[go: up one dir, main page]

TW201301560A - Layered semiconductor substrate and method of manufacturing same - Google Patents

Layered semiconductor substrate and method of manufacturing same Download PDF

Info

Publication number
TW201301560A
TW201301560A TW101123205A TW101123205A TW201301560A TW 201301560 A TW201301560 A TW 201301560A TW 101123205 A TW101123205 A TW 101123205A TW 101123205 A TW101123205 A TW 101123205A TW 201301560 A TW201301560 A TW 201301560A
Authority
TW
Taiwan
Prior art keywords
layer
semiconductor substrate
doping
lattice constant
layered semiconductor
Prior art date
Application number
TW101123205A
Other languages
Chinese (zh)
Other versions
TWI470831B (en
Inventor
彼得 史托克
甘特 薩克斯
尤特 羅薩默
撒勒 巴哈杜爾 塔帕
海爾姆特 徐文克
彼得 德瑞爾
法蘭克 穆勒
魯道夫 麥爾胡伯
Original Assignee
世創電子材料公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from EP11191535.1A external-priority patent/EP2541589B1/en
Application filed by 世創電子材料公司 filed Critical 世創電子材料公司
Publication of TW201301560A publication Critical patent/TW201301560A/en
Application granted granted Critical
Publication of TWI470831B publication Critical patent/TWI470831B/en

Links

Classifications

    • H10P14/2905
    • H10P14/3211
    • H10P14/3416
    • H10P95/50

Landscapes

  • Crystals, And After-Treatments Of Crystals (AREA)

Abstract

本發明涉及一種分層半導體基材,包括:- 一單晶第一層1,包含至少80%的矽並且具有一第一厚度和一第一晶格常數a1,該第一晶格常數a1是由一第一摻雜元素和一第一摻雜濃度來確定;以及- 一單晶第二層2,包含至少80%的矽並且具有一第二厚度和一第二晶格常數a2,該第二晶格常數a2是由一第二摻雜元素和一第二摻雜濃度確定,該第二層2與該第一層1直接接觸;以及- 一由第III族氮化物構成之單晶第三層4,以使得該第二層2位於該第一層1與該第三層4之間;其中該第二晶格常數大於該第一晶格常數,第一層1和第二層2的晶格係晶格匹配,並且該分層半導體基材的彎曲度(bow)範圍為-50微米至50微米。本發明還涉及用於製造該分層半導體基材的方法。The present invention relates to a layered semiconductor substrate comprising: - a single crystal first layer 1, comprising at least 80% germanium and having a first thickness and a first lattice constant a1, the first lattice constant a1 being Determining by a first doping element and a first doping concentration; and - a single crystal second layer 2 comprising at least 80% germanium and having a second thickness and a second lattice constant a2, the first The two lattice constant a2 is determined by a second doping element and a second doping concentration, the second layer 2 is in direct contact with the first layer 1; and - a single crystal composed of a Group III nitride Three layers 4 such that the second layer 2 is located between the first layer 1 and the third layer 4; wherein the second lattice constant is greater than the first lattice constant, the first layer 1 and the second layer 2 The lattice lattice is lattice matched, and the layered semiconductor substrate has a bowness ranging from -50 micrometers to 50 micrometers. The invention further relates to a method for making the layered semiconductor substrate.

Description

分層半導體基材及其製造方法 Layered semiconductor substrate and method of manufacturing same

本發明係關於基於矽的分層半導體,其可用作用於沉積第III族氮化物層的基材;以及用於製造該基材的方法。 The present invention relates to a germanium-based layered semiconductor that can be used as a substrate for depositing a Group III nitride layer; and a method for fabricating the substrate.

對於磊晶沉積在矽(Si)基材(氮化鎵上矽(GaN-on-Si)層結構)的氮化鎵(GaN)層,由於GaN和Si(111)之間的材料不匹配,從而引起若干基本問題。例如,GaN的熱膨脹係數(TEC)大於矽的熱膨脹係數。對於週期表中第III族的其它元素的氮化物而言同樣如此。熱膨脹係數的這一差異造成在從第III族氮化物層的沉積溫度至室溫的冷卻期間晶圓的極大的凹入彎曲。彎曲量隨著第III族氮化物層的厚度的增加而增加。如果晶圓的彎曲度太高,將無法進行進一步的處理。因此,必須保持較低的彎曲度。 For a gallium nitride (GaN) layer epitaxially deposited on a bismuth (Si) substrate (GaN-on-Si layer structure), due to material mismatch between GaN and Si (111), This causes a number of basic problems. For example, GaN has a coefficient of thermal expansion (TEC) greater than the coefficient of thermal expansion of tantalum. The same is true for the nitrides of the other elements of Group III in the periodic table. This difference in coefficient of thermal expansion results in a large concave curvature of the wafer during cooling from the deposition temperature of the Group III nitride layer to room temperature. The amount of bending increases as the thickness of the Group III nitride layer increases. If the curvature of the wafer is too high, no further processing will be possible. Therefore, it is necessary to maintain a low degree of curvature.

已提出若干解決方案。最新技術是在GaN層的堆疊內沉積產生壓應力的中間層,例如,在低溫下沉積的氮化鋁(AlN)中間層。例如,US2003/0136333A1公開了在矽基材上的GaN層的堆疊內沉積若干AlN中間層。AlN的晶格常數小於GaN的晶格常數,因此,在AlN上生長的GaN層在生長期間處於壓縮中。該壓縮補償了冷卻期間發生的拉伸應力。這實現了要生長在矽基材上的厚且無裂縫的GaN層。 Several solutions have been proposed. The most recent technology is to deposit an intermediate layer that produces compressive stress within a stack of GaN layers, such as an aluminum nitride (AlN) intermediate layer deposited at low temperatures. For example, US 2003/0136333 A1 discloses depositing a number of AlN intermediate layers in a stack of GaN layers on a tantalum substrate. The lattice constant of AlN is smaller than the lattice constant of GaN, and therefore, the GaN layer grown on AlN is in compression during growth. This compression compensates for the tensile stress that occurs during cooling. This achieves a thick and crack-free GaN layer to be grown on the tantalum substrate.

AlN層必須在低溫下沉積,這導致低生長率。因此,AlN層的沉積非常耗時。而且不得不中斷在大約1100℃的相對高溫下發生的GaN層沉積。在GaN和AlN的沉積步驟之間的溫度降低和增加花費了額外的時間。另一缺點在於,在GaN層中因AlN中間層 而產生了額外的應力,這造成GaN層中缺陷密度的增加。 The AlN layer must be deposited at low temperatures, which results in a low growth rate. Therefore, the deposition of the AlN layer is very time consuming. Moreover, GaN layer deposition that occurs at a relatively high temperature of about 1100 ° C has to be interrupted. The temperature reduction and increase between the deposition steps of GaN and AlN takes extra time. Another disadvantage is that the AlN interlayer is in the GaN layer. Additional stress is generated which causes an increase in the defect density in the GaN layer.

或者,可在晶圓的背面沉積產生與晶圓正面由GaN層造成的彎曲相反的彎曲的一或多層。US4830984教示了在矽基材晶圓的背面沉積金屬(例如鎢)矽化物層並且對其退火,以便形成具有一凸起前表面的翹起結構。此後,將砷化鎵(GaAs)層沉積在矽基材晶圓的正面。當將該結構冷卻至室溫時,在正面的GaAs層和在背面的金屬矽化物層的這兩層將張力施加到它們各自的矽基材晶圓側,從而導致GaAs層的大致平坦的表面。應力背面層的沉積需要增加若干處理步驟(背面層的沉積(潛在地翻轉晶圓),對邊緣和正面蝕刻以去除沉積物,對正面額外的拋光),這增加了製造的成本以及複雜性。此外,非Si的背面是對於所有進一步處理步驟的污染源。 Alternatively, one or more layers may be deposited on the backside of the wafer that create a bend that is opposite to the curvature of the wafer front caused by the GaN layer. No. 4,830,984 teaches depositing a metal (e.g., tungsten) telluride layer on the back side of a germanium substrate wafer and annealing it to form a raised structure having a raised front surface. Thereafter, a gallium arsenide (GaAs) layer is deposited on the front side of the germanium substrate wafer. When the structure is cooled to room temperature, tension is applied to the sides of their respective tantalum substrate wafers on both the front GaAs layer and the back metallization layer, resulting in a substantially flat surface of the GaAs layer. . The deposition of the stress backside layer requires several processing steps (deposition of the backside layer (potentially flipping the wafer), etching of the edges and front side to remove deposits, and additional polishing of the front side), which adds cost and complexity to the fabrication. Furthermore, the back side of the non-Si is a source of contamination for all further processing steps.

因此,本發明所要解決的問題是提高具有小彎曲度的高品質的第III族氮化物層,該第III族氮化物層的生產消耗的時間少於已知結構生產消耗的時間。 Accordingly, the problem to be solved by the present invention is to improve a high quality Group III nitride layer having a small degree of curvature, which is produced in a production time less than the time required for the production of known structures.

該問題由分層半導體基材解決,該分層半導體基材包括:- 一單晶第一層1,包含至少80%的矽並且具有一第一厚度和一第一晶格常數a1,該第一晶格常數a1是由一第一掺雜元素和一第一掺雜濃度確定;以及- 一單晶第二層2,包含至少80%的矽並且具有一第二厚度和一第二晶格常數a2,該第二晶格常數a2是由一第二掺雜元素和一第二掺雜濃度確定,該第二層2與該第一層1直接接觸;以及 - 一包含第III族氮化物之單晶第三層4,以使得該第二層2位於該第一層1與該第三層4之間;其中該第二晶格常數a2大於該第一晶格常數a1,第一層1和第二層2的晶格係晶格匹配,並且該分層半導體基材的彎曲度範圍為-50微米至50微米。 The problem is solved by a layered semiconductor substrate comprising: - a single crystal first layer 1, comprising at least 80% germanium and having a first thickness and a first lattice constant a1, the first a lattice constant a1 is determined by a first doping element and a first doping concentration; and - a single crystal second layer 2 comprising at least 80% germanium and having a second thickness and a second lattice a constant a2, the second lattice constant a2 being determined by a second doping element and a second doping concentration, the second layer 2 being in direct contact with the first layer 1; and - comprising a Group III nitride a single crystal third layer 4 such that the second layer 2 is located between the first layer 1 and the third layer 4; wherein the second lattice constant a 2 is greater than the first lattice constant a 1 , The lattice of the layer 1 and the second layer 2 are lattice matched, and the layered semiconductor substrate has a curvature ranging from -50 micrometers to 50 micrometers.

相對於先前技術,本發明使用全矽基的堆疊層來產生應力,以便補償由將沉積在基材上的第III族氮化物層所施加的應力。 In contrast to the prior art, the present invention uses a fully ruthenium-based stacked layer to create stresses to compensate for the stress applied by the Group III nitride layer to be deposited on the substrate.

根據本發明,藉由將至少一個元素增加到在第一和第二層之間產生晶格不匹配的至少一層中,以相對於第一層來更改基材的第二層的組成。掺雜元素及其濃度的選擇方式為第二層的晶格常數大於第一層的晶格常數。在本說明書中,術語「晶格常數」應理解為在材料的晶格鬆弛時該材料的晶格常數。如果在第二層2沉積在第一層1上的情況下,該材料形成應變異質磊晶層,則其實際的面內晶格常數偏離鬆弛狀態下材料的晶格常數。藉由調整掺雜元素、掺雜元素的濃度(以下稱為「掺雜濃度」)以及第一和第二層的厚度,產生導致晶圓凸起彎曲的特定應力量。在基材中產生的應力量足以抗衡由沉積在基材第二層上或上方的第III族氮化物層生成的應力。 According to the invention, the composition of the second layer of the substrate is altered relative to the first layer by adding at least one element to at least one layer that creates a lattice mismatch between the first and second layers. The doping element and its concentration are selected such that the lattice constant of the second layer is greater than the lattice constant of the first layer. In the present specification, the term "lattice constant" is understood to mean the lattice constant of the material when the lattice of the material is relaxed. If the material forms a deformed epitaxial layer in the case where the second layer 2 is deposited on the first layer 1, the actual in-plane lattice constant deviates from the lattice constant of the material in the relaxed state. By adjusting the doping element, the concentration of the doping element (hereinafter referred to as "doping concentration"), and the thicknesses of the first and second layers, a specific amount of stress causing the wafer bump to be bent is generated. The amount of stress generated in the substrate is sufficient to counteract the stress generated by the Group III nitride layer deposited on or above the second layer of the substrate.

本發明還可以與其它技術結合以減少彎曲度,例如,上面描述的先前技術,或者使用基材上的在先彎曲(preferential bow)。 The invention may also be combined with other techniques to reduce bowing, such as the prior art described above, or the use of a preferential bow on a substrate.

本發明的優點在於:因為消除了對於將應力補償層放入GaN的磊晶沉積的需要,並且簡化了(Al)GaN緩衝層的生長,所以提高了沉積在第二層上的第III族氮化物層的品質,並且減少了處理時間(即,降低了成本)。 An advantage of the present invention is that the III-nitride nitrogen deposited on the second layer is improved because the need for epitaxial deposition of the stress compensation layer into GaN is eliminated and the growth of the (Al)GaN buffer layer is simplified. The quality of the layer and reduces processing time (ie, reduces cost).

由於第III族氮化物層內缺少應力產生層,從而帶來了更高的品質。這減少了第III族氮化物層中的缺陷密度。採用導致這些層的高晶體品質的標準Si磊晶處理方法,能夠容易地將第一層或第二層磊晶沉積在其它層上。 The lack of a stress-creating layer within the Group III nitride layer results in higher quality. This reduces the defect density in the Group III nitride layer. The first or second layer can be easily epitaxially deposited on other layers using standard Si epitaxial processing methods that result in high crystal quality of these layers.

與在(Al)GaN緩衝層中沉積應力補償層相比較,經摻雜的Si層的沉積具有更高的產量(更低的成本)。因此,簡化了整個過程,並且降低了製造成本。 The deposition of the doped Si layer has a higher yield (lower cost) than depositing the stress compensation layer in the (Al)GaN buffer layer. Therefore, the entire process is simplified and the manufacturing cost is reduced.

將參照三個附圖來詳細描述本發明。 The invention will be described in detail with reference to the three drawings.

根據本發明的分層半導體基材由至少三層構成:單晶第一層1含有至少80%且較佳至少90%的矽。(在本說明書中,所有層的成分係以原子百分比表示。)該第一層1具有一第一厚度和一第一晶格常數a1。由該第一掺雜元素和該第一掺雜元素的濃度(在本說明書中稱作「第一掺雜濃度」)來確定第一晶格常數a1。晶格取向較佳為(111)。 The layered semiconductor substrate according to the invention consists of at least three layers: the single crystal first layer 1 contains at least 80% and preferably at least 90% of ruthenium. (In the present specification, the composition of all layers is expressed in atomic percent.) The first layer 1 has a first thickness and a first lattice constant a 1 . The first lattice constant a 1 is determined by the concentration of the first doping element and the first doping element (referred to as "first doping concentration" in the present specification). The lattice orientation is preferably (111).

單晶第二層2也含有至少80%並較佳至少90%的矽。其具有一第二厚度和一第二晶格常數a2。由該第二掺雜元素和該第二掺雜元素的濃度(在本說明書中稱作「第二掺雜濃度」)來確定第二晶格常數a1The second layer 2 of the single crystal also contains at least 80% and preferably at least 90% of ruthenium. It has a second thickness and a second lattice constant a 2 . The second lattice constant a 1 is determined by the concentration of the second doping element and the second doping element (referred to as "second doping concentration" in the present specification).

第二層2與第一層1直接接觸。在第一層1和第二層2之間沒有其它層。第二層2被沉積為形成具有與第一層1的面內晶格常數相匹配的第二層2的實際面內晶格常數的應變異質磊晶層。因為第二晶格常數a2大於第一晶格常數a1,因此第二層2相對第一層1存在壓縮應變。在沉積第三層4之前,這造成(關於與第一 層1相對的第二層2表面的)凸起彎曲,在第2圖中示出了該狀態。 The second layer 2 is in direct contact with the first layer 1. There are no other layers between the first layer 1 and the second layer 2. The second layer 2 is deposited to form a strain-of-variant epitaxial layer having an actual in-plane lattice constant of the second layer 2 that matches the in-plane lattice constant of the first layer 1. Since the second lattice constant a 2 is greater than the first lattice constant a 1 , the second layer 2 has a compressive strain with respect to the first layer 1 . This causes a convex curvature (with respect to the surface of the second layer 2 opposite the first layer 1) before depositing the third layer 4, which is shown in Fig. 2.

單晶第三層4(第3圖)由週期表的第III族的元素氮化物構成(在本說明書中稱為「第III族氮化物」)。這些第III族氮化物中,氮化鋁(AlN)、氮化鎵(GaN)以及氮化銦(InN)及其混合物尤為重要。通常,分層半導體基材包括多於一個的第III族氮化物層(稱為「第三層4」)。較佳地,第III族氮化物層的堆疊開始於AlN層,以便化學隔離矽基基材,準備二維的層的生長以及將晶格常數從第二層2的晶格常數a2調整為最上面的第III族氮化物層的晶格常數。尤佳地,最上面的第III族氮化物由GaN構成。 The third layer 4 of the single crystal (Fig. 3) is composed of an elemental nitride of Group III of the periodic table (referred to as "Group III nitride" in the present specification). Among these Group III nitrides, aluminum nitride (AlN), gallium nitride (GaN), and indium nitride (InN) and mixtures thereof are particularly important. Typically, a layered semiconductor substrate includes more than one Group III nitride layer (referred to as "third layer 4"). Preferably, the stacking of the Group III nitride layer begins with an AlN layer to chemically isolate the ruthenium-based substrate, prepare for growth of the two-dimensional layer, and adjust the lattice constant from the lattice constant a 2 of the second layer 2 to The lattice constant of the uppermost Group III nitride layer. More preferably, the uppermost Group III nitride is composed of GaN.

在根據本發明的分層半導體基材中,第二層2位於第一層1和第三層4之間。第三層4可以與第二層2直接接觸。然而,也能夠在第二層2和第三層4之間放置一個或多個中間層3。 In the layered semiconductor substrate according to the invention, the second layer 2 is situated between the first layer 1 and the third layer 4. The third layer 4 can be in direct contact with the second layer 2. However, it is also possible to place one or more intermediate layers 3 between the second layer 2 and the third layer 4.

根據本發明的分層半導體結構較佳係具有圓形晶圓的形狀。 The layered semiconductor structure according to the present invention preferably has the shape of a circular wafer.

由於第一、第二和第三層的應變的相互補償,分層半導體基材的彎曲度之範圍係如由ASTM F534 3.1.2和SEMI MF534定義的-50微米至50微米。如上面所解釋的,第III族氮化物的熱膨脹係數(TEC)大於矽。因此,當第III族氮化物層4(第1圖)在高溫下磊晶沉積在矽基材5上,並在此後冷卻至室溫時,GaN層4比矽基材5收縮得更多。這導致如第1圖中所示的具有相當大的凹入彎曲的結構。根據本發明,(由冷卻的GaN層的拉伸應變產生的)該彎曲由矽基基材中的反向應變來補償。在第2圖中示出了矽基基材的結構。矽基基材包括一具有一第一晶格常數的第一層1和一與該第一層1直接接觸的第二層2,並且該第二層2的晶 格常數大於該第一層1。 Due to the mutual compensation of the strain of the first, second and third layers, the degree of curvature of the layered semiconductor substrate ranges from -50 microns to 50 microns as defined by ASTM F534 3.1.2 and SEMI MF534. As explained above, the Group III nitride has a coefficient of thermal expansion (TEC) greater than 矽. Therefore, when the Group III nitride layer 4 (Fig. 1) is epitaxially deposited on the tantalum substrate 5 at a high temperature, and thereafter cooled to room temperature, the GaN layer 4 shrinks more than the tantalum substrate 5. This results in a structure having a relatively large concave curvature as shown in Fig. 1. According to the invention, the curvature (produced by the tensile strain of the cooled GaN layer) is compensated by the reverse strain in the ruthenium-based substrate. The structure of the ruthenium-based substrate is shown in Fig. 2. The ruthenium-based substrate comprises a first layer 1 having a first lattice constant and a second layer 2 in direct contact with the first layer 1, and the crystal of the second layer 2 The lattice constant is greater than the first layer 1.

在本發明的第一實施態樣中,第一和第二掺雜元素是相同的,並且僅第一和第二掺雜濃度不同。 In a first embodiment of the invention, the first and second doping elements are the same and only the first and second doping concentrations are different.

在第一實施態樣中,如果掺雜元素的共價原子半徑小於矽,則選擇高於第二掺雜濃度的第一掺雜濃度。因此,第一晶格常數a1變得小於第二晶格常數a2。硼(B)是共價原子半徑小於矽的典型掺雜元素,並且較佳在該情況下使用。第二掺雜濃度可以低至零,但是在該情況下,第一掺雜濃度必須大於零。 In a first embodiment, if the covalent atomic radius of the doping element is less than 矽, a first doping concentration that is higher than the second doping concentration is selected. Therefore, the first lattice constant a 1 becomes smaller than the second lattice constant a 2 . Boron (B) is a typical doping element having a covalent atomic radius smaller than 矽, and is preferably used in this case. The second doping concentration can be as low as zero, but in this case the first doping concentration must be greater than zero.

在第一實施態樣中,如果掺雜元素具有較大的共價原子半徑,則選擇高於第一掺雜濃度的第二掺雜濃度。因此,第一晶格常數a1再一次變得小於第二晶格常數a2。鍺(Ge)或銻(Sb)是共價原子半徑大於矽的典型掺雜元素,並且較佳在該情況下使用。該第一掺雜濃度可以低至零,但是在該情況下,第二掺雜濃度必須大於零。 In a first embodiment, if the doping element has a larger covalent atomic radius, a second doping concentration that is higher than the first doping concentration is selected. Therefore, the first lattice constant a 1 becomes smaller than the second lattice constant a 2 again . Germanium (Ge) or germanium (Sb) is a typical doping element having a covalent atomic radius greater than 矽, and is preferably used in this case. The first doping concentration can be as low as zero, but in this case the second doping concentration must be greater than zero.

在本發明的第二實施態樣中,第一和第二掺雜元素是不同的元素。在該情況下,第一掺雜元素的共價原子半徑小於矽,且第二掺雜元素的共價原子半徑大於矽。較佳地,第一掺雜元素是硼,且第二掺雜元素是鍺或銻。 In a second embodiment of the invention, the first and second doping elements are different elements. In this case, the covalent atomic radius of the first doping element is less than 矽, and the covalent atomic radius of the second doping element is greater than 矽. Preferably, the first doping element is boron and the second doping element is lanthanum or cerium.

藉由中間層3(如果有的話)和第三層4造成的(導致凹入彎曲)的應變量來確定所需的(導致矽基基材,即第一層1和第二層2的堆疊的凸起彎曲)應變量。依序藉由層3和層4的厚度、晶格常數以及TEC(並因此藉由其組成)來確定該凹入彎曲量。類似地,第一層1和第二層2的堆疊的凸起彎曲量取決於層1和層2的厚度和晶格常數(並因此取決於組成)。因此,選擇第一層1 和第二層2的厚度和晶格常數(並因此及其組成),以使得在第一和第二層的堆疊中引起所需的應變數。可以藉由簡單的實驗來確定這些參數的合適組合。藉由這些參數的合適選擇,能夠獲得具有-50微米至50微米、較佳-10微米至10微米的非常小的彎曲度的分層半導體基材。 Determining the desired (causing the base layer 1 and the second layer 2 by the amount of strain caused by the intermediate layer 3 (if any) and the third layer 4 (resulting in a concave bend) The stacked bumps are curved). The amount of concave bending is determined in sequence by the thicknesses of the layers 3 and 4, the lattice constant, and the TEC (and thus by its composition). Similarly, the amount of convex curvature of the stack of the first layer 1 and the second layer 2 depends on the thickness and lattice constant of the layers 1 and 2 (and thus on the composition). So choose the first layer 1 And the thickness and lattice constant of the second layer 2 (and hence its composition) such that the desired number of strains is induced in the stack of the first and second layers. A suitable combination of these parameters can be determined by simple experimentation. By suitable selection of these parameters, a layered semiconductor substrate having a very small degree of curvature of from -50 microns to 50 microns, preferably from -10 microns to 10 microns, can be obtained.

如由Frank和van der Merve(F.C.Frank和J.H.van der Merve,Proc.Roy.Soc.,A198,216(1949))首先預測的,在一給定的晶格常數差異下,當厚度超過臨界厚度時,將達到構造第一層1和第二層2的應力層堆疊的上限。對於Si(111)層結構,無法良好地建立臨界厚度,並且必須以實驗測試上限,例如,藉由測量層堆疊的彎曲度。在臨界厚度以下,彎曲度的改變可以調整到斯托尼(Stoney)和弗羅因德(Freund)等式(參見1987(1999)年弗羅因德(Freund)等人的Appl.Phys.Lett.74),以促進最終層結構的最佳化。 As first predicted by Frank and van der Merve (FCFrank and JHvan der Merve, Proc. Roy. Soc., A198, 216 (1949)), at a given lattice constant difference, when the thickness exceeds the critical thickness At this time, the upper limit of the stack of stress layers constructing the first layer 1 and the second layer 2 will be reached. For the Si (111) layer structure, the critical thickness cannot be well established, and the upper limit must be tested experimentally, for example, by measuring the curvature of the layer stack. Below the critical thickness, the change in curvature can be adjusted to the Stoney and Freund equations (see 1987 (1999) Freund et al., Appl. Phys. Lett. 74) to promote the optimization of the final layer structure.

可以藉由包括以下步驟的方法來製造根據本發明的分層半導體基材:- 生長一包含至少80%的矽以及具有第一掺雜濃度的第一掺雜元素的單晶體,該單晶體具有第一晶格常數;- 從單晶體上切下至少一個晶圓;- 將該晶圓的厚度減小到第一厚度,該晶圓構成第一層1;- 在該晶圓的一個表面上磊晶沉積第二層2;以及- 磊晶沉積由第III族氮化物構成的第三層4。 The layered semiconductor substrate according to the present invention can be fabricated by a method comprising the steps of: growing a single crystal comprising at least 80% germanium and a first doping element having a first doping concentration, the single crystal having a first a lattice constant; - cutting at least one wafer from a single crystal; - reducing the thickness of the wafer to a first thickness, the wafer forming a first layer 1; - epitaxial deposition on one surface of the wafer The second layer 2; and - epitaxial deposition of the third layer 4 composed of a Group III nitride.

在該方法的第一步驟中,生長含有至少80%的矽且較佳至少90%的矽的單晶體。較佳地,使用公知的柴可斯基(Czochralski)法來生長晶體。如果第一掺雜濃度大於零,則以適於將第一掺雜濃 度併入生長的矽單晶體的濃度,對矽熔體摻雜第一掺雜元素。所生長的單晶體具有取決於第一摻雜元素和第一摻雜濃度的第一晶格常數。採用柴可斯基法,目前能夠生長直徑高達450毫米的矽單晶體。 In the first step of the process, a single crystal containing at least 80% cerium and preferably at least 90% cerium is grown. Preferably, the crystal is grown using the well-known Czochralski method. If the first doping concentration is greater than zero, then to be suitable for the first doping concentration The concentration of the grown single crystal is incorporated into the tantalum melt to dope the first doping element. The grown single crystal has a first lattice constant that is dependent on the first dopant element and the first dopant concentration. With the Chaisky method, it is currently possible to grow single crystals with a diameter of up to 450 mm.

在第二步驟中,從單晶體上切下至少一個晶圓。根據本發明,切片方法沒有特別限制。出於經濟原因,較佳使用多線鋸(MWS),以便同時將單晶體切成多個晶圓。 In the second step, at least one wafer is cut from the single crystal. According to the present invention, the slicing method is not particularly limited. For economic reasons, a multi-wire saw (MWS) is preferably used to simultaneously cut a single crystal into a plurality of wafers.

在第三步驟中,去除由切片步驟損壞的單晶晶圓的表面層,並且將晶圓的厚度減小到等於所期望的第一厚度的值。這由機械、化學以及化學-機械處理步驟的組合來完成。機械處理步驟例如是研磨(lapping)或磨削(grinding)。化學處理可以是液相蝕刻或氣相蝕刻。拋光是廣泛使用的化學-機械處理。所得到的晶圓構成待形成的分層半導體基材的第一層1。 In a third step, the surface layer of the single crystal wafer damaged by the slicing step is removed and the thickness of the wafer is reduced to a value equal to the desired first thickness. This is done by a combination of mechanical, chemical and chemical-mechanical processing steps. The mechanical processing steps are, for example, lapping or grinding. The chemical treatment may be liquid phase etching or vapor phase etching. Polishing is a widely used chemical-mechanical treatment. The resulting wafer constitutes the first layer 1 of the layered semiconductor substrate to be formed.

在第四步驟中,由至少80%的矽且較佳至少90%的矽構成的第二層2磊晶沉積在第一層1的一個表面上。較佳地,使用化學氣相沉積(CVD)來沉積磊晶的第二層。矽源氣體較佳為三氯矽烷。如果第二掺雜濃度大於零,必須在適合將第二掺雜濃度併入生長的磊晶層的濃度下,提供供給掺雜元素的附加源氣體。在第二掺雜元素是鍺的情況下,較佳使用四氯化鍺(GeCl4)或鍺烷(GeH4)作為附加源氣體。 In a fourth step, a second layer 2 consisting of at least 80% germanium and preferably at least 90% germanium is epitaxially deposited on one surface of the first layer 1. Preferably, chemical vapor deposition (CVD) is used to deposit the second layer of epitaxial. The helium source gas is preferably trichlorodecane. If the second doping concentration is greater than zero, an additional source gas that supplies the doping element must be provided at a concentration suitable to incorporate the second doping concentration into the grown epitaxial layer. In the case where the second doping element is germanium, germanium tetrachloride (GeCl 4 ) or germane (GeH 4 ) is preferably used as the additional source gas.

如果需要,可以將一或多個中間層3磊晶沉積在第二層2的表面上。 One or more intermediate layers 3 may be epitaxially deposited on the surface of the second layer 2, if desired.

在該方法的最後一步中,由諸如AlN、GaN或InN或其混合物的第III族氮化物構成的第三層4磊晶沉積在第二層2的表面上(或 者如果存在附加之層3,則在附加之層3的表面上)。也能夠(並係較佳地)繼續生長多於一層的第III族氮化物層。較佳地,首先使用三甲基鋁(Al(CH3)3)和氨(NH3)作為前驅體來生長AlN晶種層。此後,在生長GaN層之前,具有增加的Ga濃度的AlGaN層可以用作一轉變層。通常,將三甲基鎵(Ga(CH3)3)用作前驅體,並且使用標準的MOCVD反應器在700℃和1200℃的生長溫度下進行沉積。 In the final step of the method, a third layer 4 of a Group III nitride such as AlN, GaN or InN or a mixture thereof is epitaxially deposited on the surface of the second layer 2 (or if an additional layer 3 is present, Then on the surface of the additional layer 3). It is also possible (and preferably better) to continue to grow more than one layer of the Group III nitride layer. Preferably, the AlN seed layer is first grown using trimethylaluminum (Al(CH 3 ) 3 ) and ammonia (NH 3 ) as precursors. Thereafter, an AlGaN layer having an increased Ga concentration may be used as a transition layer before the GaN layer is grown. Typically, trimethylgallium (Ga(CH 3 ) 3 ) is used as a precursor and deposition is carried out at a growth temperature of 700 ° C and 1200 ° C using a standard MOCVD reactor.

通過第一和第二掺雜元素、第一和第二掺雜濃度以及第一和第二厚度的適當選擇,能夠產生包括第III族氮化物的分層半導體基材,該分層半導體基材的彎曲度絕對值為50微米或以下或者甚至10微米或以下。 By layering the first and second doping elements, the first and second doping concentrations, and the first and second thicknesses, a layered semiconductor substrate comprising a Group III nitride can be produced, the layered semiconductor substrate The absolute value of the curvature is 50 microns or less or even 10 microns or less.

在所描述的製造方法中可包括其他步驟,例如,用於使晶圓的邊緣成形的一或多個步驟(邊緣修整、邊緣拋光)、清洗步驟、檢查步驟以及封裝步驟。 Other steps may be included in the described fabrication method, such as one or more steps (edge trimming, edge polishing), cleaning steps, inspection steps, and packaging steps for shaping the edges of the wafer.

可以將根據本發明的分層半導體基材用作用於製造電子器件(例如,電源元件)或類似發光二極體(LEDs)的光電元件的基材。 The layered semiconductor substrate according to the present invention can be used as a substrate for manufacturing photovoltaic elements of electronic devices (for example, power supply elements) or similar light emitting diodes (LEDs).

1‧‧‧第一層 1‧‧‧ first floor

2‧‧‧第二層 2‧‧‧ second floor

3‧‧‧中間層 3‧‧‧Intermediate

4‧‧‧第三層 4‧‧‧ third floor

5‧‧‧矽基材 5‧‧‧矽 substrate

第1圖圖解地示出沉積在矽基材上的第III族氮化物層(非根據本發明);第2圖圖解地示出在沉積第III族氮化物層之前之根據本發明的分層半導體基材的中間產物;第3圖圖解地示出根據本發明的分層半導體基材,該分層半導體基材包括第一和第二基於矽的層、中間層以及第III族氮化物層。 Figure 1 diagrammatically shows a Group III nitride layer deposited on a germanium substrate (not according to the invention); Figure 2 diagrammatically shows layering according to the invention prior to deposition of a Group III nitride layer An intermediate product of a semiconductor substrate; FIG. 3 diagrammatically illustrates a layered semiconductor substrate including first and second ruthenium-based layers, an intermediate layer, and a Group III nitride layer, in accordance with the present invention .

1‧‧‧第一層 1‧‧‧ first floor

2‧‧‧第二層 2‧‧‧ second floor

3‧‧‧中間層 3‧‧‧Intermediate

4‧‧‧第三層 4‧‧‧ third floor

Claims (11)

一種分層半導體基材,包括:- 一單晶第一層(1),包含至少80%的矽並且具有一第一厚度和一第一晶格常數(a1),該第一晶格常數(a1)是由一第一摻雜元素和一第一摻雜濃度來確定;以及- 一單晶第二層(2),包含至少80%的矽並且具有一第二厚度和一第二晶格常數(a2),該第二晶格常數(a2)是由一第二摻雜元素和一第二摻雜濃度來確定,該第二層(2)係與該第一層(1)直接接觸;以及- 一由第III族氮化物構成之單晶第三層(4),以使得該第二層(2)位於該第一層(1)與該第三層(4)之間;其中該第二晶格常數(a2)大於該第一晶格常數(a1),該第一層(1)和該第二層(2)的晶格係晶格匹配,並且該分層半導體基材的彎曲度(bow)範圍為-50微米至50微米。 A layered semiconductor substrate, comprising: - a first single crystal layer (1), comprising at least 80% of silicon and having a first thickness and a first lattice constant (A 1), the first lattice constant (a 1 ) is determined by a first doping element and a first doping concentration; and - a single crystal second layer (2) comprising at least 80% germanium and having a second thickness and a second a lattice constant (a 2 ), the second lattice constant (a 2 ) being determined by a second doping element and a second doping concentration, the second layer (2) being associated with the first layer ( 1) direct contact; and - a single crystal third layer (4) composed of a Group III nitride such that the second layer (2) is located in the first layer (1) and the third layer (4) Where the second lattice constant (a 2 ) is greater than the first lattice constant (a 1 ), the lattice of the first layer (1) and the second layer (2) are lattice matched, and The layered semiconductor substrate has a bowness ranging from -50 microns to 50 microns. 根據請求項1所述的分層半導體基材,其中該第一和第二摻雜元素是相同的。 The layered semiconductor substrate of claim 1 wherein the first and second doping elements are the same. 根據請求項2所述的分層半導體基材,其中該第一和第二摻雜元素係共價原子半徑小於矽的元素,且該第一摻雜濃度係高於該第二摻雜濃度。 The layered semiconductor substrate according to claim 2, wherein the first and second doping elements are elements having a covalent atomic radius smaller than 矽, and the first doping concentration is higher than the second doping concentration. 根據請求項3所述的分層半導體基材,其中該第一和第二摻雜元素是硼。 The layered semiconductor substrate of claim 3, wherein the first and second doping elements are boron. 根據請求項3或4所述的分層半導體基材,其中該第二摻雜濃度是零。 The layered semiconductor substrate of claim 3 or 4, wherein the second doping concentration is zero. 根據請求項2所述的分層半導體基材,其中該第一和第二摻 雜元素係共價原子半徑大於矽的元素,且該第二摻雜濃度係高於該第一摻雜濃度。 The layered semiconductor substrate of claim 2, wherein the first and second blends The hetero element is an element having a covalent atomic radius greater than 矽, and the second doping concentration is higher than the first doping concentration. 根據請求項6所述的分層半導體基材,其中該第一和第二摻雜元素是鍺或銻。 The layered semiconductor substrate of claim 6, wherein the first and second doping elements are tantalum or niobium. 根據請求項6或7所述的分層半導體基材,其中該第一摻雜濃度是零。 The layered semiconductor substrate of claim 6 or 7, wherein the first doping concentration is zero. 根據請求項1所述的分層半導體基材,其中該第一摻雜元素的共價原子半徑小於矽,且該第二摻雜元素的共價原子半徑大於矽。 The layered semiconductor substrate according to claim 1, wherein the first doping element has a covalent atomic radius smaller than 矽, and the second doping element has a covalent atomic radius greater than 矽. 根據請求項9所述的分層半導體基材,其中該第一摻雜元素是硼,且該第二摻雜元素是鍺或銻。 The layered semiconductor substrate of claim 9, wherein the first doping element is boron and the second doping element is lanthanum or cerium. 一種用於製造根據請求項1所述的分層半導體基材的方法,包括以下步驟:- 生長一包含至少80%的矽及具該第一摻雜濃度的第一摻雜元素的單晶體,該單晶體具有第一晶格常數;- 從該單晶體上切下至少一個晶圓;- 將該晶圓的厚度減小到該第一厚度,該晶圓構成該第一層(1);- 在該晶圓的一個表面上磊晶沉積該第二層(2);以及- 磊晶沉積由第III族氮化物構成的該第三層(4)。 A method for manufacturing the layered semiconductor substrate according to claim 1, comprising the steps of: growing a single crystal comprising at least 80% germanium and a first doping element having the first doping concentration, The single crystal has a first lattice constant; - at least one wafer is cut from the single crystal; - the thickness of the wafer is reduced to the first thickness, the wafer constitutes the first layer (1); Depositing the second layer (2) on one surface of the wafer; and - epitaxially depositing the third layer (4) composed of a Group III nitride.
TW101123205A 2011-06-30 2012-06-28 Layered semiconductor substrate and method of manufacturing same TWI470831B (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
EP11172250 2011-06-30
EP11191535.1A EP2541589B1 (en) 2011-06-30 2011-12-01 Layered semiconductor substrate and method for manufacturing it

Publications (2)

Publication Number Publication Date
TW201301560A true TW201301560A (en) 2013-01-01
TWI470831B TWI470831B (en) 2015-01-21

Family

ID=46197244

Family Applications (1)

Application Number Title Priority Date Filing Date
TW101123205A TWI470831B (en) 2011-06-30 2012-06-28 Layered semiconductor substrate and method of manufacturing same

Country Status (1)

Country Link
TW (1) TWI470831B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI622456B (en) * 2015-09-09 2018-05-01 東芝記憶體股份有限公司 Semiconductor device manufacturing method

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030132433A1 (en) * 2002-01-15 2003-07-17 Piner Edwin L. Semiconductor structures including a gallium nitride material component and a silicon germanium component
US20080173895A1 (en) * 2007-01-24 2008-07-24 Sharp Laboratories Of America, Inc. Gallium nitride on silicon with a thermal expansion transition buffer layer
EP2104135B1 (en) * 2008-03-20 2013-06-12 Siltronic AG A semiconductor wafer with a heteroepitaxial layer and a method for producing the wafer
KR101102771B1 (en) * 2008-12-24 2012-01-05 매그나칩 반도체 유한회사 Epitaxial Wafers and Manufacturing Method Thereof

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI622456B (en) * 2015-09-09 2018-05-01 東芝記憶體股份有限公司 Semiconductor device manufacturing method
US10068775B2 (en) 2015-09-09 2018-09-04 Toshiba Memory Corporation Method of bonding supporting substrate with device substrate for fabricating semiconductor device

Also Published As

Publication number Publication date
TWI470831B (en) 2015-01-21

Similar Documents

Publication Publication Date Title
US9054017B2 (en) Thick nitride semiconductor structures with interlayer structures and methods of fabricating thick nitride semiconductor structures
US8324005B2 (en) Methods of fabricating nitride semiconductor structures with interlayer structures
CN103348447B (en) Semiconductor wafers comprising gallium nitride with one or more interlayers of silicon nitride therein
JP4335187B2 (en) Nitride semiconductor device manufacturing method
US9343525B2 (en) Aluminum nitride substrate and group-III nitride laminate
TW201732871A (en) Group III nitride structure with enhanced compressive stress grown on tantalum substrate
US10192739B2 (en) Layered semiconductor substrate with reduced bow having a group III nitride layer and method for manufacturing it
TW201813010A (en) Rare earth phosphorus compound for strain management
JP7775708B2 (en) GaN substrate wafer and method for manufacturing GaN substrate wafer
TWI470831B (en) Layered semiconductor substrate and method of manufacturing same
US8466471B2 (en) Nitride semiconductor free-standing substrate and method for making same
JP6224424B2 (en) Method for manufacturing group III nitride semiconductor free-standing substrate
CN103650108B (en) Layered semiconductor substrate and manufacturing method thereof
JP7811397B2 (en) GaN-on-Si epiwafer with strain-decoupled substack
TW201513176A (en) Semiconductor wafer and method of producing semiconductor wafer
KR101474373B1 (en) Semiconductor substrate and method of manufacturing the same
KR101480949B1 (en) Compound semiconductor substrate and method for fabricating the same