Journal of Crystal Growth 225 (2001) 249–256

Preparation conditions of chromium doped ZnSe and their

infrared luminescence properties
A. Burgera,*, K. Chattopadhyaya, J.-O. Ndapa, X. Maa, S.H. Morgana,
C.I. Rablaub, C.-H. Suc, S. Fethc, Ralph H. Paged,
Kathleen I. Schaffersd, Stephen A. Payned
a
Department of Physics, Center for Photonic Materials and Devices, Fisk University, Nashville, TN 37208, USA
b
Department of Physics, West Virginia University, Morgantown, WV 26506, USA
c
Microgravity Science and Applications Department, SD 47, Science Directorate, NASA/Marshall Space Flight Center, Huntsville,
AL 35812, USA
d
Lawrence Livermore National Laboratory, L-441, P.O. Box 808, Livermore, CA 94550, USA

Abstract

We report the investigation by photoluminescence lifetime measurements of the near-IR emissions from a series of
chromium-doped ZnSe samples, correlated to their preparation conditions. The samples were polycrystalline or single
crystals prepared by post growth diffusion doping or single crystals doped during growth by the physical vapor
transport method. Room temperature lifetime values between 6 and 8 ms were measured for samples with Cr2+
concentrations from low 1017 to high 1018 cm
3 range. Lifetime data taken down to 78 K was found to be rather
temperature independent, reconfirming previous reports indicating a quantum yield of the corresponding emission of
close to 100% at room temperature. A strong decrease in the room temperature lifetime was found for chromium
concentrations higher than 1019 cm
3. # 2001 Elsevier Science B.V. All rights reserved.

PACS: 42.55.P; 78.55.C; 61.72.V

Keywords: A1. Diffusion; A1. Doping; A2. Growth from vapor; B1. Zinc compounds; B2. Semiconducting II–VI materials;
B3. Solid state lasers

1. Introduction spectral region from about 2 to 4.5 mm. [1–6]

A renewed technological interest in transition
metal (TM)-doped II–VI semiconductors was
triggered in the last years by the potential of these
materials for the development of near-to-mid
infrared tunable solid-state lasers, covering the
*Corresponding author.
E-mail address: aburger@fisk.edu (A. Burger).
Besides basic research applications, such tunable
lasers are of particular importance for environ-
mental monitoring, military countermeasures,
medical applications and remote sensing. [7,8]
Currently available tunable sources in this spectral
range often suffer from drawbacks such as
cryogenic operation (the lead salt diode lasers),
complexity (the broadly tunable optical parametric
oscillators) or narrow tuning range (Tm3+ : YAG,

0022-0248/00/$ - see front matter # 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 0 2 4 8 ( 0 1 ) 0 0 8 4 5 - 4
250 A. Burger et al. / Journal of Crystal Growth 225 (2001) 249–256
Tm3+ : YLF). To date, Co2+ : MgF2 is the only
commercially-available broadly tunable solid state
laser for the near-to-mid infrared region. It has a
wide tuning range (1750–2500 nm) [9,10], however,
it suffers from strong non-radiative decay losses.
In fact, the quantum efficiency of the Co2+
emission in MgF2 has been estimated to be less
than 3%, based on the temperature-dependent
lifetime data [11].
The spectroscopic properties of TM-doped
II–VI compounds have been investigated for more
than four decades now, starting with their study as
efficient luminescence ‘‘killers’’ in visible-emitting
phosphors [12]. The need for a compact (diode
pumped), efficient, room-temperature-operated
broadly tunable mid-IR solid-state laser has
proven a strong driving force for the investigation
of TM-doped chalcogenides as a new class of laser
media. TM ions such as Cr2+, Ni2+, Co2+ and
Fe2+ have been investigated as potential lasing
ions in ZnSe [1–3], ZnS [1,2], ZnTe [1,2], CdSe [4],
and CdMnTe [5] host crystals. Chromium-doped
chalcogenides, and particularly Cr2+ : ZnSe, have
been demonstrated to be extremely attractive for
room-temperature laser operation. So far, room-
temperature pulsed lasing has been demonstrated
for Cr2+ : ZnSe [1,2], Cr2+ : ZnS [1], Cr2+ : CdSe
[4], and Cr2+ : Cd0.85Mn0.15Te [5]. An all-solid-
state Cr2+ : ZnSe pulsed laser using strained-layer
InGaAsP/InP laser diodes for direct laser-diode
pumping has also been demonstrated [6]. More
recently, the first continuous-wave (CW) room
temperature operation of a broadly tunable
Cr2+ : ZnSe laser has also been achieved, with a
tuning range from 2138 to 2760 nm}the broadest
wavelength tuning yet demonstrated from any
room-temperature CW laser [3].
The bonding in ZnSe is partly ionic, with a
fractional ionic character of the bond f ¼ 0:676
[13]. Chromium impurity atoms enter the ZnSe
lattice substitutionally on the Zn site. As such, the
Cr atom contributes two electrons to the bonds
and assumes the lattice-neutral charge-state Cr2+.
In a laser material, the active ion should ideally
match the size and valence state of the host-lattice-
ion it replaces. To a very good approximation, the
Cr2+ ion in ZnSe does meet these requirements.
The ionic radius of the Cr2+ ion in tetrahedral
coordination in ZnSe is rCr2þ ¼ 0:905 A, ( slightly
2+
larger than that of the Zn ion it replaces, which
has r2þ (
Zn ¼ 0:880 A. A typical feature of TM
impurities in semiconductors in general, and of
chromium in II–VI compounds in particular, is the
existence of several possible charge states of a
given TM ion in a given host. Thus, although the
neutral charge state of Cr in II–VI semiconductors
is Cr2+, both Cr1+ (3d5) and Cr3+ (3d3) may exist.
The consequence of this is the presence of an
absorption band below the fundamental edge and
charge transfer processes between the valence
band and the Cr impurity ion energy level, or
between the ion and the conduction band, being
described as follows:
donor-type:
Cr2þ ðd4 Þ þ hn! Cr3þ ðd3 Þ þ eCB !
Cr2þ ðd4 Þ * ! Cr2þ þ hnIR ;
acceptor-type:
Cr2þ ðd4 Þ þ hn0 ! Cr1þ ðd5 Þ þ hnVB !
Cr2þ ðd4 Þ * ! Cr2þ þ hnIR :
In general, dopant concentration levels of the
order of 1018–1019 cm
3 are usually needed to
provide device required characteristics. Compared
with melt grown and post-growth diffusion doped
ZnSe crystals, physical vapor transport (PVT)
grown Cr-doped ZnSe possesses the highest
optical quality and lowest loss [2]. Doping during
the PVT growth process a uniform distribution of
Cr2+ ions, was reported [14]. However, those
samples had low chromium concentrations and
were not usable for device applications. A post-
growth Cr diffusion was developed [15] at Fisk
University. Polycrystalline, twinned and single
crystal wafers of melt grown ZnSe have been
diffusion doped under isothermal conditions, at
temperatures above 8008C. Concentrations in
excess of 1019 Cr+2 ions/cm3 comparable to the
previously reported values for melt grown Cr+2
doping ZnSe material, have been obtained by
diffusion doping and PVT growth. However, the
reported luminescence lifetimes varied consider-
ably from sample to sample.
 A. Burger et al. / Journal of Crystal Growth 225 (2001) 249–256 251

A specific goal of the present study was to 2. Experimental procedure

perform a systematic study that will comparatively
evaluate the luminescent properties of ZnSe
samples doped with chromium during growth by
PVT with samples doped post-growth by thermal
diffusion.
While temperature dependent lifetime measure-
ments for this particular emission have been
performed by the group at LLNL [1,2], their study
was mainly focused on a comparison of several
combinations of TM dopant ion}zinc chalcogen-
ide host. Except for the study hereby presented, no
detailed study has been reported to date that
investigates the effect of Cr2+ concentration on the
luminescence lifetime of Cr2+ : ZnSe. This is an
important issue, since it can provide the crystal
grower with important feed-back regarding the
Cr2+ concentration that should be targeted in the
growth of high quality Cr2+ : ZnSe laser crystals.
The growth-related parameters of the fourteen
Cr-doped ZnSe samples studied are summarized in
Table 1. Two single crystal samples were grown by
the PVT method at NASA/Marshall Space Flight
Center. The doping of these two samples was done
during growth using a vertical (ZnSe 38 V) and
horizontal (ZnSe 37 H) configuration. The source
materials were mixtures of ground ZnSe from
Eagle-Picher Research Laboratory and ground
CrSe from Alfa Inc. The growth temperatures
were in the 1140–11508C temperature range and
furnace translation rates of 1.9–2.2 mm/day were
used. The technical details of the procedure were
published elsewhere [16].
The majority of the samples used a diffusion
doping as a simple and efficient way of incorpor-
ating high concentrations of chromium. These

Table 1
Summary of the growth-related parameters for the Cr-doped ZnSe samples

Sample Crystallinity Growth technique Diffusion temp. Diffusion time Size (l w t) Cr2+Conc. (cm
3) Lifetime
(8C) (days) mm3 (msec)

ZS-Cr02 Large grain Window material 950 5 11 5 1.26 3.9 1019 2.4
polycrystal
ZS-Cr04 Large grain Window material 1000 5 11 6 2.30 3.1 1019 2.0
polycrystal
ZS-Cr05 Single crystal PVT 1000 2 7 3 0.66 9.1 1019 51
ZS-Cr07 Large grain Window material 800 4 12 9 1.08 6.2 1018 7.4
polycrystal
ZS-Cr08 Large grain Window material 800 4 12 9 1.28 7.9 1018 7.8
polycrystal
ZS-Cr09 Large grain Window material 850 4 10 9 1.16 1.2 1019 6.0
polycrystal
ZS-Cr10 Large grain Window material 850 4 12 10 1.19 5.0 1018 7.0
polycrystal
ZS-Cr12 Large grain Window material 805 2 10 10 0.91 4.0 1017 6.5
polycrystal
ZSWN1S Fine grain Window material 950 11 12 12 2.96 2.2 1019 1.9
polycrystal
ZSWN1B Fine grain Window material 950 11 12 12 2.94 2.5 1019 2.2
polycrystal
ZSWN2S Fine grain Window material 915 5 12 12 2.94 1.2 1018 6.7
polycrystal
ZSWN2S Fine grain Window material 915 5 12 12 2.94 2.0 1018 6.8
polycrystal
ZnSe-37H Single crystal PVT-grown NA NA 5 5 1.09 2.0 1020 }
ZnSe-38V Single crystal PVT-grown NA NA 6 6 0.53 7.1 1019 51
252 A. Burger et al. / Journal of Crystal Growth 225 (2001) 249–256
Fig. 1. Experimental set-up for the preparation of Cr2+ : ZnSe
samples by (a) PVT, (b) solid-state diffusion doping.
samples were sealed in quartz ampoules together
with the CrSe dopant source, and placed in an
isothermal oven at temperatures ranging from
800–10008C for 2–5 days. Cr concentration was
adjusted using either the temperature and/or the
diffusion time. The set-up for both the PVT
growth and diffusion doping methods are shown
in Fig. 1.
The Cr2+ concentration, nCr2þ in cm
3, is
related to the peak absorption coefficient (ap in
cm
1) of the 1.8 mm band, shown in Fig. 2, using
the following relation obtained from data reported
by Vallin et al. [17]
ap ¼ 0:144 10
17 nCr2þ : ð1Þ
Fig. 2. Room temperature absorption spectrum of a Cr-doped
ZnSe sample. For reference, the fundamental (band-to-band)
absorption of an undoped ZnSe sample at 440 nm is also
shown (see dotted line).
In this study, the Cr2+ concentration spanned
more than two orders of magnitude, from 2 1017
to 9 1019 cm
3. Photoluminescence lifetime mea-
surements from 78 to 300 K were performed, using
the pulsed output of a picosecond Nd : YAG laser
(Continuum PY61C; pulse width 5100 ps, rep rate
10 Hz). The experimental setup used for these
luminescence decay measurements is shown in
Fig. 3. For direct pumping into the 5T2 ! 5E intra-
center absorption of the Cr2+ ion (absorption
band peak ffi 1:8 mm), the 1.06 mm output of the
Nd : YAG laser was passed through a H2-Cell
Raman Shifter. A 1.9 mm central wavelength band-
pass filter was then used to select the first Stokes
line (1.91 mm) of the Raman shifted pulse for
sample excitation. For pumping into the Cr2+/
Cr1+ charge transfer band, the frequency-doubled
(532 nm) output of the pulsed Nd : YAG laser was
used. A long-pass filter (cut-on wavelength
ffi 2 mm) was used to single out for detection the
(5E ! 5T2) Cr2+ emission. Likewise, a band-pass
filter (central wavelength ffi 1 mm, bandwidthffi
100 nm) was used to single out for detection of
the 1 mm chromium-related emission. An InAs
photodiode (detection range 1.0–3.8 mm) and a
computer-interfaced digitizing oscilloscope were
used for light detection and signal monitoring and
acquisition.
 A. Burger et al. / Journal of Crystal Growth 225 (2001) 249–256 253

Fig. 3. Experimental set-up for lifetime measurements.

3. Results and discussion sorption of an undoped ZnSe sample at 440 nm

3.1. Optical absorption
A representative room temperature optical
absorption spectrum for Cr-doped ZnSe is shown
in Fig. 2. The Cr2+ concentration in this particular
sample was determined to be 5 1018 cm
3. For
reference, the fundamental (band-to-band) ab-
is also shown (see dotted line). Three absorption
bands are observed at 5 K in the Cr2+ : ZnSe
sample in the visible and near-IR region. The
intra-center absorption of interest for laser opera-
tion is the broad band 5T2–5E peaked at about
1800 nm, with a zero-phonon line (ZPL) at
2010 nm. The strong below-bandgap absorp-
tion observed around 500 nm is associated with
254 A. Burger et al. / Journal of Crystal Growth 225 (2001) 249–256

the acceptor-type charge transfer process

Cr2+ ! Cr1++hVB , as discussed earlier in this
paper. The 514.5 nm CW excitation, as well as the
532 nm pulsed excitation are well-suited to pump
into this absorption band. With increasing tem-
perature, the above mentioned absorption bands
get broader. A third absorption peaked near
680 nm could also be detected at low temperature
in samples with a chromium concentration larger
than mid 1018 cm
3. A discussion of the nature of
this absorption was previously published [18].

3.2. Photoluminescence lifetime measurements

The excitation pulse intensity was 100 mJ/pulse

and the beam size at the sample was about 3 mm
diameter. For reference, the oscilloscope trace of
the 1.9 mm pump pulse}as detected by the InAs
photodiode used for luminescence detection in the
actual experiments is also included in Fig. 4a
(dotted line). This trace shows that the response
time of the detector, and particularly the rise time,
although significantly longer than the actual laser
pulse, is not a limiting factor in our experiment.
The Cr2+ luminescence decay in Fig. 4a reveals a
simple exponential decay with a lifetime (1=e) of
about 7.3 ms and a rise (build up) time of 1.8 ms.
Such a simple decay is to be expected in view of the
direct internal excitation mechanism involving
only the ground (5T2) and excited (5E) state of
Cr2+, and was observed throughout the tempera-
ture range investigated. It is notable to mention
that pumping into the Cr2+/Cr1+charge-transfer
band with 532 nm picosecond pulses gives rise to
an emission peaked around 1 mm, with a zero-
phonon line at 934.6 nm (10 700 cm
1). This
emission was found to have a simple exponential
decay, with a strongly temperature dependant
decay-time of 1.054 ms at 78 K and about 8 ms at
280 K. Under the same excitation conditions, the
emission at wavelength longer than 2 mm is found
to be the overlap of two distinct exponentially
decaying components, with extremely different
decay times. In addition to a fast component
(10 ms) believed to be the 5E ! 5T2 radiative
transition, a slow, second component is detected.
This slow component has a lifetime of 670 ms at
78 K, which decreases to 7 ms at 300 K. It is
Fig. 4. The decay of the 5E ! 5T2 Cr2+ intracenter lumines-
cence in Cr2+ : ZnSe sample ZS Cr10 at room temperature after
pulsed excitation with 1.9 mm light on (a) linear scale and (b)
logarithmic scale. The dotted line in 4(a) shows the response
time of the InAs detector.
suggested that both the 1 mm luminescence and
the slow component of the luminescence with
wavelength longer than 2 mm could be due to
forbidden internal transitions of Cr2+. The quan-
tum efficiency of the room-temperature lumines-
cence of Cr2+ : ZnSe in the 2–3 mm spectral region
when pumped into the charge transfer band Cr2+/
Cr1+ is estimated to be about 1–2%.
Table 1 summarizes the luminescence data
obtained in this study. A total of 14 samples have
been investigated at room temperature, with the
Cr2+ concentration spanning the range from
about 2 1017 to 1020 cm
3. In spite of a certain
scattering of the experimentally measured values
for the decay time, data in Fig. 5 clearly shows a
trend. While only a slight concentration depen-
dence can be detected for Cr2+ concentrations
lower than 1019 cm
3, a strong decrease of the
 A. Burger et al. / Journal of Crystal Growth 225 (2001) 249–256
Fig. 5. Luminescence lifetime versus Cr2+ concentration for
samples prepared at: (a) Fisk .and *, (b) LLNL^ and (c)
NASA/MSFC, *. The lifetime measurements were performed
at: (a) Fisk .and * and (b) LLNL *, ^.
luminescence decay time with increasing Cr2+
concentration is apparent above this value. The
lifetime drops from more than 7 ms at nCr2þ ffi
1 1019 cm
3 to less than 3 ms at higher concentra-
tions. Samples with such low lifetime would not be
suitable for laser fabrication. Additional measure-
ments were performed on few selected samples and
it was found that the decay time is temperature
independent (in the 77–300 K range) for low Cr2+
concentration (less than 1019 ions/cm3) samples,
representative of 100% room temperature (RT)
quantum yield efficiency. This was not the case for
samples with Cr2+ concentrations larger than
1019 ions/cm3. In such samples, it is likely that an
enhanced non-radiative recombination occurs and
a room temperature quantum yield efficiency of
only about 35% was estimated.
Important conclusions could be drawn from the
results obtained in this study, revealing significant
differences between the Cr2+ : ZnSe crystal pre-
paration during growth by PVT and by post
growth thermal diffusion. Firstly, doping during
PVT growth, in the form implemented in this
study, offered little control over the Cr concentra-
tion, leading to concentrations that are too large
for the preparation of solid-state lasers. A
255
reservoir containing the dopant source (CrSe) at
a lower, independently controlled temperature
may be necessary for future experiments. One
notable property of the crystals obtained by PVT
was the axial and radial uniformity of the dopant
concentration.
The thermal diffusion doping method is rela-
tively easy to implement and offers good control
over the average chromium concentration (span-
ning over more than two orders of magnitude) by
varying the diffusion time and temperature. The
temperature of the process was at least 2008C
lower than the PVT method, thus minimizing the
probability for contamination. For the time/
temperature conditions investigated by us, the
penetration depth was comparable to or larger
than the sample thickness (2 mm). Obviously, a
concentration profile will always be the outcome
of any diffusion process. Polishing a layer of few
tens of micron off the surface of the chromium
diffused samples may be beneficial in removing the
layer consisting of high dopant concentration (also
low lifetime material), and therefore improving the
device performance.
Acknowledgements
The authors acknowledge the support of NASA
through the Fisk Center for Photonic Materials
and Devices, Grant NCC8-133, NCC8-145 and
NCC5-286 and Dr. Kennedy Reed of LLNL for
his support through the Grant DE-FG03-
94SF20368. The author at West Virginia Univer-
sity acknowledges Dr. N. C. Giles for the financial
support through the National Science Foundation
grant DMR-9806299. Portions of this work were
supported by the National Science Foundation
under Grant No. DMI-9901756.
The authors at NASA/MSFC acknowledge the
support from NASA/office of Biological and
Physical Research.
References
[1] L.D. DeLoach, R.H. Page, G.D. Wilke, S.A. Payne,
W.F. Krupke, IEEE J. Quantum Electron. 32 (1996) 885.
256 A. Burger et al. / Journal of Crystal Growth 225 (2001) 249–256
[2] R.H. Page, K.I. Schaffers, L.D. Deloach, G.D. Wilke,
F.D. Patel, J.B. Tassano Jr., S.A. Payne, W.F. Krupke, K.-
T. Chen, A. Burger, IEEE J. Quantum Electron. 33 (1997)
609.
[3] G.J. Wagner, T.J. Carrig, R.H. Jarman, R.H. Page, K.I.
Schaffers, J.-O. Ndap, X. Ma, A. Burger, OSA TOPS 26
(1999) 427.
[4] J. McKay, K.L. Schepler, G. Catella, OSA TOPS 26 (1999)
420.
[5] J.T. Seo, U. Hommerich, S.B. Trivedi, R.J. Chen,
S. Kutcher, K. Grasza, Conference of Lasers and
Electro-Optics, Vol. 6, OSA 1998 Technical Digest Series,
paper CTuC3.
[6] R.H. Page, J.A. Skidmore, K.I. Schaffers, R.J. Beach, S.A.
Payne, W.F. Krupke, in: C.R. Pollock, W.R. Bosenberg
(Eds.), OSA, Trends In Optics and Photonics, Vol. 10,
OSA, Washington, DC 1997, pp. 208–210.
[7] F.K. Tittel, Appl. Phys. B 67 (1998) 273.
[8] M. Seiter, D. Keller, M.W. Sigrist, Appl. Phys. B 67 (1998)
351.
[9] W. Koechner, Solid-State Laser Engineering, 4th edtion,
Springer, Berlin, 1996, p. 69.
[10] D. Welford, P.F. Moulton, Opt. Lett. 13 (1988) 975.
[11] P.F. Moulton, Proc. IEEE 80 (3) (1992) 348.
[12] H.A. Klassens, J. Electrochem. Soc. 100 (1953) 72.
[13] J.C. Phillips, Rev. Mod. Phys. 42 (1970) 317.
[14] G. Cantwell, W.C. Harsch, H.L. Cotal, B.G. Markey,
S.W.S. Mckeever, J.E. Thomas, J. Appl. Phys. 71 (1992)
2931.
[15] K.T. Chen, T.D. Journigan, S. Lecointe, J. Tong, A. Burger,
L.D. Deloach, K. Schaffers, R.H. Page, S.A. Payne,
Presentation at The American Conference on Crystal
Growth, ACCG-East, Atlantic City, NJ, October 1995.
[16] Ching-Hua Su , Shari Feth, M.P. Volz, R. Matyi, M.A.
George, K. Chattopadhyay, A. Burger, S.L. Lehoczky,
J. Crystal Growth 207 (1999) 35–42.
[17] J.T. Vallin, G.A. Slack, S. Roberts, A.E. Hughes, Phys.
Rev. B 2 (1970) 4313.
[18] C.I. Rablau, J.-O. Ndap, X. Ma, A. Burger, N.C. Giles,
J. Electron. Mat. 28 (6) (1999) 678.