Journal of Crystal Growth 217 (2000) 404}409

E!ect of EDTA on the light scatter in KDP crystal

Xun Sun*, Xinguang Xu, Zhangshou Gao, Youjun Fu, Shenglai Wang, Hong Zeng,
Yiping Li
State Key Laboratory of Crystal Materials, Institute of Crystal Materials, Shandong University, Jinan 250100, People's Republic of China
Received 4 April 2000; accepted 8 May 2000
Communicated by M. Schieber

Abstract

KDP crystals with di!erent concentrations of EDTA in solution were grown. Light scatter was detected with
ultramicroscopic method. It has been proven that light scatter in KDP crystal aggravates with the increasing concentra-
tion of EDTA in the growth solution. The increase with light scatter occurs since EDTA inhibits the growth of KDP
crystal at high concentration, which results in the formation of mother liquid inclusions, which in turn causes light
scatter. The density, size and shape of light scatter are varied within di!erent sectors of the KDP crystal. The mechanism
of the e!ect is discussed.  2000 Elsevier Science B.V. All rights reserved.

Keywords: Potassium dihydrogen phosphate; Ultramicroscopic method; Light scatter; EDTA; Mother liquid inclusion

1. Introduction hold of KDP crystal is related to the density of

scatter centers in the crystal. The higher the scatter
Potassium dihydrogen phosphate (KDP) is a center, the lower is the laser damage threshold of
good nonlinear crystal. The demand for high-qual- the KDP crystal. So it is very important to know
ity large single KDP crystal increases due to its the mechanism of light scatter and manage to re-
application as frequency conversion crystal in iner- duce or eliminate it.
tial con"nement fusion [1]. The growth mechanism The reason for decreasing laser damage thre-
and growth technique have been widely studied shold of KDP crystal has been reported and or-
[2}4], along with the improvement of the crystal ganic materials [5], bacteria [6], and some
growth rate and the ampli"cation of the crystal insoluble impurities [7] are thought to be three
aperture. It is also very important to improve the main factors which cause light scatter.
laser damage threshold of the KDP crystal. Many According to our research results, some anions
researches have proven that the laser damage thres- are also a good reason to cause light scatter in
KDP crystal. Anions such as pyrophosphate and
EDTA can inhibit the growing faces and then cause
* Corresponding author. Tel.: #86-0531-8564265; fax: #86- mother liquid inclusion, which will cause light scat-
0531-8565403. ter in KDP crystal. On the contrary, no scatter was
E-mail address: zsgao@[Link] (X. Sun). detected in KDP crystal grown from Fe>- and

0022-0248/00/$ - see front matter  2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 0 2 4 8 ( 0 0 ) 0 0 5 2 7 - 3
 X. Sun et al. / Journal of Crystal Growth 217 (2000) 404}409 405

Cr>-doped solution. So we think that the main

reason for light scatter in KDP crystals comes from
some organic acids and anions but not from posit-
ively charged ions.
As we know, EDTA is a strong ligand, which
easily forms steady soluble compounds with many
metal ions. When it is doped into the growth solu-
tion, it will increase the growth rate, especially the
growth rate of prism sector of KDP crystal at low
concentrations. When the concentration of EDTA
is high, it will inhibit the growth rate of the whole Fig. 1. Experimental setup.
crystal. In this paper, we report our recent results
on the e!ect of EDTA on the light scatter in KDP
crystal.

2. Crystal growth

KDP raw material is of high quality with the

concentration of Fe>, Cr> being less than 1 ppm
in the growth solution. Growth solution was "l-
tered into a growth vessel of 1000 ml. Z-direction
seeds were used to grow crystals. All the growth
processes were carried out in a water bath by using
a temperature-lowering technique with an auto- Fig. 2. Schematic of KDP crystal.
matically controlled apparatus, and the temper-
ature precision is of $0.023C. Di!erent contents of carry, which allows the samples to move in the
EDTA (AR) were added into the growth solution direction normal to the illumination beam after
after "ltering, which was carried out with a micro- being well polished. The imaging optics is located
membrane "lter of 2 lm. The temperature range of at 903 to the illumination beam. A green line can be
crystal growth is 60}303C. The rotation mode of detected while some dots appear under the micro-
`forward-stop-backwarda was adopted at a speed scope. A schematic representation of the KDP crys-
of 33 rpm. The contents of EDTA in solution are tal is shown in Fig. 2.
300, 1000, 5000 ppm of KDP, respectively. The
solutions were overheated at 803C for 24 h. In order 3.2. Ewect of EDTA of diwerent concentrations on the
to eliminate the hazard in experiments, three sam- light scatter in KDP crystal
ples were grown in the solution of each EDTA
doping concentration. Figs. 3}5 show the light scatter images in the
pyramidal sector of KDP crystals grown with dif-
ferent concentrations of EDTA in solution. As the
3. Results and discussion concentration of EDTA in solution increases from
300 to 5000 ppm, the light scatter in KDP crystal
3.1. Experimental setup aggravates accordingly. When the concentration of
EDTA reaches 5000 ppm in solution, light scatter
The experimental setup consists of a 514.5 nm becomes very heavy, which occupies the entire "eld
100 CW output argon laser (LE;EL 3500), of vision. The size, and shape of the scatter are very
a microscope and a camera to capture the images di!erent which can be explained by the fact that
(Fig. 1). The samples are put onto the adjustable when the concentration of EDTA in solution
406 X. Sun et al. / Journal of Crystal Growth 217 (2000) 404}409

Fig. 3. Light scatter in KDP crystal with di!erent concentra-

Fig. 5. Light scatter in KDP crystal with di!erent concentra-
tions of EDTA, 300 ppm.
tions of EDTA, 5000 ppm.

Fig. 4. Light scatter in KDP crystal with di!erent concentra- Fig. 6. Scatter line in KDP crystal.
tions of EDTA, 1000 ppm.

reaches some high value, it begins to inhibit the From the point of view of crystal structure, the
growth of the KDP crystal and forms mother liquid [1 0 1] pyramidal faces of the KDP crystal are
inclusion which causes light scatter in the crystal. stacked in the mode of two layers of cations over
When an EDTA molecule combines with two layers of anionic ions alternatively. Each layer
[PO ]\, it still has the ability to combine with consists completely of either cations or anions.

other EDTA molecules, this inhibits the movement Each tetrahedron of [PO ] \ forms four H-bonds

of some elementary steps for the size. As a result of with the four nearest [PO ] \ tetrahedron neigh-

subsequent movement of macrosteps, micro inclu- bors, among which three are formed within the
sions of the mother liqor are formed, which are same double layer and the other one is oriented in
viewed as scatters appeared to be bright dots under the [0 0 1] direction and penetrates a positively
the microscope. Not only those dot-shaped scatters charged double layer. In the solution, most of the
can be seen in the KDP crystal but also a line can [H PO ]\ ions exist in oligomerized form. There-
 
be seen in Fig. 6, which originates from a scatter fore, the H-bonds in the oligomer clusters must
`pointa, extends out of the "eld of vision with be broken before they join into the crystal lattices.
a little gap in the middle of the scatter line ending at So the process of crystal growth is a process of
pyramidal}prism sector. the repeating of dehydration of [H PO ]\ and
 
 X. Sun et al. / Journal of Crystal Growth 217 (2000) 404}409 407

forming H-bond with the neighboring PO\ tet-


rahedrons at the growing position.
In addition, comparing with K>, [H PO ]\ is
 
a massive molecule and special orientation must be
taken. So, from the point of view of surface reac-
tion, the growth rate of the pyramidal face is con-
trolled by the stacking rate of [H PO ]\. After the Fig. 7. Structure of EDTA molecule.
 
negatively charged double layer is formed, potassi-
um ions will join into the lattice site rapidly because
of its small volume, light weight and no preferred
orientation as compared with [H PO ]\. So, the
 
pyramidal face is always positively charged in
the growth process. This interfacial character of the
growing crystal has already been veri"ed by means
of high-resolution X-ray di!raction [8]. It is this
interfacial character that makes the cations not
easy to access. So, the e!ect of metal ions on the
growth of pyramidal faces is insigni"cant.
But this character makes the anions, which are
negatively charged, much easier to be adsorbed,
especially for anions that can form high H-bonds.
In addition to the static interaction with the surface Fig. 8. Light scatter in di!erent sectors of KDP crystal, prism
positive charge, they can also form H-bonds in the sector.
direction of [0 0 1] with the [H PO ]\ in the crys-
 
tal lattice beneath the outer positively charged Pyrophosphate has been thought to be another
double layer. reason for the scatter in KDP crystal based on our
The prismatic faces are stacked in a layer mode. recent results. Pyrophosphate mainly causes scatter
Each layer is composed of both cations and anions in the pyramidal sector of KDP crystal but very
in the ratio of 1 : 1. In the usual case, cations are little in the prism sector.
easy to be adsorbed on these faces, especially those Although the scatters caused by these two
which can form strong coordination bonds with (EDTA and pyrophosphate) are di!erent in the
[PO ] \. But when the concentration of anions is KDP crystal, the principle is perhaps the same.

so high that signi"cant anionic adsorption can oc- That is to say, some soluble large molecules, such as
cur through much weaker interaction of H-bond, EDTA and pyrophosphate will cause scatter in
anions will have an inhibiting e!ect on the growth KDP crystal when the concentration is high in the
of prismatic face and even tapering phenomena solution, which inhibits the crystal growth and
happens [9,10]. form mother liquid inclusions, which cause light
EDTA (Fig. 7) will combine with [H PO ]\ by scatter.
 
forming H-band [9]. However, because the H-band
energy is much less than that of ion-band, the e!ect 3.3. The diwerent sectors of KDP crystal with
is not very evident when the concentration of diwerent light scatter
EDTA in the solution is low. As the concentration
of EDTA reaches a certain degree, this inhibition In order to study the e!ect of EDTA on light
becomes very clear. That is to say, EDTA will a!ect scatter in di!erent sectors of KDP crystal,
the pyramidal sector and prism of KDP crystal by 1000 ppm EDTA-doped KDP crystal was chosen
forming H-band with [H PO ]\. It is this dual- as a sample to study the light scatter in pyramidal
 
e!ect action of EDTA that causes the light scatter sector, prism sector and pyramidal}prism sector.
in KDP crystal. The images captured are shown in Figs. 4, 8 and 9.
408 X. Sun et al. / Journal of Crystal Growth 217 (2000) 404}409

Fig. 9. Light scatter in di!erent sectors of KDP crystal, pyr- Fig. 10. Light scatter in di!erent sectors of KDP crystal, light
amidal}prism sector. scatter in pyrimadal}pyramidal sector boundary.

Other samples have a similar behavior. Comparing

Fig. 4 with Fig. 8, it can be seen that the light scatter angle of illumination beam, which will be discussed
increases a little. Those scatter points in the prism in another paper.
not only have larger sizes but also have an extend-
ing distribution compared to those found in the
pyramidal sector, which concentrate on a line along
4. Conclusion
the [0 0 1] direction. The light scatter in pyr-
amidal}prism sector is the heaviest one among the
E A preliminary conclusion can be drawn from our
three for its density, di!erent size and di!erent
studies that EDTA is one reason for light scatter
shape, which varies from spherical to plates or
in KDP crystal. It will inhibit the crystal growth
cylindrical at di!erent observation angles. Al-
in both prism and pyramidal sectors at a high
though these scatters are di!erent in orientation
concentration in the solution to cause mother
and size, they mainly concentrate in a band along
liquid inclusion, which is the main reason for
the [0 0 1] direction. The reason for this is that the
light scatter in KDP crystal.
growth rates of the pyramidal sector and prism are
E There is no scatter in sector boundary.
di!erent here, and the crystal lattice does not match
E Light scatter in pyramidal}prism sector is
very well, which adsorbs impurities in the growth
heavier than those in pyramidal sector and prism
solution easily.
sector.
The camera in the observation system is moved
to give a macroscopic detection. No scatter can be
These are the preliminary conclusions on the
detected in the sector boundary (Fig. 10) compared
e!ect of EDTA on light scatter, and further study is
with other sectors in the KDP crystal. These phe-
still needed.
nomena can be seen in all the samples in our
experiments. As to the formation of sector bound-
ary, which is related to the segregation of impu-
rities, mismatch of the crystal lattice and the Acknowledgements
growth rate, we still do not have adequate evidence
to prove which one is the main reason for its forma- The State High Technology Program "nancially
tion. supports this work for Inertial Con"nement Fusion
We have also found that light scatter in KDP and National Science Foundation (project
solution caused by EDTA varies with intensity and 59823003).
 X. Sun et al. / Journal of Crystal Growth 217 (2000) 404}409 409

References [5] A. Yokotani et al., Appl. Phys. Lett. 48 (16) (1986) 1030.
[6] M.F. Singleton et al., Appl. Phys. Lett. 52 (11) (1988) 857.
[1] J.D. Lindl, Phys. Plasmas 2 (1955) [7] Mingshan Yan et al., J. Synthetic Crystals 15 (1) (1986).
[2] Land, T.L. Matin, S. Potapenko, G.T. Palmore, J.J. De [8] S.A. de Vries, E. Vlieg, P. Goedtkindt, W.J. Huisman,
Yoreo, Nature 399 (1999) 442. M.J. Zwanenburg, Abstracts of ICCG'12, 1998, Jerusalem,
[3] N. Zaitseva, J. Atherton, R. Rozsa, L. Carman, I. Smolsky, Israel, p. 143.
M. Runkel, R. Ryon, L. James, J. Crystal Growth 197 [9] Youjun Fu, Zhangshou Gao et. al., Progress in Crystal
(1999) 911. Growth and Character., in press.
[4] G. Ponomarev, S. Govorkov, W. Ruderman, Abstracts of [10] Youjun Fu, Zhangshou Gao, Jiamin Liu, Yiping Li, Hong
ICCG'12, 1998, Jerusalem, Israel, p. 76. Zeng, Minhua Jiang, J. Crystal Growth 198/199 (1999) 682.