Raman micro-spectrometry and its applications to the identification of inclusions in natural rubies
显微拉曼光谱法及其在天然红宝石中内含物鉴定中的应用

Laboratoire SPMS, École Centrale de Paris, Grande voie des Vignes, 92295 Châtenay-Malabry Cedex, France
SPMS 实验室， École Centrale de Paris， Grande voie des Vignes， 92295 Châtenay-Malabry Cedex， France

Raman spectrometry is an analytical method based on the interaction of light and molecular and/or lattice motions (vibration, rotation and translation motions). This technique presents several advantages : it is non-destructive, quick, precise, very easy to use and requires no sampling. Distinctive Raman imaging is now also available.
拉曼光谱法是一种基于光与分子和/或晶格运动（振动、旋转和平移运动）相互作用的分析方法。这种技术具有以下几个优点： 它是非破坏性的、快速的、精确的、非常容易使用并且不需要采样。现在还提供独特的拉曼成像功能。

Concerning the study of minerals and gems, considerable progress was made once the coupling of a Raman spectrometer with an optical microscope became possible. Historically, this system was developed at the same time by French and American researchers [1-5]. A dedicated microscope, through which the exciting laser beam and the backscattered signal both pass is used to examine a sample and to send the Raman signals to the spectrometer. Such a device allows the exciting beam to be focussed on a small volume of the sample and the Raman spectrum of this volume to be collected. It is thus possible to study microcrystals (Fig. 1). Technical improvements have been added to the basic technique to increase its performances. The confocal mount is an important one which consists in placing a hole in the image focal plane of the microscope, in order to suppress from the signal all beams not from the focal plane. The confocal mount allows a good vertical resolution to be achieved whereas this would be impossible with a standard mount [6]. It is consequently possible to measure two- or three-dimensional maps of samples, which is very interesting for the study of inhomogeneous samples, of di- or triphasic inclusions, and for numerous other applications [7-8].
在矿物和宝石的研究方面，一旦拉曼光谱仪与光学显微镜的耦合成为可能，就取得了相当大的进步。从历史上看，该系统是由法国和美国的研究人员同时开发的 [1-5]。专用显微镜，激发激光束和背向散射信号都通过该显微镜，用于检查样品并将拉曼信号发送到光谱仪。这种设备允许激发光束聚焦在一小块样品上，并收集该体积的拉曼光谱。因此可以研究微晶（图 1）。基本技术中添加了技术改进以提高其性能。共聚焦安装座是一个重要的安装座，包括在显微镜的图像焦平面上放置一个孔，以便从信号中抑制所有不是来自焦平面的光束。共聚焦安装座可以实现良好的垂直分辨率，而标准安装座则无法实现 [6]。因此，可以测量样品的二维或三维图，这对于研究不均匀样品、双相或三相夹杂物以及许多其他应用非常有趣[7-8]。

The identification of inclusions in natural stones provides interesting information on their genesis such as their conditions of formation, secondary reactions, or unnatural treatments. In our Laboratory, Raman micro-spectrometry has been used for analysing natural ruby samples. Our goal was to investigate the possibility of distinguishing several ruby deposits by examining the chemical nature and the frequency of inclusions in a given natural gem.
天然宝石中内含物的鉴定提供了有关其成因的有趣信息，例如它们的形成条件、次级反应或不自然处理。在我们的实验室中，拉曼显微光谱法已被用于分析天然红宝石样品。我们的目标是通过检查特定天然宝石的化学性质和内含物的频率来研究区分几种红宝石矿床的可能性。

The first step in such a study consists in building a Raman database of various minerals that might be found as inclusions in natural rubies or other stones. We built our own database [9], since such spectra were rare, incomplete or not available. A database already existed, but it was not complete enough for our needs [10]. At present, it is possible to find interesting updated regularly data at the following Internet address:
此类研究的第一步包括建立一个拉曼数据库，其中包含可能在天然红宝石或其他宝石中发现的各种矿物。我们建立了自己的数据库 [9]，因为这样的光谱很少见、不完整或不可用。数据库已经存在，但它还不够完整，无法满足我们的需求 [10]。目前，可以在以下 Internet 地址找到有趣的定期更新数据：
http://minerals.gps.caltech.edu/files/raman/Caltech_data/index.htm
The analysis of inclusions can be carried out easily, quickly, and with a very simple sampling operation. Thick blades are the best samples, but it is also possible to analyse cut gemstones. Rough surfaces are not recommended as it is difficult to examine the inside of the sample and to focus the laser beam accurately on the inclusion. For our study, we used 1 mm thick polished ruby samples. They were studied with a LABRAM I (DILOR, ISA) using a SpectraPhysics Ar+ laser as the source. A $\times 50$$\times 50$xx50\times 50 microscope objective with a long working distance ( 8 mm ) allowed the examina-
夹杂物的分析可以轻松、快速地进行，并且采样作非常简单。厚刀片是最好的样品，但也可以分析切磨的宝石。不建议使用粗糙的表面，因为很难检查样品内部并将激光束准确地聚焦在夹杂物上。在我们的研究中，我们使用了 1 mm 厚的抛光红宝石样品。它们是用 LABRAM I （DILOR， ISA） 和 SpectraPhysics Ar+ 激光器作为光源进行研究的。具有长工作距离 （ 8 mm ） 的 $\times 50$$\times 50$xx50\times 50 显微镜物镜允许检查-

Figure 1. Principle scheme of Raman micro-spectrometry applied to the analysis of inclusions.
图 1.拉曼显微光谱法应用于夹杂物分析的原理方案。
tion of inclusions deep inside the host crystal. Inclusions ranging from some microns to hundreds of microns in size were studied.
主晶体深处的内含物。研究了大小从几微米到几百微米不等的夹杂物。

There are several problems that can arise: when studying inclusions, the host matrix can partly absorb the laser beam, depending on its wavelength, consequently making the measurement of deep inclusions difficult. For rubies where the 514.5 nm wavelength of an $\mathrm{Ar}+$$\mathrm{Ar}+$Ar+\mathrm{Ar}+ laser is used, inclusions located beyond 800 mm in depth are difficult to study. Turning the sample upside down solved the problem in our case. Another difficulty which can frequently be encountered is fluorescence. In the case of ruby, there is very strong fluorescence due to the well-known R-lines situated in the red region and used in the ruby laser (Fig. 2). Consequently, when the 632.8 nm line of a He -Ne laser is used as
可能会出现几个问题：在研究夹杂物时，主基体可以部分吸收激光束，具体取决于其波长，因此很难测量深层夹杂物。对于使用 514.5 nm 波长激光的 $\mathrm{Ar}+$$\mathrm{Ar}+$Ar+\mathrm{Ar}+ 红宝石，深度超过 800 mm 的内含物很难研究。将样品倒置解决了我们案例中的问题。另一个经常遇到的困难是荧光。对于红宝石，由于红宝石区域中使用了众所周知的 R 线，因此会产生非常强的荧光（图 2）。因此，当 He -Ne 激光器的 632.8 nm 线用作

Figure 2. Raman spectra of a ruby with two different exciting wavelengths. The influence of fluorescence and its intensity compared to that of the Raman scattering can clearly be seen. Left part: although a fluorescence background is still present, the Raman peaks of corundum can be observed. Right part: Strong R-lines of ruby are observed with a 632.8 nm He-Ne excitation, completely hiding the Raman peaks of corundum.
图 2.具有两种不同激发波长的红宝石的拉曼光谱。可以清楚地看到荧光及其强度与拉曼散射相比的影响。左半部分：虽然荧光背景仍然存在，但可以观察到刚玉的拉曼峰。右图：在 632.8 nm He-Ne 激发下观察到红宝石的强 R 线，完全隐藏了刚玉的拉曼峰。
excitation radiation only a very strong fluorescence emission is observed, completely hiding the very weak Raman lines of corundum. With a 514.5 nm excitation of an Ar+ laser, a strong background is still observed on the Raman spectrum, and increases for high wavenumbers (they are closer to the R-lines). A solution can be to work with a 488.0 nm wavelength, in order to be even further away from the R-lines. The spectra are less noisy, and the measurement is more sensitive. The 488.0 nm wavelength presents another advantage as it is located almost at the minimum absorption of ruby. Consequently, the fluorescence excited by this wavelength is weaker than that excited with 514.5 nm wavelength. However, in all cases, it is impossible to measure the peak of ${\mathrm{H}}_2\mathrm{O}$${\mathrm{H}}_2\mathrm{O}$H_(2)O\mathrm{H}_{2} \mathrm{O}, located around $3400{\mathrm{cm}}^{-1}$$3400{\mathrm{cm}}^{-1}$3400cm^(-1)3400 \mathrm{~cm}^{-1}, too close to the R-lines. It must also be noted that the Raman peaks are about a hundred thousand times weaker than the ruby fluorescence. In the case of natural rubies, as the fluorescence background depends on impurities and/or trace elements, the variation of the background compared to the Raman peaks intensities can in some cases indicate the origin of the stone (Fig. 3). This is the reason why all our samples were examined with the 514.5 nm wavelength.
激发辐射仅观察到非常强的荧光发射，完全隐藏了刚玉非常微弱的拉曼线。在 Ar+ 激光器的 514.5 nm 激发下，在拉曼光谱上仍然观察到强背景，并且随着高波数的增加而增加（它们更接近 R 线）。一种解决方案是使用 488.0 nm 波长，以便远离 R 线。光谱的噪声更小，测量更灵敏。488.0 nm 波长具有另一个优势，因为它几乎位于红宝石的最小吸收处。因此，该波长激发的荧光比 514.5 nm 波长激发的荧光弱。然而，在所有情况下，都不可能测量位于 $3400{\mathrm{cm}}^{-1}$$3400{\mathrm{cm}}^{-1}$3400cm^(-1)3400 \mathrm{~cm}^{-1} 周围的 ${\mathrm{H}}_2\mathrm{O}$${\mathrm{H}}_2\mathrm{O}$H_(2)O\mathrm{H}_{2} \mathrm{O} 的峰值，该峰离 R 线太近。还必须注意的是，拉曼峰比红宝石荧光弱约 10 万倍。对于天然红宝石，由于荧光背景取决于杂质和/或微量元素，因此与拉曼峰强度相比，背景的变化在某些情况下可以表明宝石的来源（图 3）。这就是为什么我们所有的样品都使用 514.5 nm 波长进行检查的原因。

There are some other difficulties to overcome. The Raman spectrum of a single crystal is strongly dependent on its orientation due to the polarisation effect. This is true for the host matrix and the inclusions. As the last ones can be randomly placed in the host crystal, the recorded Raman spectra can drastically vary in shape and intensity for inclusions of the same nature but differently oriented. It is sometimes difficult to identify an inclusion just by comparison with the spectra of the database. It is necessary to turn the sample and study it in different positions in order to change the orientation of both the host matrix and the inclusion. A correct spectrum can be obtained by averaging these different spectra.
还有一些其他困难需要克服。由于极化效应，单晶的拉曼光谱在很大程度上取决于其取向。对于主体矩阵和内含物来说，情况确实如此。由于最后的拉曼光谱可以随机放置在主晶体中，因此对于性质相同但方向不同的内含物，记录的拉曼光谱在形状和强度上可能会发生巨大变化。有时仅通过与数据库的光谱进行比较很难识别内含物。有必要转动样品并在不同的位置进行研究，以改变主基质和夹杂物的方向。通过平均这些不同的光谱，可以获得正确的光谱。

The last problem is that inclusions of different minerals can have the same or very close Raman spectra. For example, this is the case for the different forms of ${\mathrm{CaCO}}_3$${\mathrm{CaCO}}_3$CaCO_(3)\mathrm{CaCO}_{3} : calcite, aragonite and other carbonates have very similar spectra [11], making them difficult to distinguish.
最后一个问题是，不同矿物的包裹体可以具有相同或非常接近的拉曼光谱。例如，不同形式的 ${\mathrm{CaCO}}_3$${\mathrm{CaCO}}_3$CaCO_(3)\mathrm{CaCO}_{3} 方解石、文石和其他碳酸盐具有非常相似的光谱 [11]，因此很难区分。

The present paper shows the results recently obtained when examining 164 and 100 ruby samples originating from the Luc Yen and Quy Chau mines respectively, both recently discovered and exploited in Vietnam. Figures 2-5 and photo 1 show some interesting features of inclusions found in these rubies. In addition, we have detected complex inclusions containing both gaseous and liquid ${\mathrm{CO}}_2$${\mathrm{CO}}_2$CO_(2)\mathrm{CO}_{2} (Fig. 4, photo. 2), and another very interesting type of inclusion, which seems to contain sulphured compounds. These last inclusions present a spectrum which evolves with time: the peaks of a certain type of sulphur (probably with ${\mathrm{S}}_3{}^{2-}$${\mathrm{S}}_3{}^{2-}$S_(3)^(2-)\mathrm{S}_{3}{ }^{2-} ions)
本文展示了最近在检查分别产自陆克彦和归洲矿的 164 和 100 个红宝石样品时获得的结果，这两颗样品都是最近在越南发现和开采的。图 2-5 和照片 1 显示了这些红宝石中发现的内含物的一些有趣特征。此外，我们还检测到了同时包含气态和液 ${\mathrm{CO}}_2$${\mathrm{CO}}_2$CO_(2)\mathrm{CO}_{2} 态的复杂内含物（图 4，照片 2），以及另一种非常有趣的内含物类型，它似乎含有含硫化合物。这些最后的内含物呈现出一个随时间演变的光谱：某种类型的硫（可能带有 ${\mathrm{S}}_3{}^{2-}$${\mathrm{S}}_3{}^{2-}$S_(3)^(2-)\mathrm{S}_{3}{ }^{2-} 离子）的峰

Figure 3. General shapes of the Raman spectra of African, Vietnamese and Burmese rubies.
图 3.非洲、越南和缅甸红宝石的拉曼光谱的一般形状。

Figure 4. Raman spectra of titanite and of zircon inclusions in a Luc Yen ruby.
图 4.Luc Yen 红宝石中钛矿和锆石内含物的拉曼光谱。

Photo 1. Bottle shaped diphasic inclusion: liquid and gaseous ${\mathrm{CO}}_2$${\mathrm{CO}}_2$CO_(2)\mathrm{CO}_{2}.
照片 1.瓶形双相内含物：液态和气态 ${\mathrm{CO}}_2$${\mathrm{CO}}_2$CO_(2)\mathrm{CO}_{2} 。