本發明者等人為了解決上述課題而進行了各種研究。其結果為，獲得以下見解。 與溶劑之親和性較高之研磨粒容易潤濕，分散性良好。另一方面，與溶劑之親和性較低之研磨粒不易潤濕，而容易引起研磨粒之凝聚。若使用分散性良好之研磨粒，則可藉由抑制研磨時之研磨粒之凝聚，使研磨粒有效率地發揮作用，而提高研磨速度。可認為若分散性良好，則與研磨對象物之接觸點及接觸次數增加，藉此研磨速度變高。可認為於親水性較高之研磨粒中，由於覆蓋表面之水分子而妨礙研磨粒間之接觸，藉此可抑制凝聚，而於親水性較低之研磨粒中，無法完全防止研磨粒間之接觸，故而容易引起凝聚。 本發明係基於該等見解而完成。以下，對本發明之一實施形態之研磨用組合物進行詳細說明。 本發明之一實施形態之研磨用組合物包含二氧化矽研磨粒、pH值調整劑、及水。二氧化矽研磨粒與水之親和性AV為0.51以上，且親和性AV係由下式(1)表示。 AV＝Rsp/TSA (1) 式(1)中，Rsp係由下式(2)表示，TSA為二氧化矽研磨粒之總表面積。 Rsp＝(Rav/Rb)－1 (2) 式(2)中，Rav係於使二氧化矽研磨粒分散之狀態下觀測之NMR弛豫時間之倒數。Rb係於未使二氧化矽研磨粒分散之狀態下觀測之NMR弛豫時間之倒數。 二氧化矽研磨粒可使用在該領域中常用者，例如可使用膠體二氧化矽、發煙二氧化矽等。 二氧化矽研磨粒之含量為研磨用組合物整體之0.5～60質量%。二氧化矽粒子之含量之上限較佳為50質量%，進而較佳為40質量%。二氧化矽粒子之含量之下限較佳為1質量%，進而較佳為5質量%。 本實施形態之研磨用組合物進而包含pH值調整劑。作為用於將研磨用組合物調整至鹼性側之化合物，例如可列舉：氫氧化鉀、氫氧化鈉、碳酸氫鉀、碳酸鉀、碳酸氫鈉、碳酸鈉等鹼性化合物。作為用於將研磨用組合物調整至酸性側之化合物，例如可列舉：鹽酸、硫酸、硝酸、磷酸等酸性化合物。本實施形態之研磨用組合物之pH值較佳為8.5～11.0。 本實施形態之研磨用組合物中，除上述以外，可任意地調配在研磨用組合物之領域中通常已知之調配劑。 本實施形態之研磨用組合物係藉由將二氧化矽研磨粒、pH值調整劑、及其他調配材料適宜地混合並加水而製作。或者，本實施形態之研磨用組合物可藉由將二氧化矽研磨粒、pH值調整劑、及其他調配材料依序混合至水中而製作。作為將該等成分混合之方法，可使用均化器、超音波等在研磨用組合物之技術領域中常用之方法。 以上所說明之研磨用組合物於用水稀釋成為適當之濃度後，被用於藍寶石基板之研磨。 [實施例] 以下，藉由實施例對本發明更具體地進行說明。本發明並不限定於該等實施例。 製作表1所示之實施例1～4及比較例1～3之研磨用組合物。 [表1]
[粒徑之測定方法] 二氧化矽研磨粒之平均粒徑係使用大塚電子股份有限公司製造之粒徑測定系統「ELS-Z2」，藉由動態光散射法而測定。 [中值粒徑之測定方法] 二氧化矽研磨粒之中值粒徑(D50
)係使用美國CPS Instruments公司製造之「碟式離心式高解析粒度分佈測定裝置(DC24000UHR)」，藉由微差式離心沈降法(Differential Centrifugal Sedimentation)測定粒度分佈，並將該粒度分佈轉換為累積分佈，求出該累積分佈中之累積達到50%之粒徑。 [親和性之測定方法] 藉由以下所說明之脈衝NMR，對二氧化矽研磨粒(以下亦簡稱為「粒子」)與分散介質之界面特性進行評價。 接觸或吸附於粒子表面之分散介質分子與分散介質本體中之分散介質分子(未與粒子表面接觸之自由狀態之分散介質分子)對磁場變化之回應不同。一般而言，吸附於粒子表面之液體分子之運動受限制，而本體液中之液體分子可自由移動。其結果為，吸附於粒子表面之液體分子之NMR弛豫時間較本體液中之液體分子之NMR弛豫時間變短。於使粒子分散之液體中觀測之NMR弛豫時間成為反映粒子表面上之液體體積濃度與自由狀態之液體體積濃度的2個弛豫時間之平均值。 再者，於使粒子分散之液體中觀測之弛豫時間常數Rav係由下式表示。 Rav＝PsRs＋PbRb Pb：本體液之體積濃度 Ps：粒子表面積上之液體之體積濃度 Rs：吸收至粒子表面上之吸收層液體分子之弛豫時間常數 Rb：本體液中之液體分子之弛豫時間常數 又，比表面積S與弛豫時間常數Rav之關係係由下式表示。 Rav＝ΨPSLρP(Rs－Rb)＋Rb ΨP：粒子體積濃度 L：吸收至粒子表面上之液體吸收層之厚度 ρP：粒子密度 將於使粒子分散之液體中觀測之弛豫時間之倒數(NMR弛豫時間常數)設為Rav，將於使粒子分散前之液體中觀測之弛豫時間之倒數(NMR弛豫時間常數)設為Rb，計算Rsp＝(Rav/Rb)－1。Rsp係分散介質與粒子表面之親和性之指標，表示若粒子之總表面積相同，則Rsp越大，分散介質與粒子表面之親和性越高。 於本實施形態中，將Rsp除以粒子之總表面積所得之值定義為該研磨用組合物中之粒子與分散介質之「親和性」。 Rav、Rb係使用Xigo nanotools公司製造之脈衝NMR裝置Acorn area測定弛豫時間(具體而言，使二氧化矽研磨粒分散後之NMR弛豫時間、及使二氧化矽研磨粒分散前之NMR弛豫時間)，並求出其倒數。測定條件設為磁場：0.3 T、測定頻率：13 MHz、測定核：1
H NHR、測定方法：CPMG(Carr-Purcell-Meiboom-Gill)脈衝序列法、樣品量：1 ml、溫度：25℃。 粒子之總表面積TSA係根據下式而求出。 TSA＝S×V×ΨP×ρP S：粒子比表面積 V：照射射頻波之部分之NMR管體體積 ΨP：粒子體積濃度 ρP：粒子密度 粒子比表面積S係根據下式而求出。 S＝6/(n×ρP) n：粒子密度 粒子體積濃度ΨP係根據下式而求出。 ΨP＝(λ/100)/[(1－(λ/100))×ρP]×κ λ：粒子之重量%濃度 κ：空白樣品(分散介質)之密度 繼而，使用所製作之實施例1～4及比較例1～3之研磨用組合物，對直徑4英吋之藍寶石基板之c面進行研磨。研磨裝置使用Strasbaugh公司製造之單面研磨機。研磨墊使用胺基甲酸酯之研磨墊。以稀釋後之二氧化矽研磨粒之含量成為19重量%之方式將研磨用組合物稀釋，並以300 ml/分鐘之供給速度進行供給。設為壓盤之旋轉速度為140 rpm、研磨頭之旋轉速度為130 rpm、研磨負荷為500 gf/cm2
，而進行15分鐘之研磨。 [研磨速度之測定方法] 對研磨前後之藍寶石晶圓之質量變化量進行測定，算出藍寶石晶圓之厚度變化量，將每單位時間之厚度變化量設為研磨速度。 [試驗結果之評價] 由於在粒子比表面積S相同時，Rsp越大表示親和性越高，故而圖1所示之圖線之斜率越大，親和性越高。實施例1～4之親和性高於比較例1～3之親和性。如前述之表1所記載般，親和性較高之實施例1～4相較於親和性較低之比較例1～3，可獲得高達約2～4倍之研磨速度。 以上，對本發明之實施形態進行了說明。上述實施形態僅為用於實施本發明之例示。因此，本發明並不限定於上述實施形態，於不脫離其主旨之範圍內，可將上述實施形態適當變化而實施。The present inventors conducted various studies in order to solve the above problems. As a result, the following findings were obtained. The abrasive grains having a high affinity with a solvent are easily wetted and have good dispersibility. On the other hand, the abrasive grains having a low affinity with a solvent are less likely to be wetted, and are liable to cause aggregation of the abrasive grains. When the abrasive grains having good dispersibility are used, the abrasive grains can be efficiently acted upon by suppressing aggregation of the abrasive grains during polishing, thereby increasing the polishing rate. When the dispersibility is good, the contact point and the number of contacts with the object to be polished are increased, whereby the polishing rate is increased. It is considered that in the abrasive grains having a high hydrophilicity, since the water molecules covering the surface interfere with the contact between the abrasive grains, aggregation can be suppressed, and in the abrasive grains having low hydrophilicity, the abrasive grains cannot be completely prevented. Contact, it is easy to cause condensation. The present invention has been completed based on these findings. Hereinafter, the polishing composition according to an embodiment of the present invention will be described in detail. The polishing composition according to an embodiment of the present invention comprises cerium oxide abrasive grains, a pH adjuster, and water. The affinity of the cerium oxide abrasive grains to water is 0.51 or more, and the affinity AV system is represented by the following formula (1). AV = Rsp / TSA (1) In the formula (1), Rsp is represented by the following formula (2), and TSA is the total surface area of the cerium oxide abrasive grains. Rsp = (Rav / Rb) - 1 (2) In the formula (2), Rav is the reciprocal of the NMR relaxation time observed in the state in which the ceria abrasive grains are dispersed. Rb is the reciprocal of the NMR relaxation time observed without dispersing the ceria abrasive grains. The cerium oxide abrasive grains can be used in the field, for example, colloidal cerium oxide, fumed cerium oxide, or the like can be used. The content of the cerium oxide abrasive grains is 0.5 to 60% by mass based on the entire polishing composition. The upper limit of the content of the cerium oxide particles is preferably 50% by mass, and more preferably 40% by mass. The lower limit of the content of the cerium oxide particles is preferably 1% by mass, and more preferably 5% by mass. The polishing composition of the present embodiment further contains a pH adjuster. Examples of the compound for adjusting the polishing composition to the alkaline side include basic compounds such as potassium hydroxide, sodium hydroxide, potassium hydrogencarbonate, potassium carbonate, sodium hydrogencarbonate, and sodium carbonate. Examples of the compound for adjusting the polishing composition to the acidic side include acidic compounds such as hydrochloric acid, sulfuric acid, nitric acid, and phosphoric acid. The pH of the polishing composition of the present embodiment is preferably 8.5 to 11.0. In the polishing composition of the present embodiment, in addition to the above, a compounding agent generally known in the field of polishing compositions can be arbitrarily formulated. The polishing composition of the present embodiment is produced by appropriately mixing cerium oxide abrasive grains, a pH adjuster, and other formulation materials, and adding water. Alternatively, the polishing composition of the present embodiment can be produced by sequentially mixing cerium oxide abrasive grains, a pH adjuster, and other formulation materials in water. As a method of mixing the components, a method commonly used in the technical field of the polishing composition such as a homogenizer or an ultrasonic wave can be used. The polishing composition described above is diluted with water to an appropriate concentration and then used for polishing a sapphire substrate. [Examples] Hereinafter, the present invention will be more specifically described by way of examples. The invention is not limited to the embodiments. The polishing compositions of Examples 1 to 4 and Comparative Examples 1 to 3 shown in Table 1 were produced. [Table 1] [Measurement Method of Particle Size] The average particle diameter of the cerium oxide abrasive grains was measured by a dynamic light scattering method using a particle size measurement system "ELS-Z2" manufactured by Otsuka Electronics Co., Ltd. [Method for Measuring Median Particle Diameter] The median diameter (D 50 ) of the cerium oxide abrasive particles is a "disc centrifugal high-resolution particle size distribution measuring device (DC24000UHR)" manufactured by CPS Instruments, Inc., USA. The particle size distribution was measured by Differential Centrifugal Sedimentation, and the particle size distribution was converted into a cumulative distribution, and the particle size in which the cumulative distribution reached 50% was determined. [Method for Measuring Affinity] The interfacial properties of the ceria abrasive grains (hereinafter also simply referred to as "particles") and the dispersion medium were evaluated by pulse NMR as described below. The dispersing medium molecules that are in contact with or adsorbed on the surface of the particles are different from the dispersing medium molecules in the bulk of the dispersing medium (dispersing medium molecules that are not in contact with the surface of the particles) in response to changes in the magnetic field. In general, the movement of liquid molecules adsorbed on the surface of the particles is restricted, and the liquid molecules in the bulk liquid are free to move. As a result, the NMR relaxation time of the liquid molecules adsorbed on the surface of the particles becomes shorter than the NMR relaxation time of the liquid molecules in the bulk liquid. The NMR relaxation time observed in the liquid in which the particles are dispersed becomes an average of two relaxation times reflecting the liquid volume concentration on the particle surface and the liquid volume concentration in the free state. Further, the relaxation time constant Rav observed in the liquid in which the particles are dispersed is represented by the following formula. Rav=PsRs+PbRb Pb: volume concentration of bulk liquid Ps: volume concentration of liquid on particle surface area Rs: relaxation time constant of liquid molecules absorbed into the surface of the particle Rb: relaxation time constant of liquid molecules in bulk liquid Further, the relationship between the specific surface area S and the relaxation time constant Rav is represented by the following formula. Rav=ΨPSLρP(Rs-Rb)+Rb ΨP: particle volume concentration L: thickness of liquid absorbing layer absorbed onto the surface of the particle ρP: reciprocal of the relaxation time observed in the liquid which disperses the particle (NMR relaxation) The time constant is set to Rav, and the reciprocal of the relaxation time (NMR relaxation time constant) observed in the liquid before the dispersion of the particles is set to Rb, and Rsp = (Rav / Rb) - 1 is calculated. The index of the affinity of the Rsp-based dispersion medium to the surface of the particles means that if the total surface area of the particles is the same, the larger the Rsp, the higher the affinity of the dispersion medium to the surface of the particles. In the present embodiment, the value obtained by dividing Rsp by the total surface area of the particles is defined as the "affinity" between the particles in the polishing composition and the dispersion medium. Rav and Rb were measured using a pulsed NMR apparatus Acorn area manufactured by Xigo nanotools Co., Ltd. to determine the relaxation time (specifically, the NMR relaxation time after dispersing the cerium oxide abrasive particles, and the NMR relaxation before dispersing the cerium oxide abrasive particles Yu time), and find the reciprocal. The measurement conditions were set to magnetic field: 0.3 T, measurement frequency: 13 MHz, measurement core: 1 H NHR, measurement method: CPMG (Carr-Purcell-Meiboom-Gill) pulse sequence method, sample amount: 1 ml, temperature: 25 °C. The total surface area TSA of the particles was determined according to the following formula. TSA=S×V×ΨP×ρP S: particle specific surface area V: NMR tube volume of the portion irradiated with the radio frequency wave ΨP: particle volume concentration ρP: particle density The particle specific surface area S is obtained by the following formula. S=6/(n×ρP) n: particle density The particle volume concentration ΨP is obtained from the following formula. ΨP = (λ / 100) / [(1 - (λ / 100)) × ρP] × κ λ: weight % concentration of the particles κ: density of the blank sample (dispersion medium), and then the produced Example 1 - 4 and the polishing compositions of Comparative Examples 1 to 3 were polished on the c-plane of a 4 inch diameter sapphire substrate. The grinding apparatus used a single-side grinder manufactured by Strasbaugh. The polishing pad uses a urethane polishing pad. The polishing composition was diluted with the content of the diluted cerium oxide abrasive grains to be 19% by weight, and supplied at a supply rate of 300 ml/min. The rotation speed of the platen was 140 rpm, the rotation speed of the polishing head was 130 rpm, and the polishing load was 500 gf/cm 2 , and the polishing was performed for 15 minutes. [Method for Measuring Polishing Rate] The amount of change in the mass of the sapphire wafer before and after the polishing was measured, and the amount of change in thickness of the sapphire wafer was calculated, and the amount of change in thickness per unit time was defined as the polishing rate. [Evaluation of Test Results] When the particle specific surface area S is the same, the larger the Rsp is, the higher the affinity is. Therefore, the slope of the graph shown in Fig. 1 is larger, and the affinity is higher. The affinities of Examples 1 to 4 were higher than those of Comparative Examples 1 to 3. As described in the above Table 1, Examples 1 to 4 having higher affinity were able to obtain a polishing rate of up to about 2 to 4 times as compared with Comparative Examples 1 to 3 having low affinity. The embodiments of the present invention have been described above. The above embodiments are merely examples for carrying out the invention. Therefore, the present invention is not limited to the above-described embodiments, and the above-described embodiments can be appropriately modified and implemented without departing from the spirit and scope of the invention.