledgements
We thank Prof． Hirotomo Ueno of the Kagoshima University for taking samples from Sakurajima Volcano． We also thank Dr． M． Ohkawa of the Hiroshima University for the XRD measurement． This study was supported by a Grant－in－ Aid from Toyo University．References
Coe， R． S．（1967）Paleo－intensities of the earth’s magnetic field determined from Tertiary and Quarternary rocks． J．G．R．，72，3247－3262． Kono， M． and Tanaka， H．（1984）Analysis of the Thelliers method of paieointensity deter－ mination 1：Estimation of statistical errors． J， Geomag． Geoelect．，36，267－284． Ueno， N．（1997）Preliminary report on geornagnetic paleo－intensity study of historic vo1－ canic rocks in Kyusyu， Japan． Journa］of the Toyo Univ，， GeneraユEducation （Nat． Sci，），4ユ，19－38． Ueno， N．， Z． Zheng and H， Ueno（2004）Self－reversal PTRM evidence found in Sakura－ jima volcanic rocks． Journal of Toyo Univ．， Natural Science，4899－132． Zheng， Z．， X． Zhao， and N． Ueno（2004）Anew pre－treatment pTRM method for paleoin－ tensity determination of igneous rocks， Abstracts－2004 Japan Earth and Planetary Science Joint Meeting EO12－022．120 Naoko UENo et al． 要 旨
上野直子・鄭 重・佐藤高晴：桜島安永溶岩の岩石磁気一古地球磁場強度実験への応用一
外部磁場を獲得する温度（ブロッキング温度）が異なるチタノマグネタイト相が近接し
て存在すると，狭い温度範囲で自己反転残留磁化が獲得されることがUeno他（2004）で
確認された．その際に用いた桜島安永溶岩（試料番号SF17）について，基礎データとし
て液体窒素温度までの低温Js－T，低温時のヒステリシス， XRD分析を行い，さらに古地
球磁場強度測定に及ぼす影響を考慮した．古地球磁場強度測定では，通常使われているテ
リエ法Coe Versionの他に，外部磁場に比例しない残留磁化を獲得した温度範囲の残留
磁化を排除して，部分熱残留磁化と自然残留磁化が比例する温度区間だけを使って古地球
磁場強度を求める新しい方法（鄭Version）を用いた．
液体窒素までの低温中で振動試料型磁力計を用いて10kGの磁場中における飽和磁化
の温度変化と約50℃おきのヒステリシスを測定した．ヒステリシスパラメーターをもち
いたDay Plotは単磁区を含む擬似単磁区のチタノマグネタイトが主な磁性鉱物であるこ
とを示している．磁石で抽出した強磁性鉱物を用いたX線回折（X－Ray Diffraction）ではイルメナイトー
ヘマタイト系列の信号がないことを確認し，狭い温度範囲での自己反転残留磁化は榛名の
ヘモイルメナイトによる自己反転とは異なる反転獲得機構であることを確信した．ウルボ
スピネルーマグネタイト系列のチタノマグネタイトの回析線が検出された．このチタノマ
グネタイトの同析線はウルボスピネルよりもマグネタイトの回析線の方に近い．格子距離
がTiの分子比と直線的に比例すると仮定すると， Tiの分子比はx＝o．25～o．30であった．古地球磁場強度測定に使用されるテリエ法はNRMがTRM起源であり，古地球磁場
強度に比例していることが前提である．常温からキュリー温度まで，20度から50度の温
度間隔で，無磁場中の加熱後の残留磁化と同一温度での一定磁場中での加熱後の残留磁化
を用いてアライダイアグラムから計算した場合（テリエ法Coe Version）は常温から約200
度までのデータから46μTが得られた．
新しく開発した，鄭Versionテリエ法では，まずNRMを熱消磁した温度区間にその区
間のみに50μTの定磁場で部分熱残留磁化（pTRM）をNRMに垂直になるように直接つ
けて（Coe Versionでは室温から熱消磁温度までつける），この温度区間に消磁された
NRMと獲得されたpTRMから温度区間ごとに見かけの古地球磁場強度を計算する．次
に，50μTの定磁場でNRM方向に室温からNRMの最高の熱消磁温度まで人「的TRM
をつけて，この人工的TRMをNRMの場合と同様に温度区分熱消磁および定磁場での
部分熱残留磁化の直接付加から人工的TRMの古地球磁場強度を計算する．この人工的
TRMの古地球磁場強度は50μTのはずなので，計算値と50μTの比の値を補正項として
NRMによって得られた見かけの古地球磁場強度を補正する． NRMおよび人工的TRM
の区間別熱消磁後につけた部分熱残留磁化同上を比較して差が大きく出る温度区間の値は
実験中に化学変化が起きたためと考え除去する．有意な差が出ない温度以下で新しい試料
で再度実験すればより説得力のある結果が得られる．この鄭方法で求めた古地球磁場強度
は44．3±0．6μTであった．これは観測値から期待される44μTと一致する．SF17では，
Rock magnetic properties of an An－ei Lava， ln Sakurajima Volcano 121
