On the Preparation of Silicontetraiodide, its Some Thermodynamic Properties and Some Experiments on the Crystal Growth of Silicon by the Disproportionate Reaction of Silicondiiodide 利用統計を見る

(1)

On the Preparation of Silicontetraiodide, its Some Thermodynamic

Properties and Some Experiments on the Crystal Growth

of Silicon by the Disproportionating Reaction

of Silicondiiodide

（Received l2 September 1967）

KotaroUCHIMURA TakashiTAKUMA MasakiYUIZUMI ToshioKUNUGI

Synopsis

In carring out the vapor deposition of silicon on seed crystals in closed tubes and in open tubes using iodine as carrier gas， the most important reactions are，（1） Si十212−一→Sil4（G）（2） SiI4 i＝＝＝≧SiI2（G）十12 （3） Si十Sil4 Pt 2Sil2 About these reactions， reaction temperature and equilibrium constants were measured and from the measured value，］F°，4H°and AS°were calculated． Further using above reactions， some simple experiments on the vapor deposition of silicon were carried out． 1． Int】roduction Although many trials on the preparation of single crystals or thin films of silicon by using silicon tetrachloride or trichlorsilane have been studied， there have been reported few studies on the methods by silicontetraiodide． Therefore the authors reported in this paper on the preparation of silicontetraiodide， the measurement of its vapor pressure， the thermal decomposition of silicontetraiodide at high temperature， and the equilibrium of dispro・ portionating reaction of silicontetraiodide， and intended to try to growth single crystals or to make epitaxial growth by silicontetraiodide． 2． Experimental procedure and the results 2．1 Prelりarationげsilicon彦etraiodide 2．1．1Measurement and the results The reaction between silicon and iodine was ＊Apart of this report was reported at the 33th annual meeting of the Electrochemical Society of Japan， April 1966， in Tokyo． To be pub． lished in the Denki−Kagaku in the near future．＊＊Institute of the inorganic synthesis， Yamanashi University．＊＊＊Graduate of Yamanashi Universit y．＊＊＊＊Department of applied chemistry， Faculty of Engineering， Yamanashi University． studied by measuring pressure change to clarify the reaction temperature and reaction mecha− nism． The starting materials used was： Si：（1）Particle size ca．200 mesh Purity 99．9％（SupPlied by Hirano Seizaemon Shoten）（2）Plate of single crystal with（111）plane I2： Reagent for chemical analysis （Supplied by Shuzui Hikotaro Shoten） The measurements were made using Bourdon gauge illustrated in Fig．1． Experiment l a）In part B of Fig．1were inserted single crystals， which was etched by the mixture of hydroflouric acid， acetic acid and nitric acid， and in part A was inserted magnetic hammer and capillary tube， in the latter preliminary distillated small quantity of pure iodine． After the system was evacuated to the order of 10−3∼10−4 mmHg， the capillary was broken by the hammer with cooling part Bby ice． Under the vaccum， iodine was in・ troduced into part B by warming A and cooling B，and then C， D were sealed． The furnace was set and the temperature was gradually raised untill 1050°C． The speed of the temperature rise was about 100°C／h，

一149一

(2)

Dec．1967 Reports of the Faculty of Engineering， Yamanashi University _No．18 Steel Hammer sealed with glass Ampule filled with iqdine Fig．1 Bourdon Gauge used for meaSUrement 800 700 600 500 命：400已玉 300 200 100 Experiment 〃〃〃 F’irst Reheating heatin9 1 −○一一●− 2 −△一一一▲− 3 −ロー一■rv 4 −▽一一▼一＼△ ペノ＼N

k

＼ H ／＼／＼ 1 2 ！ノ！！！！／ 3，／口！口 A Experiment l b）After the experi− ment l a）， the pressure change was measured again on the reaction pro− ducts of the experiment l a）．、 Experiment 2 0。A、。。G、。。3。。4。。5。。6。。7。6。。。．一一 t（℃）＊Deviation from the ideal gal expansion line can be considered to be caused by the decomposition of iodine of I2（g）＝21（g） Fig．2 Pressure change of the mixture of Si（s）十12 Non−etched silicon plate was used and the other operation was the same as in experiment l a）． The weight of used iodine was 50 mg． Experiment 3 Etched silicon crystals and iodine（ca．55 mg）were used． Experiment 4 The silicon powder and iodine （ca．38 mg）were used． The results of measurement were shown in Fig．2． 2．1．2 Consideration of the results The reaction temperature of silicon and iodine depends on the surface state of silicon single crystal plate． Namely， when the surface was pre−etched befour insertion， the reaction temperature was in the range between 500°C and 600°C， whereas nonetched crystals reacted at about 800°C． This seemed to be caused by oxide firm covering the surface of silicon crystal plate． From the Fig．2， it was clearly seen that the pressure of the reaction product was just the half of the iodine pressure． The expected reactions between silicon and iodine are as follows． Si（s）十212（9）＝Sil4（9） Si（s）十12（9）＝SiI2（9） 2Si（s）十312（9）・Si21，（9） Of the above three reaction ， one satisfies the observed

only

relation． more， the saturated vapor pressure curve（GH）＊ of the products obtained by experiment l b） coincided completely with those of silicontetr．a一（1）（2）（3） the first Further一 iodide already published， and the products were slightly yellowish crystalline solids＊＊． There・ fore we assumed that silicon and iodine reacted directly to form silicontetraiodine． 2．2 ノレleasurement げvopor Pre∬ure （’f silicon− tetraiodide ・2．2．1 Procedure and its results The Bourdon Guage made of pyrex glass as the same type as described in Fig．1， was used． The experiments were carried out as described in experiment l a）． After inserting the capil− lary into part A which are丘11ed with puri丘ed silicontetraiodide by vacuum distillation， setting the furnace， and the vapor pressure was measured until 300°C． Fig．3gives the results of the vapor pressure measurements of silicon− tetraiodide． 2．2．2 Discussion From the results of vapor pressure measure− ment of silicontetraiodide the relation between log P（mmHg）and 1／T was plotted in Fig．4．

From Fig．4， the following experimental

formula was obtained： or 2．69 10g P（mmHg）＝一 ×103十7．57 2．69 1091）（atn、）＝一 ×103十4．69 ＊This will be discussed later．＊＊The chemical composition of the crystal was confirmed that Si：1＝1：4by chemical analysis．

(3)

the Crystal Growth of Silicon by the Disproportionating Reaction of Silicondiiodide 800 700 600 500 旬： E 400 300 200 100 ’Oo 100 Pressent Authors Anderson Beltz 200 t（℃） 300 3．0 2．0 ③ 宅曽’ 一 1．0 1．0． 2．0 1／T×IO3 Present・AuthOrs Anderson， Beltz 3．0 Fgi．3 Vapor pressure of SiI4 Fig．4 Vapor pressure of SiI4 This gave the next thermodynamic functions： ∠IF°120∼300。c＝−RTInl）＝12．3×103−21．4 T cal／mol AH°120∼300・c＝12．3×103 dS°120∼300。c＝21．4 e．u． On the other the following formula fo・r the of silicontetraiodide； 3862．7 1・gP・mm…＝23，3809−− s・ cal／mol hand， Andersen et al．i）_gave vapor p「essu「e

一4．9934109T

1）（mmHg）and log P（mmHg）which were calculated from this formula were，indicated with black dotts in Fig．3and 4． From these figures， it was found that the results obtained by． present authors agreed well with those of Andersen． Further， Andersen’s formula leads to the following relations： ∠IF°＝−93．5 T−←17．65×102十9．9 T ln7「 From this，45 and AH were calculated ；、一誓≧−81・6−22・71・gT−AS T・∂；（∂FT）一・7・65×…−9・9T−AH If mean value of temperature of our experi− ments：500°K（227°C）．「was put、into above for一 mula， the following values were obtained： 4H227・c＝12．68×103 cal／mol ∠IS22，。c・＝20．3 e．u． These values for the evaporation of silicon− tetraiodide agreed well with author’s results． 2．3 1）ecomψosition reaction of silicontetraiodide at high temPeratorres For the study of the equilibrium：Si（s）十 SiI4（9）＝＝≧2SiI2（9）， it is necessary to make clear the behavior of silicontetraiodide at high temperatures、 For this reason， the decompo− sition reaction of silicontetraiodide at high temperatures was studied． 2．3．1 Procedure and its results The Bourdon Gauge made of fused silica which is the same type as in Fig．1was used． In this time， only the purified silicontetraiodide was introduced into the part A of the Bourdon Gauge and the behavior of silicontetraiodide at high temperatures was studied． The rate of．temperature increase waS about 100°C／h， specially at temperatures between 900∼1100°C， it was controlled to about 50°C／h． 2．3．2 Discussion of the results The expected reaction for the decompo・ §itiqn of silicOntetraiodide are as follows：

一151一

(4)

Dec．1967 Reports of the Faculty of Engineering， Yamanashi University _No．18 SiI4（9）＝SiI2（9）十21（9）（or I2）（1） SiI4（9）＝SiI3（9）十1（9）（2） Sil4（9）＝SiI（9）十31（9）（3）

and

Sil4（9）＝Si（s）十41（9）（or 21，）．（4） With regard to reaction（2）， SiI3 is considered to be unstable， and it decomposes into SiI2 and I2 following to the next reaction： SiI3（9）＝SiI2（9）十1（9） With regard to reaction（3）， it has not been yet recognized the presence of SiI． With regard to reaction（4）， it was reported that reaction （4） did not occur unless the total pressure is less than 15 mmHg2）． In all our experiments， the total pressure

were more than 15mmHg． So we assumed

that reaction（4）did not occur． Accordingly， pressure lncrease ln our experlments were considered to be caused by reaction（1）． Since the decomposition temperatures were higher than ca．800°C， we assumed that I2 would be decomposed to 210wing to the small partial pressure of iodine． Therefore， the equilibriums were calculated by the following formula： Sil，（9）＝SiI2（9）十21（9） When the degree of dissociation of silicon− tetradiqdide and the original pressure were denoted as x and pot respectively， then the partial pressures at equilibrium would be ex− pressed as follows：・ PSii、＝P。t（1−x）， PSiI，＝P。・・￠、 800 700 600 500 ◎ 84・0 300 200 100 P・＝2P。t・・v So the total pressure is expressed as： ΣP＝Ps ii、＋Pgil，＋P、＝P。t（1＋2x） Since the pressure change at lower temper・ ature of silicontetraiodide with increasing temperature agrees with the expansion line of ideal gas， it can be shown by the next formula： P・1−P・・（・＋2；3） The equilibrium constants were： K・−Ps㌶12−4｛誓3 From these Formulas， the equilibrium con・ stant can be calculated from the experiment， and the relation between log Kp and 1／T was shown in Fig．6． From the above relation， the following formula was obtained．． 1・gK・（mmHg）一一2㌢4×1・・＋2・・7

This gave the following thermodynamic

functions for the equilibrium： 4F°900∼lloo・c＝106．8×103−68．5 T ca1／mol AH°gooん1100・c＝106．8×103 cal／mol ∠tS°goo∼1100。c＝68．5 e．u． On the other hand， Morcher2）reported the following formula for the equilibrium to the Same reaCt10n：、（℃） Fig．5 Decomposition of SiI4（g）at ・high temperatures 邑官 5 富』

竃

’弘 1／T×IO3 0．9 Fig．6 Equilibrium constant of the reac− l tion Sil，（9）＝Sil，十21（9），

(5)

the Crystal GroWth of Silicon by the Disproportionating Reaction of Silicondiiodide 800 700 600 500 載宅連400 300 200 100 0 100 200 30ぴ 400 500 600 700 800 900 1000 1100 温度．（℃） Fig．7 Pressure change caused by the reaction Si（s）十SiI4（9）：：：こ2SiI2（9） 1・9（Ps㌶12）mm−・8・759−2・・86／T （1361∼1526°K） From above formula， AF°， AH°and AS°were obtained as follows： AF°＝97．0×103−59．5 T cal／moI AH°＝97．0×103 cal／mol ∠IS°＝59．5・ e．u． Considering the fact that Morcher’s values were calculated ones which is based on the hydrolysis of silicontetraiodide by Berthelot， it can be said that the auther’s resultsl show good agreement with the above values． 2・4 ExPerimental studγ of the e4uilibrium： Si（s）十SiI4（9）＝＞2SiI2（9） To carry out the epitaxial growth of silicon using SiI4 in a closed vesse1， the accurate studies of the above reaction have a great importance． Nevertheless， only few studies on the above reaction have been reported． 2．4．1 Procedure and its results The same type of Bourdon Gauge made of fused silica as shown in Fig．1was used． The operation was the sam6 as in the study of the reaction of silicon and iodine， but in this time， instead−of iodine， siliconteraiodide was used． The results were shown in Fig．7． 2．4．2 Discussion of the results On the reactiohs between silicontetrai6dide and silicon， following reactions may be con− sidered：盲：巨二旦、3．0 Fig．8 Si（s）十SiI4（9）’＝2SiI2（9） SiI4（9）＝SiI2（9）十21（9） 21（9）十Si（s） On the reaction and silicon， generally occur． But reaction（2）・and（3）must sidered at the same our experiment on the reactiOn between silicon and iodine， they react violently，』 ’ So that reaction（2）and（3）may be considered to lead to the・reaction（1）finally． Accordingly only the reaction（1）may be considered to take place in the tube．・When theしoriginal pressure of silicontetraiodide and the rate．of reaction at t°C were den6ted Pぱand x’、respec・ ti「vely， equilibrium constants＾may be calculated from the experimental data as before．∵The relation between log五p and 1／T was§hown in Fig．8． This was arranged on a“strai’№? line and this may be expressed as：ジ． 1 17．5 1・gK・（mmHg）＝一一T×103＋；15・2’1・fl This gave the ，”following thermgdynamib functionS，： dF°go回100・c＝80．2×10L56．3τ ca1／mol AH°90・・vl1…c＝80．2×103 cal／lng1 ∠fS°goo∼Iloo。c：＝56．3 e．u． On the reaction（1）， Schafer and Morcher published ．，．the following formula：○，∴・，・バ 0．8 ． 0．9＼ 1／T×lO3 Equilibrium constants of the reaction of the rea￠tion Si（s）十’Sil4（9）＝ 2Sil，（9）（1）（2）＝SiI2（9） tt （3） between silicon’tetraiodide ’re：action（1）is said to be con− time．・But according to

一153一

(6)

Dec．1967 一一一一＿1｛壁g吐亙一．Of中g Faculty of Engineering， Yamanashi University _No．18 E _F D A

一

B ■ 45㎝ C （a） c＝−C 1； D E BA 25cm （b） Fig．9 Apparatus used for the crystal growth of Silicon l・9（；il｛：）二mHg−1・，32・−12・93・／7 （12，000 1，600°K） From this formula， AF°， AH°and 4S were calculated as follows： 4F°＝59．2×10L38．6 T cal／mol AH°＝59．2×103 cal／mol ∠IS°＝＝38．6 e．u． There are considerable discrepancy between above results and those of authers． It must be paid attention that the Schafer’s values were based on the calculated ones based on the hydrolysis of silicontetraiodide by Berthelot， and they2）themselves said that these values were not so accurate． On the other hand our data are based on the experiments． 2．5 Some emperiments on the grouth of silicon ssingle cr2stαls吻吻ακiα19γ・wth Some simple experiments on the silicon single crystals and epitaxial growth were carried out on the bases above described disproportionation reaction of silicon diiodide．； 2．5．1 Experimental procedure and its results． Experiment 1． The equipment used was shown in Fig．9（a）and the experiments 1・ were carried out as following manners；in part A of the apparatus was inserted powdered silicon， and in part C was inSerted a cappilary filled with purified silicon tetraiodide． After the evaquation of the system to about 10−4−10−5 mmHg， the cappilary was brqken by a magnetic hammer and the silicon tetraiodide was intro− duced into part A． D， E， F were sealed off， and the furnace was brought around the ap− paratus， and measuring the pressure， the temperature of the system was raised． After 2 hours， high temperature zone A and lower temperature zone B were brought to the desired temperatures． When the epitaxial growth were carried out， silicon single crystal of the（111）cut which was etched before use， was plased on the silica boat in the part B． Experiment 2． Experiment l was carried out in horizontal apParatus， but in experiment 2was carried out in vertical apparatus， and insted of SiI4，12 was used． In part A was inserted powdered silicon， in part C was inserted acapilary filled with desired quantity of purified iodine． After the evaquation of the system to about 10−3−10『4 mmHg， D was sealed， and the capillary was broken by heating by a sharp hand torch． Iodine was transfered to part B． Finaly part E was sealed off， and the high temperature zone A and low temperature zorle were brought to the desired temperature． The experimental results were shown in Table l and 2． An example of the whisker crystal was shown in Photo．1． 2．5．2 Consideration of the results The variety of experimental conditions studied here are relatively restricted and therefore， de丘nitive remarkes cannot be given from the above experiments， but some remarks can be made． When the pressure of the system are less than 80 mmHg， silicon was not transported and the higher the ．． pressure the more large quantity of silicon was transported． When the temperature difference of the two zone was small， the transported silicOn have a tendency to deposited in polycrystaline state， and． when the temperature difference was large， whisker was found to form． Comparing the horizbntal type and vertical type， the latter is more effective in transporting the silicon． This rnay _{be due to the thermal convection How of the} gas・ 3． Conclusio皿 1・．Reaction temperature between iodine and ・ silicon was measured． This temperature depends upon the surface of silicon， the reaction product was confirmed to be Sil4． 2，The vapor pressure of SiI4 was measured and from the results the thermodynamic functions of evaporation were derived；AF°，

(7)

the Crystal Growth of Silicon by the Disproportionating Reaction of Silicondiiodide 4H°and AS°． These values coincided with those reported already． 3．The decomposition reaction of SiI4（9）only was studied and the reaction SiI4＝Sil2十21 was confirmed to occure and the relation between the equilibriurn constants and the temperature was studied to obtain 4F°，4H° and∠IS°for above reaction． 4．The equilibrium for the reaction Sil4十Si＝ 2Sil2 was studied and the relation between equilibrium constants and temperature was obtained to calculate aF°， AH°and∬°from these results． 5．Some experiments on the growth of silicon crystals by the above reaction were carried out， and some results were shown． Acknowledgment Greatful acknowledgment is made to Asahi Glass Company for free supply of silicontetra− iodide of high purity． References 1）H．C． Anderson， L．旺Beltz， J． Am． Chem． Soc．， 75，4828 （1953）． 2）H．Schafer， B． Morcher， Z． anorg． Chem．290， 279 （1957）． Table l Some experiments on the deposltlon by the disproportionation reaction of silicondiiodide．（In the case of apParatus（a）） No． 1 2 3 4 5 6 hlgher temp． lower temp． press．・・n… n・ 1（mmHg） 1， OOO 1，090 1，100 1，100 1，100 1，080 600 750 950 950 950 920 574 833 875 571 788 718 reaction tlme（hr） 9 9 5 6 10 13 silicon deposited （mg／hr） 2．9 5．5 20．0 38．0 21．0 27．6 condition of the products silicon deposited on the reaction tube poly crystaline poly crystaline good whisker deposited in large quantity needle like crystal and polycrystals Table 2 （In the case of apparatus（b）） No． 1 2 3 4 5 higher temp． zone（°C） 1，050 1，050 1，050 1，050 1，150 lower temp． zone（°C） 950 950 950 950 950 press．（mmHg） 1，460 1，170 770 80 765 reaction time（hr） 6 11 5．5 5 5 sllicon deposlted mg／hr 59．2 51．5 46．0 0．0 49．5 conditlon of the products poly crystaline poly crystaline poly crystaline no silicon deposlted at lower tem． perature zone poly crystaline and a little whisker Photo．1 An example of whisker crystals of silicon 一ユ55一