Correction Method for Ion Yield Transients in the Near Surface Region of Sims Depth Profiles
- PDF / 171,867 Bytes
- 4 Pages / 420.48 x 639 pts Page_size
- 41 Downloads / 144 Views
NC NC
INTRODUCTION As solid state device features continue to decrease in size, it has become more important to characterize dopant concentrations within the first several hundred angstroms of the surface. Secondary ion mass spectrometry (SIMS) is the technique of choice for dopant depth profiling due to its high sensitivity and good depth resolution. In order to increase the sensitivity of SIMS, electropositive elements (e.g. oxygen) or electronegative elements (e.g. cesium) are used as primary ion species to enhance positive or negative secondary ion yields, respectively. This has the disadvantage, however, of causing secondary ion yields to vary by up to several orders of magnitude over the first few hundred angstroms of a depth profile as the implanted primary ion concentration increases [1,2]. Secondary ion yields stabilize once the primary ion reaches a steady state concentration, which occurs at a depth proportional to the range of the primary ions in the solid. This ion yield transient artifact hinders quantification of dopant concentrations until the primary ion concentration reaches steady state. Until recently, this anomaly was usually ignored since implant depths in the device were well beyond the range of the primary ion used in SIMS. With the trend towards shallower junction depths, however, this is not necessarily the case. One method which has been used as a first order approximation for removing the ion yield transient, when using an oxygen primary ion beam for anlaysis in a Si matrix, is to ratio the profile of the element of interest to the profile of Si [3] or 0 [4]. This assumes that the ion yield transient of the element of interest is identical to the Si or 0 transient. This study presents a novel empirical method based on the use of bulk doped standards for characterizing the ion yield transient of the element of interest. An ion yield correction function is then generated which may be applied to depth profiles from subsequent samples. The method has been used to characterize the ion yield transient of B in Si using an 02 primary ion beam. The resulting ion yield correction function was applied to 4 KeV and 10 KeV B implants. The effect of the primary oxygen ion beam energy on the shape of the ion yield transient was also investigated. EXPERIMENTAL The standard was Si with B bulk-doped to .005-.015 0/cm. The ion yield correction function was applied to two2 different B implants. One was implanted at 4 KeV _t) a dose of 1E15 cm . The other was a 10 KeV implant with a 6.1E14 cm dose. All samples were dipped in HF for 30 sec. then rinsed in deionized water immediately prior to analysis in order to obtain a reproducible surface oxide layer. The experiments were done with a Cameca IMS-3f ion microscope. The instrumental conditions used are summarized in Table I. Depth scales in the depth profile plots were quantified using surface profilometry of the sputter craters.
Mat. Res. Soc. Symp. Proc. Vol. 54.
1986 Materials Research Society
706
B ion yield transient
RT/CM3
B, 0 and Si
RRB. UNIT
Data Loading...
