Crystal structures of high-pressure phases formed in Si by laser irradiation

Crystal structures of high-pressure phases formed in Si by laser irradiation Abstract Internal modification induced in Si by a permeable pulse laser was investigated by transmission electron microscopy. A laser induced modified volume (LIMV) was a cylindrical rod along the track of a laser beam with the head at the focus of the laser beam. In the LIMV, beside voids, dislocations, micro-cracks and what had been supposed to be an unidentified high-pressure phase (hpp) of Si were observed in LIMV. The so-called ‘hpp’ was identified mostly as diamond Si. laser, (scanning) transmission electron microscopy, high-pressure phase Introduction Laser-induced subsurface modification is of practical as well as academic interest. Practically, it has been successfully applied to a dry-and-nearly debris-free dicing of Si wafer, i.e. the separation of a Si wafer into individual dies [1]. Full understanding of the dicing process necessitates a detailed knowledge on structure of laser induced modified volume (LIMV), which is also of academic interest. Verburg et al. [2] were the first who studied LIMV using transmission electron microscope (TEM) as well as scanning electron microscope (SEM). They showed that LIMV contained amorphous Si and high-pressure phase (hpp) of Si. They claimed that diffraction pattern obtained in LIMV contained diffraction spots the spacing of which did not match with the d-spacings of diamond Si (DS) but matched with those of Si-III/Si-XII. However, they could not construct the reciprocal lattices, since the number of the extra spots was too small. The present authors [3] have carried out comprehensive TEM study on LIMV and showed that LIMV consisted of five distinguished features as summarized in Fig. 1 and Table 1. However, unfortunately there were some ambiguities in the definition of modified features, so that they will be redefined as shown in the third column of Table 1 in the present study. The reason for this is as follows: While the present authors [3] have obtained electron diffraction patterns from what was supposed to be a hpp, which was formed adjacent to a decorated void, they could not identify the crystal structure either, since the diffraction pattern was too complicated. Table 1. Definition of microstructural features associated with LIMV   Ref. [3]  Figure 1 in the present study  ①  Sessile dislocations 1  Sessile dislocations 1  ②  Front void (Type 2 void)  Genuine void(gV)  ③  Sessile dislocations 2  Sessile dislocations 2  ④  2nd void(Type 1 void)  Decorated void(dV)  ④’  hpp  hpp  ④’’  Nano-crack  Nano-crack  ⑤  Heavily dislocated region  Heavily dislocated region  ⑤’  Large crack  Large crack    Ref. [3]  Figure 1 in the present study  ①  Sessile dislocations 1  Sessile dislocations 1  ②  Front void (Type 2 void)  Genuine void(gV)  ③  Sessile dislocations 2  Sessile dislocations 2  ④  2nd void(Type 1 void)  Decorated void(dV)  ④’  hpp  hpp  ④’’  Nano-crack  Nano-crack  ⑤  Heavily dislocated region  Heavily dislocated region  ⑤’  Large crack  Large crack  Here, a decorated void is referred to one which is accompanied with dark (or grey) contrast while a genuine void to one which is not accompanied with such a contrast. Fig. 1. View largeDownload slide Schematic illustration of LIMV (reproduced with permission from [3]). Fig. 1. View largeDownload slide Schematic illustration of LIMV (reproduced with permission from [3]). The object of the present study is to identify the structure of the so-called hpp Si by means of TEM. Experimental procedures Laser processing Detail of the set-up of laser processing is summarized in Table 2. A laser beam was injected into a 0.6 mm thick (001) Si wafer along <110>. Table 2. Set-up of laser processing Wavelength  Energy  Pulse duration  Depth of focus (z)  Spacing laser scan (d)  Scanning speed  1342 nm  0.3 W (3.3 μJ)  90 ns  38 μm  4 μm  4 mm/s  Wavelength  Energy  Pulse duration  Depth of focus (z)  Spacing laser scan (d)  Scanning speed  1342 nm  0.3 W (3.3 μJ)  90 ns  38 μm  4 μm  4 mm/s  Preparation of TEM samples Plan-view TEM sample containing LIMV’s in the plane of foil was prepared using a focused ion beam apparatus (Hitachi High-Technologies FB-2200FIB). The orientation of the TEM foil is [011>. At first, the thickness of the sample was adjusted to ~1 μm in order that the LIMV’s under investigation were embedded in the foil as completely as possible. It was confirmed that at least one void was completely embedded in the TEM specimen. However, it was also found that the specimen was too thick for clear electron diffraction patterns to be obtained from hpp’s. Therefore, the specimen was further thinned using a gentle mill (Technoorg Linda Gentle Mill 3, Ion Mill Model IV8) so as to obtain clear enough diffraction patterns. (S)TEM observation TEM sample thus prepared was examined in a JEM-2100plus operated at an accelerating voltage of 200 kV and also in a high-voltage electron microscope JEM-1000KRS of Nagoya University operated at an accelerating voltage of 1000 kV. The observation was made using the conventional bright field (BF) and dark field (DF) imaging modes [4]. In a JEM-2100plus, scanning TEM (STEM) was also used for high-angle annular dark field (HAADF) imaging and also EDX mapping. Results General S/TEM observation Figure 2a and b shows laser-beam tracks 1 ~ 4 taken in BF and HAADF modes, respectively. Both in the BF mode (Fig. 2a) and HAADF mode (Fig. 2b), none of gV1, gV2, gV3 and gV4 are accompanied with any peculiar contrast. Thus, gV1, gV2, gV3 and gV4 are genuine voids. This is in good agreement with EDX mappings of Si shown in an inset in Fig. 2b. On the other hand, dV1 and dV2 are decorated with black contrasts (hpp1 and hpp2) in the BF mode (Fig. 2a) and with white contrasts in the HAADF mode (Fig. 2b). This HAADF-STEM observation strongly suggests that hpp1 and hpp2 are denser than the matrix DS: This strongly suggests that hpp1 and hpp2 are a non-DS phase, most probably a high-pressure phase of Si, with a higher density than DS. This is the reason why they were referred to high-pressure phase of Si tentatively in Ref. [3]. Figure 2c schematically summarizes the result. Fig. 2. View largeDownload slide TEM micrograph of tracks 1~4 in the BF (a) and HAADF (b) modes. Both in the BF mode (a) and HAADF mode (b) none of gV1, gV2, gV3 and gV4 are accompanied with any peculiar contrast, while dV1 and dV2 are decorated with black contrasts (hpp1 and hpp2) in the BF mode (a) and with white contrasts in the HAADF mode (b). Inset in (b) shows EDX mapping of Si and (c) shows schematic illustration. Fig. 2. View largeDownload slide TEM micrograph of tracks 1~4 in the BF (a) and HAADF (b) modes. Both in the BF mode (a) and HAADF mode (b) none of gV1, gV2, gV3 and gV4 are accompanied with any peculiar contrast, while dV1 and dV2 are decorated with black contrasts (hpp1 and hpp2) in the BF mode (a) and with white contrasts in the HAADF mode (b). Inset in (b) shows EDX mapping of Si and (c) shows schematic illustration. Electron diffraction from ‘hpp’ General description Figures 3–5 show selected area diffraction (SAD) patterns taken from encircled areas in the respective micrographs. Surprisingly, most of the SAD’s are either network of spots from single crystalline DS (A, B, D, F of hpp1 (Fig. 3), A and B of hpp2(Fig. 4) and A and B of hpp3 (Fig. 5)) or Debye rings from polycrystalline DS (E and G of hpp1 (Fig. 3), C of hpp2 (Fig. 4)). However, detailed inspection shows that some of these SAD’s contain extra spots which cannot be explained easily by assuming DS. Fig. 3. View largeDownload slide SAD patterns taken from encircled areas in the track 1. Fig. 3. View largeDownload slide SAD patterns taken from encircled areas in the track 1. Fig. 4. View largeDownload slide SAD patterns taken from encircled areas in the track 2. Fig. 4. View largeDownload slide SAD patterns taken from encircled areas in the track 2. Fig. 5. View largeDownload slide SAD patterns taken from encircled areas in the track 3. Fig. 5. View largeDownload slide SAD patterns taken from encircled areas in the track 3. Detailed analysis of SAD Genuine void and periphery of hpp region Figure 6 shows a typical example of SAD from single crystal DS taken at a genuine void gV2(A). Neither extra spot nor hallo-ring resulting from an amorphous phase is observed. This evidences that a genuine void accompanies no structural change. This is in good agreement with the previous result [3]. Fig. 6. View largeDownload slide A typical example of SAD from single crystal DS taken at a genuine void gV2(A). Fig. 6. View largeDownload slide A typical example of SAD from single crystal DS taken at a genuine void gV2(A). On the other hand, SAD taken near the periphery of the hpp region consists of not only network from single crystalline DS but also hallo-ring, whose origin must be attributed to an amorphous phase, as shown in Fig. 7. Figure 7 even contains extra spots (indicated by circles), albeit very weak. Possible origins of these extra spots will be discussed later in detail Fig. 7. View largeDownload slide SAD taken at hpp1(A). It consists of not only network from single crystalline DS but also hallo-ring, whose origin must be attributed to an amorphous phase. Also, extra spots (indicated by circles) are present. Fig. 7. View largeDownload slide SAD taken at hpp1(A). It consists of not only network from single crystalline DS but also hallo-ring, whose origin must be attributed to an amorphous phase. Also, extra spots (indicated by circles) are present. In the middle of the hpp region, SAD consists mainly of Debye rings from DS, as shown in Fig. 8. This SAD contains extra spots, which are indicated by A and B. Possible origins of these extra spots will be discussed later in detail. Fig. 8. View largeDownload slide SAD taken at hpp2(C). Concentric circles are Debye rings from DS. It consists mainly of Debye rings from DS. Fig. 8. View largeDownload slide SAD taken at hpp2(C). Concentric circles are Debye rings from DS. It consists mainly of Debye rings from DS. SAD from twinned DS Figure 9 shows an enlarged diffraction pattern of hpp2(D). This is a typical diffraction pattern from twin as shown in Fig. 9b [4]. Two patterns with B = <011> common and the mirror plane of {111} are superimposed, as shown in Fig. 9c. Fig. 9. View largeDownload slide (a) SAD taken at hpp2(D). (b) Model of twin in FCC (Reproduced with permission from [4]). (c) Two patterns with B =<011> common and the mirror plane of {111} are superimposed, leading to extra spots shown by crosses. Fig. 9. View largeDownload slide (a) SAD taken at hpp2(D). (b) Model of twin in FCC (Reproduced with permission from [4]). (c) Two patterns with B =<011> common and the mirror plane of {111} are superimposed, leading to extra spots shown by crosses. Origin of the extra spots other than twin Figure 10a shows a typical example taken from F of hpp1 (Fig. 3) at a larger camera constant, the key diagram of which is shown in Fig. 10b. There are three networks constituted by extra spots, as shown by red circle, blue square and green triangle. All of them coincide perfectly with the network from DS (indicated by black circles (●) displaced by some distances. This indicates that extra spots are caused by double reflection of diffraction network of DS [4]. Fig. 10. View largeDownload slide (a) SAD taken at hpp1(F). (b) Key diagram. Fig. 10. View largeDownload slide (a) SAD taken at hpp1(F). (b) Key diagram. Figure 11 shows schematically mechanism of double diffraction. Here, hpp region coexists with DS matrix. If the intensity of one of the diffracted beams from hppregion is strong enough, then this diffracted beam behaves as if the incident beam (secondary incident beam) for underlying DS and forms a secondary diffraction pattern from DS (indicated by cross). If the diffraction pattern from the hppregion under consideration is not strong enough to form a network of their own spots, only the secondary incident beam (◎) appears. On the other hand, the incident electron beam passing hpp region without experiencing diffraction forms a normal diffraction pattern of single crystalline DS as indicated by black circles(●). Therefore, ghpp in Fig. 11 indicates a diffraction spot from the hpp region. Inspection of Fig. 10a suggests that at least three grains in the hpp region formed the secondary incident beam, as they are indicated by red circle, blue square and green triangle, respectively. Fig. 11. View largeDownload slide Schematic illustration of double diffraction. Fig. 11. View largeDownload slide Schematic illustration of double diffraction. It is natural to assume that the secondary incident spot (origin) of a double-diffraction network is strongest one in the network. Thus, B, C and D are the candidates for the double-diffraction networks indicated by red circle, blue square and green triangle, respectively. Spots B and C can be reasonably explained by assuming ghpp = DS 400 and DS 200, respectively; only spot D may correspond to BS8 200 (The d-spacings for Si in a variety of crystal structure have been reviewed [5] and are listed in Table 3). Diffraction spot denoted by E cannot be explained by double diffraction, but all the other ‘extra spots’ are explained by assuming the double diffraction: Table 3. Lattice spacings of DS and hpps of Si [5] DS [a]  Hex. [b]  BC8 [c]  Ortho [d]  hkl  d (Å)  hkl  d (Å)  hkl  d(Å)  hkl  d(Å)  111  3.134  010  2.116  200  3.318  200  2.368  200*  2.714  002  2.076  211  2.709  020  2.251  220  1.919  011  1.886  220  2.678  101  2.245  311  1.637  012  1.482  222  1.915  011  2.219  400  1.357  110  1.222  311  1.774  220  1.632  331  1.245          211  1.619  422  1.108          121  1.590  DS [a]  Hex. [b]  BC8 [c]  Ortho [d]  hkl  d (Å)  hkl  d (Å)  hkl  d(Å)  hkl  d(Å)  111  3.134  010  2.116  200  3.318  200  2.368  200*  2.714  002  2.076  211  2.709  020  2.251  220  1.919  011  1.886  220  2.678  101  2.245  311  1.637  012  1.482  222  1.915  011  2.219  400  1.357  110  1.222  311  1.774  220  1.632  331  1.245          211  1.619  422  1.108          121  1.590  Ortho(VI) [e]  Hex. [f]  Cubic [g]  hkl  d (Å)  hkl  d (Å)  hkl  d (Å)  200  3.960  001  2.373  111  1.923  111  3.091  010  2.188  200  1.267  020  2.380  011  1.609  220  1.181  002  2.368  110  1.264  311  1.007  021  2.126  002  1.187  222  0.9642  Ortho(VI) [e]  Hex. [f]  Cubic [g]  hkl  d (Å)  hkl  d (Å)  hkl  d (Å)  200  3.960  001  2.373  111  1.923  111  3.091  010  2.188  200  1.267  020  2.380  011  1.609  220  1.181  002  2.368  110  1.264  311  1.007  021  2.126  002  1.187  222  0.9642    [a]  [b]  [c]  [d]  [e]  [f]  [g]  Point group  m3  6/mmm  mmm  mmm  mmm  6/mmm  m3m  Space group  Ia3  P63/mmc  Ia3  Imma  Cmca  P6/mmm  Fm-3m  Lattice constant (Å)  a = 5.4282  a = 2.444  a = 6.636  a = 4.737  a = 7.92  a = 2.527  a = 3.34    b = 2.444    b = 4.502  b = 4.759  b = 2.527  c = 4.152    c = 2.55  c = 4.736  c = 2.373  File number in [5]  2  18  1  12  14  17  13    [a]  [b]  [c]  [d]  [e]  [f]  [g]  Point group  m3  6/mmm  mmm  mmm  mmm  6/mmm  m3m  Space group  Ia3  P63/mmc  Ia3  Imma  Cmca  P6/mmm  Fm-3m  Lattice constant (Å)  a = 5.4282  a = 2.444  a = 6.636  a = 4.737  a = 7.92  a = 2.527  a = 3.34    b = 2.444    b = 4.502  b = 4.759  b = 2.527  c = 4.152    c = 2.55  c = 4.736  c = 2.373  File number in [5]  2  18  1  12  14  17  13  Now, with this discussion in mind, let us discuss the origins of extra spots which appear in Fig. 8. Here, apparent extra spots A and B are observed. Supposing that these two extra spots are ones directly diffracted from hpp region, the corresponding d-spacing of both OA and OB is 4.09 Å. This may be reasonably accounted for by 200 of a Si-VI, which is 3.96 Å [5]. Alternative explanation is that extra spots A and B result from double diffraction of DS with the secondary incident beam at Ao. In that case, AoB corresponds to DS 111 and AoA to DS 200. Other possibilities may exist. In any case, the intensity of the extra spots is very weak, and it can be concluded that the amount of non-DS phases is to be very small. Conclusions The aforementioned results can be summarized as follows: Most of hppregions consist of DS. DS is either single crystalline or polycrystalline or twinned. Some evidence is obtained for an amorphous Si. Evidence of non-DS Si phase is quite scanty. It is evident that Verburg et al.’s claim [2] that formation of non-DS and/or amorphous Si is responsible for the formation of voids is of doubtful validity. However, the present conclusion raises another serious question as follows: That polycrystalline DS and twinned DS are formed is evidence to indicate that the original single crystalline DS experiences a heavy plastic deformation. Such a heavy plastic deformation is not to lead to any volume difference in the hppregions. Then, what is responsible for the shrinkage leading to formation of a void? A still more serious question is what is responsible for formation of genuine voids, which appear almost completely free from deformation (dislocation) and phase transformation. Acknowledgements This work was carried out under Nanotechnology Platform of Ministry of Education, Culture, Sports, Science and Technology, Japan. Assistance in preparing the plan-view TEM sample by Ms.Yoko Yoshida in Knowledge Hub Aichi is acknowledged. References 1 Ohmura E, Fukuyo F, Fukumitsu K, and Morita H ( 2006) Internal modified-layer formation mechanism into silicon with nanosecond laser. J. Archiev. Mater. Manuf. Eng.  17: 381– 384. 2 Verburg P C, Smillie L A, Römer G R B E, Haberil B, Bradby J E, Williams J S, and Huis in’t Veld A J ( 2015) Crystal structure of laser-induced subsurface modifications in Si. Appl. Phys. A  120: 683– 691. Google Scholar CrossRef Search ADS   3 Iwata H, Kawaguchi D, and Saka H ( 2017) Electron microscopy of voids in Si formed by permeable laser irradiation. Microscopy.  66: 328– 336. doi:10.1093/jmicro/dfx024. Google Scholar CrossRef Search ADS   4 Saka H ( 1997) Kesshou-densikenbikyougaku (Electron microscopy of crystals) (in Japanese) , ( Uchidarokakuho, Tokyo). 5 For example, Seto Y, CSManager, software. Available at http://pmsl.planet.sci.kobe-u.ac.jp/~seto/?page_id=27&lang=en © The Author(s) 2018. Published by Oxford University Press on behalf of The Japanese Society of Microscopy. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Microscopy Oxford University Press

Crystal structures of high-pressure phases formed in Si by laser irradiation

Loading next page...
 
/lp/ou_press/crystal-structures-of-high-pressure-phases-formed-in-si-by-laser-3PpkSkL4iK
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of The Japanese Society of Microscopy. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0022-0744
eISSN
1477-9986
D.O.I.
10.1093/jmicro/dfx128
Publisher site
See Article on Publisher Site

Abstract

Abstract Internal modification induced in Si by a permeable pulse laser was investigated by transmission electron microscopy. A laser induced modified volume (LIMV) was a cylindrical rod along the track of a laser beam with the head at the focus of the laser beam. In the LIMV, beside voids, dislocations, micro-cracks and what had been supposed to be an unidentified high-pressure phase (hpp) of Si were observed in LIMV. The so-called ‘hpp’ was identified mostly as diamond Si. laser, (scanning) transmission electron microscopy, high-pressure phase Introduction Laser-induced subsurface modification is of practical as well as academic interest. Practically, it has been successfully applied to a dry-and-nearly debris-free dicing of Si wafer, i.e. the separation of a Si wafer into individual dies [1]. Full understanding of the dicing process necessitates a detailed knowledge on structure of laser induced modified volume (LIMV), which is also of academic interest. Verburg et al. [2] were the first who studied LIMV using transmission electron microscope (TEM) as well as scanning electron microscope (SEM). They showed that LIMV contained amorphous Si and high-pressure phase (hpp) of Si. They claimed that diffraction pattern obtained in LIMV contained diffraction spots the spacing of which did not match with the d-spacings of diamond Si (DS) but matched with those of Si-III/Si-XII. However, they could not construct the reciprocal lattices, since the number of the extra spots was too small. The present authors [3] have carried out comprehensive TEM study on LIMV and showed that LIMV consisted of five distinguished features as summarized in Fig. 1 and Table 1. However, unfortunately there were some ambiguities in the definition of modified features, so that they will be redefined as shown in the third column of Table 1 in the present study. The reason for this is as follows: While the present authors [3] have obtained electron diffraction patterns from what was supposed to be a hpp, which was formed adjacent to a decorated void, they could not identify the crystal structure either, since the diffraction pattern was too complicated. Table 1. Definition of microstructural features associated with LIMV   Ref. [3]  Figure 1 in the present study  ①  Sessile dislocations 1  Sessile dislocations 1  ②  Front void (Type 2 void)  Genuine void(gV)  ③  Sessile dislocations 2  Sessile dislocations 2  ④  2nd void(Type 1 void)  Decorated void(dV)  ④’  hpp  hpp  ④’’  Nano-crack  Nano-crack  ⑤  Heavily dislocated region  Heavily dislocated region  ⑤’  Large crack  Large crack    Ref. [3]  Figure 1 in the present study  ①  Sessile dislocations 1  Sessile dislocations 1  ②  Front void (Type 2 void)  Genuine void(gV)  ③  Sessile dislocations 2  Sessile dislocations 2  ④  2nd void(Type 1 void)  Decorated void(dV)  ④’  hpp  hpp  ④’’  Nano-crack  Nano-crack  ⑤  Heavily dislocated region  Heavily dislocated region  ⑤’  Large crack  Large crack  Here, a decorated void is referred to one which is accompanied with dark (or grey) contrast while a genuine void to one which is not accompanied with such a contrast. Fig. 1. View largeDownload slide Schematic illustration of LIMV (reproduced with permission from [3]). Fig. 1. View largeDownload slide Schematic illustration of LIMV (reproduced with permission from [3]). The object of the present study is to identify the structure of the so-called hpp Si by means of TEM. Experimental procedures Laser processing Detail of the set-up of laser processing is summarized in Table 2. A laser beam was injected into a 0.6 mm thick (001) Si wafer along <110>. Table 2. Set-up of laser processing Wavelength  Energy  Pulse duration  Depth of focus (z)  Spacing laser scan (d)  Scanning speed  1342 nm  0.3 W (3.3 μJ)  90 ns  38 μm  4 μm  4 mm/s  Wavelength  Energy  Pulse duration  Depth of focus (z)  Spacing laser scan (d)  Scanning speed  1342 nm  0.3 W (3.3 μJ)  90 ns  38 μm  4 μm  4 mm/s  Preparation of TEM samples Plan-view TEM sample containing LIMV’s in the plane of foil was prepared using a focused ion beam apparatus (Hitachi High-Technologies FB-2200FIB). The orientation of the TEM foil is [011>. At first, the thickness of the sample was adjusted to ~1 μm in order that the LIMV’s under investigation were embedded in the foil as completely as possible. It was confirmed that at least one void was completely embedded in the TEM specimen. However, it was also found that the specimen was too thick for clear electron diffraction patterns to be obtained from hpp’s. Therefore, the specimen was further thinned using a gentle mill (Technoorg Linda Gentle Mill 3, Ion Mill Model IV8) so as to obtain clear enough diffraction patterns. (S)TEM observation TEM sample thus prepared was examined in a JEM-2100plus operated at an accelerating voltage of 200 kV and also in a high-voltage electron microscope JEM-1000KRS of Nagoya University operated at an accelerating voltage of 1000 kV. The observation was made using the conventional bright field (BF) and dark field (DF) imaging modes [4]. In a JEM-2100plus, scanning TEM (STEM) was also used for high-angle annular dark field (HAADF) imaging and also EDX mapping. Results General S/TEM observation Figure 2a and b shows laser-beam tracks 1 ~ 4 taken in BF and HAADF modes, respectively. Both in the BF mode (Fig. 2a) and HAADF mode (Fig. 2b), none of gV1, gV2, gV3 and gV4 are accompanied with any peculiar contrast. Thus, gV1, gV2, gV3 and gV4 are genuine voids. This is in good agreement with EDX mappings of Si shown in an inset in Fig. 2b. On the other hand, dV1 and dV2 are decorated with black contrasts (hpp1 and hpp2) in the BF mode (Fig. 2a) and with white contrasts in the HAADF mode (Fig. 2b). This HAADF-STEM observation strongly suggests that hpp1 and hpp2 are denser than the matrix DS: This strongly suggests that hpp1 and hpp2 are a non-DS phase, most probably a high-pressure phase of Si, with a higher density than DS. This is the reason why they were referred to high-pressure phase of Si tentatively in Ref. [3]. Figure 2c schematically summarizes the result. Fig. 2. View largeDownload slide TEM micrograph of tracks 1~4 in the BF (a) and HAADF (b) modes. Both in the BF mode (a) and HAADF mode (b) none of gV1, gV2, gV3 and gV4 are accompanied with any peculiar contrast, while dV1 and dV2 are decorated with black contrasts (hpp1 and hpp2) in the BF mode (a) and with white contrasts in the HAADF mode (b). Inset in (b) shows EDX mapping of Si and (c) shows schematic illustration. Fig. 2. View largeDownload slide TEM micrograph of tracks 1~4 in the BF (a) and HAADF (b) modes. Both in the BF mode (a) and HAADF mode (b) none of gV1, gV2, gV3 and gV4 are accompanied with any peculiar contrast, while dV1 and dV2 are decorated with black contrasts (hpp1 and hpp2) in the BF mode (a) and with white contrasts in the HAADF mode (b). Inset in (b) shows EDX mapping of Si and (c) shows schematic illustration. Electron diffraction from ‘hpp’ General description Figures 3–5 show selected area diffraction (SAD) patterns taken from encircled areas in the respective micrographs. Surprisingly, most of the SAD’s are either network of spots from single crystalline DS (A, B, D, F of hpp1 (Fig. 3), A and B of hpp2(Fig. 4) and A and B of hpp3 (Fig. 5)) or Debye rings from polycrystalline DS (E and G of hpp1 (Fig. 3), C of hpp2 (Fig. 4)). However, detailed inspection shows that some of these SAD’s contain extra spots which cannot be explained easily by assuming DS. Fig. 3. View largeDownload slide SAD patterns taken from encircled areas in the track 1. Fig. 3. View largeDownload slide SAD patterns taken from encircled areas in the track 1. Fig. 4. View largeDownload slide SAD patterns taken from encircled areas in the track 2. Fig. 4. View largeDownload slide SAD patterns taken from encircled areas in the track 2. Fig. 5. View largeDownload slide SAD patterns taken from encircled areas in the track 3. Fig. 5. View largeDownload slide SAD patterns taken from encircled areas in the track 3. Detailed analysis of SAD Genuine void and periphery of hpp region Figure 6 shows a typical example of SAD from single crystal DS taken at a genuine void gV2(A). Neither extra spot nor hallo-ring resulting from an amorphous phase is observed. This evidences that a genuine void accompanies no structural change. This is in good agreement with the previous result [3]. Fig. 6. View largeDownload slide A typical example of SAD from single crystal DS taken at a genuine void gV2(A). Fig. 6. View largeDownload slide A typical example of SAD from single crystal DS taken at a genuine void gV2(A). On the other hand, SAD taken near the periphery of the hpp region consists of not only network from single crystalline DS but also hallo-ring, whose origin must be attributed to an amorphous phase, as shown in Fig. 7. Figure 7 even contains extra spots (indicated by circles), albeit very weak. Possible origins of these extra spots will be discussed later in detail Fig. 7. View largeDownload slide SAD taken at hpp1(A). It consists of not only network from single crystalline DS but also hallo-ring, whose origin must be attributed to an amorphous phase. Also, extra spots (indicated by circles) are present. Fig. 7. View largeDownload slide SAD taken at hpp1(A). It consists of not only network from single crystalline DS but also hallo-ring, whose origin must be attributed to an amorphous phase. Also, extra spots (indicated by circles) are present. In the middle of the hpp region, SAD consists mainly of Debye rings from DS, as shown in Fig. 8. This SAD contains extra spots, which are indicated by A and B. Possible origins of these extra spots will be discussed later in detail. Fig. 8. View largeDownload slide SAD taken at hpp2(C). Concentric circles are Debye rings from DS. It consists mainly of Debye rings from DS. Fig. 8. View largeDownload slide SAD taken at hpp2(C). Concentric circles are Debye rings from DS. It consists mainly of Debye rings from DS. SAD from twinned DS Figure 9 shows an enlarged diffraction pattern of hpp2(D). This is a typical diffraction pattern from twin as shown in Fig. 9b [4]. Two patterns with B = <011> common and the mirror plane of {111} are superimposed, as shown in Fig. 9c. Fig. 9. View largeDownload slide (a) SAD taken at hpp2(D). (b) Model of twin in FCC (Reproduced with permission from [4]). (c) Two patterns with B =<011> common and the mirror plane of {111} are superimposed, leading to extra spots shown by crosses. Fig. 9. View largeDownload slide (a) SAD taken at hpp2(D). (b) Model of twin in FCC (Reproduced with permission from [4]). (c) Two patterns with B =<011> common and the mirror plane of {111} are superimposed, leading to extra spots shown by crosses. Origin of the extra spots other than twin Figure 10a shows a typical example taken from F of hpp1 (Fig. 3) at a larger camera constant, the key diagram of which is shown in Fig. 10b. There are three networks constituted by extra spots, as shown by red circle, blue square and green triangle. All of them coincide perfectly with the network from DS (indicated by black circles (●) displaced by some distances. This indicates that extra spots are caused by double reflection of diffraction network of DS [4]. Fig. 10. View largeDownload slide (a) SAD taken at hpp1(F). (b) Key diagram. Fig. 10. View largeDownload slide (a) SAD taken at hpp1(F). (b) Key diagram. Figure 11 shows schematically mechanism of double diffraction. Here, hpp region coexists with DS matrix. If the intensity of one of the diffracted beams from hppregion is strong enough, then this diffracted beam behaves as if the incident beam (secondary incident beam) for underlying DS and forms a secondary diffraction pattern from DS (indicated by cross). If the diffraction pattern from the hppregion under consideration is not strong enough to form a network of their own spots, only the secondary incident beam (◎) appears. On the other hand, the incident electron beam passing hpp region without experiencing diffraction forms a normal diffraction pattern of single crystalline DS as indicated by black circles(●). Therefore, ghpp in Fig. 11 indicates a diffraction spot from the hpp region. Inspection of Fig. 10a suggests that at least three grains in the hpp region formed the secondary incident beam, as they are indicated by red circle, blue square and green triangle, respectively. Fig. 11. View largeDownload slide Schematic illustration of double diffraction. Fig. 11. View largeDownload slide Schematic illustration of double diffraction. It is natural to assume that the secondary incident spot (origin) of a double-diffraction network is strongest one in the network. Thus, B, C and D are the candidates for the double-diffraction networks indicated by red circle, blue square and green triangle, respectively. Spots B and C can be reasonably explained by assuming ghpp = DS 400 and DS 200, respectively; only spot D may correspond to BS8 200 (The d-spacings for Si in a variety of crystal structure have been reviewed [5] and are listed in Table 3). Diffraction spot denoted by E cannot be explained by double diffraction, but all the other ‘extra spots’ are explained by assuming the double diffraction: Table 3. Lattice spacings of DS and hpps of Si [5] DS [a]  Hex. [b]  BC8 [c]  Ortho [d]  hkl  d (Å)  hkl  d (Å)  hkl  d(Å)  hkl  d(Å)  111  3.134  010  2.116  200  3.318  200  2.368  200*  2.714  002  2.076  211  2.709  020  2.251  220  1.919  011  1.886  220  2.678  101  2.245  311  1.637  012  1.482  222  1.915  011  2.219  400  1.357  110  1.222  311  1.774  220  1.632  331  1.245          211  1.619  422  1.108          121  1.590  DS [a]  Hex. [b]  BC8 [c]  Ortho [d]  hkl  d (Å)  hkl  d (Å)  hkl  d(Å)  hkl  d(Å)  111  3.134  010  2.116  200  3.318  200  2.368  200*  2.714  002  2.076  211  2.709  020  2.251  220  1.919  011  1.886  220  2.678  101  2.245  311  1.637  012  1.482  222  1.915  011  2.219  400  1.357  110  1.222  311  1.774  220  1.632  331  1.245          211  1.619  422  1.108          121  1.590  Ortho(VI) [e]  Hex. [f]  Cubic [g]  hkl  d (Å)  hkl  d (Å)  hkl  d (Å)  200  3.960  001  2.373  111  1.923  111  3.091  010  2.188  200  1.267  020  2.380  011  1.609  220  1.181  002  2.368  110  1.264  311  1.007  021  2.126  002  1.187  222  0.9642  Ortho(VI) [e]  Hex. [f]  Cubic [g]  hkl  d (Å)  hkl  d (Å)  hkl  d (Å)  200  3.960  001  2.373  111  1.923  111  3.091  010  2.188  200  1.267  020  2.380  011  1.609  220  1.181  002  2.368  110  1.264  311  1.007  021  2.126  002  1.187  222  0.9642    [a]  [b]  [c]  [d]  [e]  [f]  [g]  Point group  m3  6/mmm  mmm  mmm  mmm  6/mmm  m3m  Space group  Ia3  P63/mmc  Ia3  Imma  Cmca  P6/mmm  Fm-3m  Lattice constant (Å)  a = 5.4282  a = 2.444  a = 6.636  a = 4.737  a = 7.92  a = 2.527  a = 3.34    b = 2.444    b = 4.502  b = 4.759  b = 2.527  c = 4.152    c = 2.55  c = 4.736  c = 2.373  File number in [5]  2  18  1  12  14  17  13    [a]  [b]  [c]  [d]  [e]  [f]  [g]  Point group  m3  6/mmm  mmm  mmm  mmm  6/mmm  m3m  Space group  Ia3  P63/mmc  Ia3  Imma  Cmca  P6/mmm  Fm-3m  Lattice constant (Å)  a = 5.4282  a = 2.444  a = 6.636  a = 4.737  a = 7.92  a = 2.527  a = 3.34    b = 2.444    b = 4.502  b = 4.759  b = 2.527  c = 4.152    c = 2.55  c = 4.736  c = 2.373  File number in [5]  2  18  1  12  14  17  13  Now, with this discussion in mind, let us discuss the origins of extra spots which appear in Fig. 8. Here, apparent extra spots A and B are observed. Supposing that these two extra spots are ones directly diffracted from hpp region, the corresponding d-spacing of both OA and OB is 4.09 Å. This may be reasonably accounted for by 200 of a Si-VI, which is 3.96 Å [5]. Alternative explanation is that extra spots A and B result from double diffraction of DS with the secondary incident beam at Ao. In that case, AoB corresponds to DS 111 and AoA to DS 200. Other possibilities may exist. In any case, the intensity of the extra spots is very weak, and it can be concluded that the amount of non-DS phases is to be very small. Conclusions The aforementioned results can be summarized as follows: Most of hppregions consist of DS. DS is either single crystalline or polycrystalline or twinned. Some evidence is obtained for an amorphous Si. Evidence of non-DS Si phase is quite scanty. It is evident that Verburg et al.’s claim [2] that formation of non-DS and/or amorphous Si is responsible for the formation of voids is of doubtful validity. However, the present conclusion raises another serious question as follows: That polycrystalline DS and twinned DS are formed is evidence to indicate that the original single crystalline DS experiences a heavy plastic deformation. Such a heavy plastic deformation is not to lead to any volume difference in the hppregions. Then, what is responsible for the shrinkage leading to formation of a void? A still more serious question is what is responsible for formation of genuine voids, which appear almost completely free from deformation (dislocation) and phase transformation. Acknowledgements This work was carried out under Nanotechnology Platform of Ministry of Education, Culture, Sports, Science and Technology, Japan. Assistance in preparing the plan-view TEM sample by Ms.Yoko Yoshida in Knowledge Hub Aichi is acknowledged. References 1 Ohmura E, Fukuyo F, Fukumitsu K, and Morita H ( 2006) Internal modified-layer formation mechanism into silicon with nanosecond laser. J. Archiev. Mater. Manuf. Eng.  17: 381– 384. 2 Verburg P C, Smillie L A, Römer G R B E, Haberil B, Bradby J E, Williams J S, and Huis in’t Veld A J ( 2015) Crystal structure of laser-induced subsurface modifications in Si. Appl. Phys. A  120: 683– 691. Google Scholar CrossRef Search ADS   3 Iwata H, Kawaguchi D, and Saka H ( 2017) Electron microscopy of voids in Si formed by permeable laser irradiation. Microscopy.  66: 328– 336. doi:10.1093/jmicro/dfx024. Google Scholar CrossRef Search ADS   4 Saka H ( 1997) Kesshou-densikenbikyougaku (Electron microscopy of crystals) (in Japanese) , ( Uchidarokakuho, Tokyo). 5 For example, Seto Y, CSManager, software. Available at http://pmsl.planet.sci.kobe-u.ac.jp/~seto/?page_id=27&lang=en © The Author(s) 2018. Published by Oxford University Press on behalf of The Japanese Society of Microscopy. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

MicroscopyOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial