1887
banner image
No data available.
Please log in to see this content.
You have no subscription access to this content.
No metrics data to plot.
The attempt to load metrics for this article has failed.
The attempt to plot a graph for these metrics has failed.
Diffusion of Mg dopant in metal-organic vapor-phase epitaxy grown GaN and AlxGa1−xN
Rent:
Rent this article for
USD
10.1063/1.4792662
/content/aip/journal/jap/113/7/10.1063/1.4792662
http://aip.metastore.ingenta.com/content/aip/journal/jap/113/7/10.1063/1.4792662
View: Figures

Figures

Image of FIG. 1.
FIG. 1.

SIMS depth profiles of Mg for five representative GaN:Mg samples with nominal Mg concentration of 5 × 1019 cm−3 grown at Tg between 925 °C and 1050 °C versus sputter depth scale. The SIMS resolution limit for the Mg distribution as determined from the Al profiles is shown by the bold dashed lines (8 nm/decade and 14 nm/decade) for the samples grown at lowest and highest Tg , respectively. The starting point of Mg doping is indicated by the dashed-dotted line. The grey bar indicates the concentration levels over a depth scale of about 1 μm of diffused Mg atoms at 1100 °C for 1 h into undoped GaN. 12 The arrows indicate back-diffusion b, diffusion d, segregation s, and post growth diffusion p.

Image of FIG. 2.
FIG. 2.

Difference of Mg profiles (Mg (X °C) – Mg (925 °C) where X = 955, 970, 1005, 1035, and 1050, respectively) versus sputter depth scale for five representative GaN:Mg samples, i.e., the Mg profile of the sample grown at Tg  = 925 °C for which Mg diffusion is negligible, was subtracted from profiles of the samples grown at Tg  > 925 °C. The back-diffusion side has positive values while in the range from intersection to surface the concentration of missing Mg atoms is given by negative values. The arrows indicate back-diffusion b, diffusion d, segregation s, and post growth diffusion p.

Image of FIG. 3.
FIG. 3.

Integrated contributions of the difference of Mg profiles above and below the intersection (see, Fig. 2 ) versus Tg . The result has the dimension of a sheet concentration or dose. The positive part shows the sheet concentration of back-diffused atoms (circles); the negative part the sum of missing back-diffused and segregated atoms (squares). Polynomial fits (full lines) are just viewing aids. The difference of both contributions represented by the polynomial fits yields the segregated atoms indicated by the dashed line. The inset shows the Mg distribution of a GaN:Mg layer, indicated by the shaded bar, which was placed between undoped layers.

Image of FIG. 4.
FIG. 4.

Two SIMS depth profiles of Mg from 110 nm Al0.1Ga0.9N:Mg layers without GaN cap with nominal Mg concentration of 5 × 1019 cm−3, grown at Tg of 925 °C and 1050 °C, versus sputter depth scale shown on the left side (Al profiles are omitted). Subtraction of the low from the high temperature distribution yields the resulting difference profile versus sputter depth shown on the right side.

Image of FIG. 5.
FIG. 5.

SIMS depth profiles of Mg for four representative Al0.1Ga0.9N:Mg samples with nominal Mg concentration of 5 × 1019 cm−3, grown at Tg between 925 °C and 1035 °C, versus sputter depth scale. The resolution limit of 8 nm/decade is shown by bold dashed line for the sample grown at highest Tg . The starting point of Mg doping is indicated by the dash-dotted line. Since the flux of the Al precursor is considerably higher, the Al signal (full line) is divided by 500.

Image of FIG. 6.
FIG. 6.

Difference of Mg profiles (Mg (X °C) – Mg (925 °C) where X = 955, 1005, and 1035, respectively) versus sputter depth scale for three representative Al0.1Ga0.9N:Mg samples, i.e., the Mg profile of the sample grown at Tg  = 925 °C for which Mg diffusion is negligible, was subtracted from profiles of the samples grown at Tg  > 925 °C. The back-diffusion side has positive values while in the range from intersection to surface the concentration signal is strongly influenced by the GaN/Al0.1Ga0.9N interface (indicated by the bold dashed line).

Image of FIG. 7.
FIG. 7.

Integrated contributions from back-diffusion of difference of Mg profiles for GaN:Mg (circles) and Al0.1Ga0.9N:Mg (squares) versus Tg . The result has the dimension of a sheet concentration or dose. Polynomial fits (full lines) are just a viewing aid.

Image of FIG. 8.
FIG. 8.

SIMS depth profiles of Mg and Al marker for two GaN:Mg sampleswith nominal Mg concentration of 5 × 1019 cm−3 annealed at Tanneal  = 990 °C and Tanneal  = 1050 °C versus sputter depth scale. The high temperature distribution is adjusted to the Al marker and, therefore, shifted by 46 nm. The flux of the Al precursor is considerably higher and the signal is divided by 500.

Image of FIG. 9.
FIG. 9.

SIMS depth profiles of Mg and Al for two Al0.1Ga0.9N:Mg sampleswith nominal Mg concentration of 5 × 1019 cm−3 annealed at Tanneal  = 970 °C and Tanneal  = 1050 °C versus sputter depth scale. The high temperature annealed distribution is adjusted to the Al marker and therefore shifted by 5 nm. The flux of the Al precursor is considerably higher and the signal is divided by 500.

Image of FIG. 10.
FIG. 10.

Etch rate of the annealed GaN:Mg (circles) and Al0.1Ga0.9N:Mg (squares) samples versus annealing temperature. Linear fits (full lines) are just a viewing aid.

Image of FIG. 11.
FIG. 11.

SIMS depth profiles of Mg for GaN:Mg samples from Fig. 1 at Tg  = 925 °C and 1050 °C (left side) and Al0.1Ga0.9N:Mg from Fig. 5 (right side) at Tg  = 925 °C (circles) and 1035 °C (squares) versus the square of the diffusion width. Dotted lines are fits assuming a Gaussian distribution.

Image of FIG. 12.
FIG. 12.

Diffusion depth profiles of Mg for five representative GaN:Mg samples grown at Tg between 925 °C and 1050 °C plotted relative to the intersection at 98 nm (see, Fig. 1 ), fitted by erfc (lines) assuming a concentration dependent diffusion coefficient.

Image of FIG. 13.
FIG. 13.

Diffusion depth profiles of Mg for four representative Al0.1Ga0.9N:Mg samples grown at Tg between 925 °C and 1035 °C plotted relative to the intersection at 71 nm (see Fig. 5 , fitted by erfc (lines) assuming a concentration dependent diffusion coefficient.

Image of FIG. 14.
FIG. 14.

Concentration dependence of the diffusion coefficient of GaN, (left side) and Al0.1Ga0.9N (right side) for different Tg versus Mg concentration.

Image of FIG. 15.
FIG. 15.

Exponent γγ 925 of the concentration dependence of the diffusion coefficient for GaN (circles) and Al0.1Ga0.9N (squares), as determined from the fitted erfc, versus Tg . Lines are just viewing aids.

Image of FIG. 16.
FIG. 16.

Arrhenius plot of Ds Ds 925 at Ns  = 1.25 × 1019 cm−3 versus 1/kTg . The slopes for both, GaN (circles) and Al0.1Ga0.9N (squares), yield an activation energy E A = 5.0 eV and 5.2 eV, respectively.

Image of FIG. 17.
FIG. 17.

SIMS depth profile of Mg in GaN versus sputter depth scale is shown on a linear scale on the left side. The transient of the Mg doping is fitted by a linear function given by bold line down to about 20% of the concentration and extrapolated by a dashed line. The beginning of doping as determined by the growth time is indicated by the dashed-dotted line. Beyond the dashed-dotted line, the profile proceeds, which essentially is related to broadening of the SIMS signal indicated by the dashed line. This resolution limit is shown in more clarity on the right side where the profiles of Mg and Al marker (thin lines) are drawn on a semi logarithmic scale. The flux of the Al precursor is considerably higher and the signal is divided by 500. The bold dashed lines mark the resolution limit for both Al and Mg signals, which is 8 nm/decade. The other lines have the same denotation as on the left side.

Image of FIG. 18.
FIG. 18.

SIMS depth profiles of Mg in GaN (full lines) doped at levels of 3 × 1019 cm−3, 6 × 1019 cm−3, and 1 × 1020 cm−3 versus sputter depth scale are shown on a linear scale on the left side. The linear fits are given by bold lines and extrapolated by dashed lines. The beginning of doping as determined by the growth time is indicated by the dashed-dotted line. Added is the In distribution of a triple QW, which is used as a marker (full line). The resolution limit is shown on the right side where the profiles of Mg and In markers (thin lines) are drawn on a semi logarithmic scale. The flux of the In precursor is considerably higher, and the signal is divided by 100. The bold dashed lines mark the resolution limit for the Mg signals, which is 8 nm/decade. The other lines have the same denotation as on the left side.

Image of FIG. 19.
FIG. 19.

Delay-length of the distributions of the samples described in Figs. 17 and 18 versus N Mg. The results related to the In marker are given by full squares and to the Al marker by an open square. The results related to the In marker were used for a linear fit (dashed line).

Image of FIG. 20.
FIG. 20.

SIMS depth profile of Mg in Al0.1Ga0.9N:Mg versus sputter depth scale is shown on a linear scale on the left side. The linear fit is given by a bold line and extrapolated by a dashed line. The beginning of doping as determined by the growth time is indicated by the dashed-dotted line. The resolution limit is shown on the right side where the profiles of Mg and Al (thin lines) are drawn on a semi logarithmic scale. The flux of the Al precursor is considerably higher, and the signal is divided by 500. The bold dashed lines mark the resolution limit for both Al and Mg signals, which is 8 nm/decade. The other lines have the same denotation as on the left side.

Loading

Article metrics loading...

/content/aip/journal/jap/113/7/10.1063/1.4792662
2013-02-21
2014-04-24
Loading

Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: Diffusion of Mg dopant in metal-organic vapor-phase epitaxy grown GaN and AlxGa1−xN
http://aip.metastore.ingenta.com/content/aip/journal/jap/113/7/10.1063/1.4792662
10.1063/1.4792662
SEARCH_EXPAND_ITEM