The basic essentials of magnetic sector SIMS analysis

1. Usefulness

1.1 Characteristics

The magnetic sector SIMS (for “Secondary Ion Mass Spectrometry”) analysis is used for measuring the atomic composition of solid materials samples. Analysis are performed on plane surfaces. An analysis provides profiles of elements concentrations as a function of depth. Depths ranging from a few nanometers up to a few tens of microns can be reached.

The main feature of this technique is its extreme sensitivity: a magnetic sector spectrometer can detect, in some cases, one atom of minor element among several billions (109) of matrix atoms (i.e. concentrations near from 1013 Although it is particularly sensitive, this technique can also be used for measuring the material matrix composition with an absolute accuracy of the order of a few percent. The depth resolution of the profiles can reach approximately 1nm. However, it must be emphasized that some performances are incompatible with each other (typically, sensitivity and depth resolution). In that case, the operator must find a compromise or, on the contrary, privilege the performance that corresponds to the relevant information.

1.2 Applications

This technique is much used in the field of semiconductors for measuring dopants, impurities or major elements concentrations in a wafer as a function of depth. It is implemented for material research, technological process monitoring, reverse engineering, failures study, …

2. Physical phenomena occurring within the sample

2.1 Ionic bombardment of the sample

During an analysis, the material to be analysed is bombarded by a beam of accelerated and focused ions. This beam and the ions it is made up of are known as “primary” beam and ions. Under typical analysis conditions, the primary beam diameter at sample surface (i.e. the probe size) has a value of a few microns. The impact energy of primary ions lies between 0.5 and 15keV according to the selected conditions. Cs+ and O2+ are the most frequently used primary ions.

Ions impacts on the sample cause the ejection of surface atoms, so that the primary beam gradually digs a crater in the sample. A scanning is applied to the primary beam so as to homogenize the bombardment and to form a flat-bottomed square crater. The crater covers a surface ranging between 20×20μm2 to 500×500μm2. A fraction of ejected atoms is ionized and forms the ions known as “secondary”: these ions are the ones that are exploited to determine the concentrations of elements contained in the sample. An ionic optics enables to limit the detection to the ions that come from a disc located at the center of the square crater.

2.2 Insulating samples

In order to correctly evacuate the electric charge formed on the analysed sample, it is preferable that this one be conductive. However, insulating materials are also routinely analysed thanks to specific processes (surface metallization, electrons beam bombardment).

3. Analysis conditions

3.1 Primary and secondary ions

An analysis is based on secondary ions detection. Generally, the detected ions are monoatomic. However, it can sometimes be advantageous to detect ions made up of one atom of the analysed element plus one or several atoms coming from the matrix or the primary beam (particularly, see the MCs+ mode described below). Positive and negative ions cannot be detected simultaneously. As a consequence, the analysis conditions have to be adjusted on the equipment according to the chosen ions polarity. For monoatomic secondary ions, the settings are adjusted to detect positive ions if electropositive elements are searched (for example Al), while they are adjusted to detect negative ions for elements that have a high electronic affinity (for example F). Electropositive or electronegative elements can sometimes be measured by detecting respectively negative or positive ions, but the sensitivity is then considerably decreased. The primary ions nature has an important effect on the secondary ions polarity. O2+ primary ions are generally used to produce positive secondary ions whereas Cs+ primary ions are used to create negative ions. Cs+ ions are also used to produce MCs+ type ions where M can be a metal or a matrix element. This mode is particularly useful to quantify a matrix composition because the ionization efficiency of MCs+ type species is only little matrix dependant.

3.2 Depth resolution

The depth resolution of a concentration profile depends on:

  • he primary ions impact energy: a high impact energy leads to a “mixing” of superficial atomic layers which turns out to be the ultimate limiting feature to depth resolution,
  • the area the detected ions come from (crater center only or the whole crater): indeed, if secondary ions emitted from the crater edges are detected, they tend to scramble the information on depth.

The best depth resolution is obtained:

  • by decreasing the primary ions impact energy (down to less than 1keV),
  • by spatially filtering the secondary ions origin in order to preserve only the ions that come from the flat crater center region.

Depth resolution has its drawbacks: the detection sensitivity is lowered by these conditions.

3.3 High mass resolution

Sometimes, an interfering signal can be superimposed to the signal of the searched ion: it occurs when a different ionic species with a very close mass is generated and detected. This phenomenon is called “mass interference”. As an example, when one wishes to measure a phosphorus concentration into a silicon wafer, a mass interference occurs between 31P and 30Si+1H. When the masses are not too close to each other, they can be separated, but the high mass resolution settings can considerably diminish the transmitted signal, and thus the sensitivity.

3.4 Atmospheric elements

Atmospheric elements (H, C, N, O) are present in the vacuum chamber residual atmosphere, even when ultra-high vacuum conditions are adopted. When one wishes to measure a very weak concentration of these elements in a sample, the detected signals actually do not correspond merely to the atoms present inside the sample: they correspond also to atmospheric atoms adsorbed on the sample surface. The concentrations below which this phenomenon becomes prevalent are called “detection limits”. When the concentrations in the sample are lower than these ones, the analysis can only provide an overvaluation of these concentrations. The detection limits can be lowered by:

  • a long pumping time before carrying out the analysis,
  • a high sputtering rate (to the detriment of depth resolution).

4. Concentrations and depths calculation

4.1 Minor element concentration (impurity, dopant)

A minor element is an element present in small quantity in the sample, at a concentration typically lower than 1% of the total number of atoms. For such a dilution, the measured signal is considered to be proportional to the element concentration in the material. This concentration is determined thanks to a standard sample which must respect the following conditions:

  • its matrix must have the same composition as the matrix of the analysed sample,
  • the concentration or the dose (at/cm2) of the required element inside the standard must be known. By measuring the signals in the same way in the analysed sample and in the standard sample, one determines the value of a coefficient named “RSF” (for “Relative Sensitivity Factor”) which enables the calculation of a Z element concentration in the analysed sample by using the SIMS formula:

    C(Z)=RSF x S(Z)/S(ref)

    C(Z): Z element concentration.

    S(Z): signal intensity relative to Z element.

    S(ref): signal intensity relative to a reference species made up of one or several matrix atoms.

4.2 Major element concentration

A matrix element concentration is more difficult to determine than a minor element concentration because of the matrix effect. Matrix effect is due to the fact that ionization rates of atoms are highly dependant on their chemical environment within the analysed material. As a consequence, when one measures the signal of an ion made up of one matrix atom, this signal is generally not proportional to the element concentration. The analysis mode known as “MCs+” (see section 3.1) makes it possible to reduce considerably (but not completely) the matrix effect . This mode is the one which is generally used to measure matrix compositions. It is necessary then to use a standard made up of the same alloy than the analysed sample and whose composition is known. If the sample and the standard have sufficiently close compositions, the quantification accuracy can be better than one percent.

4.3 Depth

The most conventional way to establish the depth scale of a concentration profile is to measure the crater depth thanks to a profilometer and then to deduce one or several sputtering rates according to cases. If the crater crosses several layers whose sputtering rates are different, depths can be determined by taking into account the sputtering rate ratios from a layer to another (for example, an InP layer is etched approximately 1.5 time faster than an InGaAs layer).

5. Typical use of magnetic sector SIMS analysis

Measuring the atomic composition of a solid sample consists in identifying the elements which compose it and determining their concentrations (regardless of their chemical bonds). One can distinguish the major elements, which constitute the “matrix”, from the minor elements (typically less than 1% of atoms).

For some applications, minor elements have a major impact on material properties. For example, electronic or optical properties of semiconductor materials are linked to the presence of doping atoms at concentrations that are often lower than 1ppm. Thus, to obtain the desired properties, it is necessary to control doping elements concentrations, but also unwanted impurities concentrations which must generally be much lower than dopant concentrations.

Moreover, some semiconductor devices require a control of the doping depths within a few nanometers only. The recent development of silicon integrated circuits leads to the fabrication of shallow junctions by very weak energy doping elements implantation (at a few tens of electron-volts). Thus, the characterization of such structures requires:

  • detecting very low concentrations of minor elements (<1ppm),
  • obtaining concentrations profiles as a function of depth, with a resolution of a few nanometers for some structures.

Performances reached by the Secondary Ions Mass Spectrometry (SIMS) technique correspond very precisely to this type of request. This method is particularly known for its great sensitivity since it enables the detection of very low concentrations down to 1 atom of minor element among 109 material atoms. In many cases, it also enables to measure the matrix composition when alloys are to be analysed (laser diodes heterostructures, multilayer mirrors …).

Three mass spectrometers types exist: quadrupolar, magnetic sector and time of flight. Magnetic sector SIMS analysis is very common in the field of semiconductor industry due to its high sensitivity. Nevertheless, its performances open its use to other fields such as metallurgy, biology or geology.