Complex Formation of Trimethylaluminum and Trimethylgallium with Ammonia

Subscriber access provided by JIANGSU UNIV LIB

The Journal of Physical Chemistry A is published by the American Chemical Society.

1155 Sixteenth Street N.W., Washington, DC 20036

Article

Complex Formation of Trimethylaluminum and Trimethylgallium with Ammonia: Evidence for a Hydrogen-Bonded Adduct

George T. Wang, and J. Randall Creighton

J. Phys. Chem. A, 2006, 110 (3), 1094-1099 ? DOI: 10.1021/jp054133o

Downloaded from https://www.wendangku.net/doc/c210745916.html, on December 3, 2008

More About This Article

Additional resources and features associated with this article are available within the HTML version:?Supporting Information

?Links to the 2 articles that cite this article, as of the time of this article download

?Access to high resolution figures

?Links to articles and content related to this article

?Copyright permission to reproduce figures and/or text from this article

Complex Formation of Trimethylaluminum and Trimethylgallium with Ammonia:Evidence for a Hydrogen-Bonded Adduct

George T.Wang*and J.Randall Creighton

Sandia National Laboratories,Sandia National Laboratories,P.O.Box 5800,MS-0601,Albuquerque,New Mexico 87185

Recei V ed:July 26,2005;In Final Form:No V ember 16,2005

We have investigated the formation of gas-phase adducts of trimethylaluminum and trimethylgallium with ammonia using room-temperature Fourier transform infrared experiments and density functional theory calculations.Our results indicate for the first time that,at higher partial pressures,a product distinct from the well-known (CH 3)3M:NH 3adduct grows in for both M )Al and M )https://www.wendangku.net/doc/c210745916.html,parison of the experimental and calculated IR spectra,along with calculations of the energetics,indicates that this second product is the result of hydrogen bonding of a second NH 3molecule to the (CH 3)3M:NH 3adduct and can be written as (CH 3)3M:NH 3???NH 3.The binding energy of this hydrogen-bonded adduct is calculated to be 26.8kcal/mol for M )Al and 18.4kcal/mol for M )Ga and is lower in energy (more stable)relative to the 1:1(CH 3)3M:NH 3adduct by 7.2kcal/mol for M )Al and 6.6kcal/mol for M )Ga.In contrast,an alternative complex involving the formation of two separate M -N donor -acceptor bonds,which is written as H 3N:(CH 3)3M:NH 3,is calculated to be lower in energy relative to (CH 3)3M:NH 3by only 0.1kcal/mol for M )Al and 0.2kcal/mol for M )Ga and is not observed experimentally.These results show that hydrogen bonding plays an important role in the interaction of ammonia with metal organic precursors involving Al,Ga,and In,under typical metal organic chemical vapor deposition AlGaInN growth conditions.

Introduction

The group III nitrides,including GaN and AlGaInN alloys,are an important class of semiconductors currently used in a number of optoelectronic applications,including light emitting diodes and lasers.Currently,metal organic chemical vapor deposition (MOCVD),typically employing the precursors trimethylaluminum (TMAl),trimethylgallium (TMGa),trim-ethylindium (TMIn),and ammonia (NH 3),is the dominant technique used to deposit device-quality III-nitride materials.Unfortunately,gas-phase reactions between these organometallic precursors and ammonia can have parasitic effects that make control and reproducibility of the growth process difficult.1-9Upon mixing with ammonia,it is widely known that TMAl,TMGa,and TMIn will undergo the reversible formation of a donor -acceptor complex via attack of the lone-pair electrons of NH 3at the electron-deficient metal atom of the organometallic precursor,as shown in

Formation of the (CH 3)3M:NH 3adduct is of interest because it represents the first step from which further parasitic chemical reactions may take place and thus has been widely studied.4,9-24Creighton et al.previously showed that these parasitic reactions can even lead to the formation of gas-phase nanoparticles.25,26In this paper,we have taken a combined experimental and theoretical approach to elucidate the nature of the adduct formation between ammonia and trimethylaluminum or trim-ethylgallium.We report here the gas-phase infrared spectra of the complexes of TMAl and TMGa with NH 3at room temperature,along with density functional theory quantum

chemical calculations of the energetics and infrared frequencies.We have also isolated experimentally and identified,for the first time,the formation of an additional,distinct “2:1”adduct involving a single TMAl or TMGa molecule and two NH 3molecules,which increases in relative concentration as the partial pressures are increased.We can thus expand reaction 1as

Previous theoretical work has considered the complexation of

a second NH 3molecule with the metal atom of the (CH 3)3M:NH 3adduct,forming a doubly coordinated H 3N:(CH 3)3M:NH 3adduct with two M -N bonds.19,21A reasonable assumption thus may be to assign the identity of the experimentally observed 2:1adduct as this doubly coordinated H 3N:(CH 3)3M:NH 3adduct.However,we present here for the first time experimental and theoretical results that indicate that this adduct instead involves the hydrogen bonding of a second NH 3molecule to the (CH 3)3M:NH 3adduct,rather than coordination to the metal atom.

Theoretical and Experimental Methods

Main-group chemistry has been widely studied using density functional theory (DFT)methods.DFT calculations using the B3LYP hybrid functional 27were employed in this study to examine the chemistry between TMAl and TMGa with NH 3.The Gaussian 03software package was used for all the calculations.28Geometry optimizations were carried out without symmetry constraints using the 6-31G(d)basis set to locate the stationary points on the potential energy surface.Species

*Corresponding author.E-mail:gtwang@https://www.wendangku.net/doc/c210745916.html,.

M(CH 3)3+NH 3T (CH 3)3M:NH 3

M )Al,Ga,In (1)

M(CH 3)3+2NH 3T (CH 3)3M:NH 3+NH 3T

(CH 3)3M:(NH 3)2M )Al,Ga,In (2)1094J.Phys.Chem.A 2006,110,1094-1099

10.1021/jp054133o CCC:$33.50?2006American Chemical Society

Published on Web 12/24/2005

involving hydrogen bonding were reoptimized using a larger 6-311++G(d,p)basis set incorporating diffuse and polarization functions on the hydrogen atoms.Single-point energy and frequency calculations using the6-311++G(d,p)basis set were performed for each stationary point to obtain the zero-point energies,thermal corrections,and infrared frequencies.The calculated stationary points on the potential energy surface were verified by analysis of the normal modes as minima by the absence of imaginary frequencies.All energies reported in this paper have been zero-point-corrected.Enthalpies were calculated at298K and1atm of pressure.Calculated frequencies were scaled using factors primarily determined using a least-squares fitting analysis comparing the calculated versus experimental spectra.29The scaling factors ranged from0.95to0.97. Calculated infrared spectra were synthesized using the calculated frequencies and intensities,assuming Gaussian line shapes with a26cm-1full width at half-maximum(fwhm).Peak assign-ments were made via visualization of the normal modes from the Gaussian log file.

Gas-phase infrared spectroscopy was performed with a Mattson RS-1FTIR spectrometer at2cm-1resolution.A heatable long path length gas cell was mounted in the sample compartment of the instrument.Briefly,the IR beam enters and exits through a single KCl window(6mm thickness)and is folded once with a Au-coated spherical mirror(r)40.6cm), giving an internal path length of～80cm.This intermediate value of path length gives a reasonable absorbance for the organometallic precursors and adducts without producing an excessive absorbance from the gas-phase NH3(which is200-800×higher in concentration).In all spectra shown for(CH3)3M +NH

3

mixtures,the very large NH3spectrum has been removed

for clarity.The gas cell was connected in parallel with our research MOCVD reactor and operated at flow rates and pressures in the same nominal range used for AlGaInN deposition.Gases were mixed before injection into the cell,with concentrations kept below the onset of adduct condensation.30 A long internal gas inlet tube allowed the gases to preheat before they were fully introduced into the cell.

The total pressure was varied from50to300Torr,with a total flow rate of6500cm3(STP)min-1.For this flow rate,at 300K and100Torr total pressure,the mean residence time in the cell is3.54s(internal volume)3.2L).Hydrogen was used as the carrier gas.The ammonia flow rate was fixed at1000 cm3(STP)min-1,giving P(NH3))7.7Torr at the50Torr total pressure condition.TMAl and TMGa were delivered using a standard bubbler configuration to give P(TMAl-monomer)) 11.7mTorr31and P(TMGa))31.4mTorr32at the50Torr total pressure condition.We assume that TMAl is100% dimerized at the bubbler conditions and give flow rates and partial pressures on a TMAl monomer basis.The partial pressure of the reactant scales with total pressure,so at300Torr the values are6-fold greater than the50Torr values given above. The flow rates of TMAl and TMGa in this study were set at 1.52and4.08cm3(STP)min-1,respectively.All spectra were taken at room temperature(24°C).

Theoretical Results

The optimized geometries of the adducts of TMAl and NH3 considered in this paper are shown in Figure1.The calculated geometries of the complexes of TMGa and NH3have virtually identical conformations to those of TMAl and NH3and hence are not separately shown.Selected geometric parameters of interest are also shown in Figure1for both TMAl and TMGa. Calculated binding energies and changes in enthalpies of all products relative to the reactants are given in Table1.

The optimized geometry of the1:1(CH3)3Al:NH3adduct is shown in https://www.wendangku.net/doc/c210745916.html,plex formation is energetically favorable,with a calculated binding energy(-?E ZPE)of19.6 kcal/mol and an enthalpy change of-20.4kcal/mol relative to the reactants.This compares to previous calculations of?E) -20.2to-25.9kcal/mol18,19,23and?H)-27kcal/mol21for

(CH3)3Al:NH3.The structure of(CH3)3Al:NH3has been previ-ously discussed in detail,and the calculated Al-N bond length of 2.117?for(CH3)3Al:NH3agrees well with previous calculations.12,18For(CH3)3Ga:NH3,we calculate a binding energy of11.8kcal/mol and a change in enthalpy of-12.6 kcal/mol.These values are slightly below previous calculations of?E)-14.4to-20kcal/mol19,23,33and?H)-15.9to -20.5kcal/mol.21,22,24,33Previous experimental estimates of?H range from-15.2to-16.3kcal/mol.33,34The calculated Ga-N bond length of2.185?is similar in magnitude to previous calculations22-24and an electron diffraction study which reported a Ga-N bond length of2.161?.14

It is also possible for a second NH3molecule to attack the metal atom of the1:1(CH3)3M:NH3complex to form a second M-N donor-acceptor bond,as shown in Figure1c.We previously reported the observation in the magnesocene(MgCp2 or Mg(C5H5)2)+NH3system of an analogous adduct also involving two donor-acceptor bonds,i.e.,H3N:MgCp2:NH3, at higher partial pressures.35The2:1H3N:(CH3)3M:NH3com-plex is calculated(relative to the reactants(CH3)3M+2NH3) to have a binding energy of19.7kcal/mol and?H of-21.1 kcal/mol where M)Al and a binding energy of12.0kcal/mol and?H of-12.9kcal/mol where M)Ga.The stabilization energy added by the second M-N donor-acceptor bond relative to the1:1(CH3)3M:NH3adduct is only0.1kcal/mol for M) Al and0.2kcal/mol for M)Ga.Thus,it is seen that the formation of the2:1H3N:(CH3)3M:NH3complex is

only Figure1.Optimized geometries of(a)unassociated(CH3)3M and NH3, (b)1:1(CH3)3M:NH3,(c)2:1H3N:(CH3)3M:NH3,and(d)2:1hydrogen-bonded(CH3)3M:NH3???NH3.

Complex Formation of(CH3)3M(M)Al,Ga)with NH3J.Phys.Chem.A,Vol.110,No.3,20061095

energetically favorable over the 1:1complex by a minimal amount.The calculated average length of the two Al -N bonds in H 3N:(CH 3)3Al:NH 3is 2.261?,a substantial increase from the calculated Al -N bond length of 2.087?in the 1:1(CH 3)3-Al:NH 3adduct.This increase in bond length indicates a significant decrease in the strength of the Al -N donor -acceptor bond upon formation of the second Al -N donor -acceptor bond.A similar phenomenon is observed for H 3N:(CH 3)3Ga:NH 3,where the calculated average length of the two Ga -N bonds is 2.437?,versus 2.185?in the 1:1(CH 3)3Ga:NH https://www.wendangku.net/doc/c210745916.html,pared to these results,Nakamura et al.calculated that formation of the 2:1H 3N:(CH 3)3M:NH 3adduct results in a greater (but still weak)stabilization energy of 5.7kcal/mol for M )Al and 3.1kcal/mol for M )Ga,relative to the (CH 3)3M:NH 3adduct.19Simka et al.calculated the formation of H 3N:(CH 3)3Ga:NH 3as being 5.2kcal/mol more energetically favor-able compared to (CH 3)3Ga:NH 3.21

We also consider here for the first time an alternative complex resulting from hydrogen bonding of the nitrogen atom of a second NH 3molecule to a hydrogen atom of the NH 3molecule in the 1:1(CH 3)3M:NH 3adduct,as shown in Figure 1d.The binding energy of this 2:1hydrogen-bonded adduct,which we write as (CH 3)3M:NH 3???NH 3,is calculated to be 26.8kcal/mol for M )Al and 18.4kcal/mol for M )Ga.This formation of this hydrogen bond represents a stabilization energy relative to the 1:1(CH 3)3M:NH 3adduct of 7.2kcal/mol for M )Al and 6.6kcal/mol for M )Ga.In comparison,recent theoretical studies of the hydrogen-bonded ammonia dimer (NH 3)2have calculated a weaker interaction energy of 1.6-3.8kcal/mol.36-39The stronger interaction in the 2:1hydrogen-bonded (CH 3)3M:NH 3???NH 3adduct is likely due to the greater electron deficiency of the NH 3molecule (compared to unassociated NH 3)involved in the donor -acceptor bond with TMAl or TMGa.

Significantly,the theoretical calculations thus indicate that formation of the 2:1hydrogen-bonded (CH 3)3M:NH 3???NH 3adduct is energetically favored over the 2:1H 3N:(CH 3)3M:NH 3complex by a substantial amount.The calculated Al -N and Ga -N bond lengths of 2.087and 2.197?,respectively,in the hydrogen-bonded (CH 3)3M:NH 3???NH 3adduct are almost un-changed from their values in the 1:1(CH 3)3M:NH 3adduct,indicating that this donor -acceptor bond is not significantly perturbed upon formation of the hydrogen bond.The N ???H hydrogen bond length is calculated to be 2.011?for (CH 3)3-Al:NH 3???NH 3and 2.045?for (CH 3)3Ga:NH 3???NH 3.Experimental Results

Trimethylaluminum +NH 3.We have collected the room-temperature gas-phase infrared spectra of the products resulting from mixing TMAl or TMGa with NH 3.Figure 2b shows the room-temperature IR spectrum of TMAl mixed with NH 3at a total pressure of 50Torr.The spectrum contains no discernible contribution from uncomplexed,gas-phase TMAl (not shown),and thus the TMAl is considered to be completely associated with NH 3.The spectrum at 50Torr is largely similar to the spectrum reported previously at 99°C at 100Torr by Creighton

et al.,26which was identified as the (CH 3)3Al:NH 3adduct in part by comparison with earlier condensed-phase spectra of (CH 3)3Al:NH 3.10,11,18When the pressure is increased to 300Torr,the infrared spectrum (Figure 2c)shows the disproportionately strong growth of several peaks (marked by vertical lines)which were very weak or not visible in the spectrum taken at 50Torr.This indicates an increase in the relative concentration of a second distinct product as the pressure is increased,where the growth peaks represent the new modes of the second product.This observation is consistent with reaction 2,whereby increas-ing the pressure will shift the equilibrium from the 1:1(CH 3)3-Al:NH 3adduct toward a 2:1(CH 3)3Al:(NH 3)2adduct involving a second NH 3molecule,according to Le Chatelier’s principle.Examination of the calculated spectrum of the 1:1(CH 3)3Al:NH 3adduct,shown in Figure 2a,shows an excellent fit to the spectrum at 50Torr (Figure 2b).Moreover,the calculated 1:1(CH 3)3Al:NH 3spectrum (Figure 2a)fails to predict the two growth peaks in the 300Torr data (Figure 2c)at 1053and 3126cm -1that can be associated with the 2:1adduct.The results thus show that,at 50Torr,the product mixture at room temperature consists primarily of the 1:1(CH 3)3Al:NH 3adduct and that,at 300Torr,the product mixture shifts toward a greater fraction of the 2:1(CH 3)3Al:(NH 3)2adduct.

To spectrally isolate and identify the 2:1adduct,the contribu-tion of the 1:1(CH 3)3Al:NH 3complex was removed from the 300Torr spectrum via subtraction of a multiple (5.2)of the 50Torr spectrum.The multiple is less than the simple ratio of the pressures (300/50)6)because at 300Torr there is propor-tionately less 1:1(CH 3)3Al:NH 3adduct than at 50Torr,due to an increase of the 2:1adduct.The resulting difference spectrum,

TABLE 1:Relative Energies and Enthalpies (kcal/mol)of the Calculated Species Shown in Figure 1,Calculated at the B3LYP/6-311++G(d,p)Level of Theory

M )Al

M )Ga ?E

?E ZPE ?H ?E ?E ZPE ?H (CH 3)3M +2NH 3

000000(CH 3)3M:NH 3+NH 3-22.5-19.6-20.4-14.6-11.8-12.6H 3N:(CH 3)3M:NH 3-24.8-19.7-21.1-16.3-12.0-12.9(CH 3)3M:NH 3???NH 3

-31.4

-26.8

-27.6

-22.7

-18.4

-

19.0

Figure 2.(a)Calculated IR spectrum of (CH 3)3Al:NH 3(frequency scale factor )0.96),(b)experimental IR spectrum of (CH 3)3Al:NH 3+NH 3at 50Torr,and (c)(CH 3)3Al:NH 3+NH 3at 300Torr.

1096J.Phys.Chem.A,Vol.110,No.3,2006Wang and Creighton

which represents the 2:1adduct,is shown in Figure 3a.To determine the identity of the 2:1complex,we plotted the calculated spectrum of the hydrogen-bonded (CH 3)3Al:NH 3???NH 3adduct,shown in Figure 3b,along with that of the H 3N:(CH 3)3Al:NH 3complex,shown in Figure 3c.It is seen that the spectrum of the hydrogen-bonded adduct matches the experi-mental 2:1adduct spectrum quite accurately and correctly predicts the experimental peaks at 1054,1245,3126,and 3361cm -1,which are modes that increase in relative intensity as the pressure is increased and can thus be uniquely assigned to the 2:1adduct.In contrast,the calculated spectrum of the H 3N:(CH 3)3Al:NH 3complex is a much poorer fit overall and notably fails to predict the 2:1adduct experimental peaks at 1054,1245,3126,and 3361cm -1.

On the basis of this analysis,we identify the 2:1(CH 3)3Al:(NH 3)2adduct,which increases in relative concentration as the pressure is increased,as the hydrogen-bonded (CH 3)3Al:NH 3???NH 3adduct.This assignment is consistent with the theoretical calculations of the energetics,which predicts that the hydrogen-bonded (CH 3)3Al:NH 3???NH 3adduct is lower in energy (more stable)by 7.1kcal/mol relative to the 2:1H 3N:(CH 3)3Al:NH 3complex.The observed experimental frequencies are listed in Table 2,along with selected mode assignments taken from the corresponding calculated hydrogen-bonded (CH 3)3Al:NH 3???NH 3frequencies.

Trimethylgallium +NH 3.The results for TMGa +NH 3are very similar to those of TMAl +NH 3.Figure 4b shows the room-temperature IR spectrum of TMGa mixed with NH 3at a total pressure of 50Torr.The spectrum at 50Torr is largely identical to that reported by Creighton et al.26at 53°C and 100Torr,which was identified as the (CH 3)3Ga:NH 3adduct.The spectrum is also in good agreement with previous reports of the gas-phase spectrum of (CH 3)3Ga:NH 3.13-15The calculated

(CH 3)3Ga:NH 3adduct is shown in Figure 4a and is a good fit with the 50Torr spectrum in Figure 4b.As with TMAl +NH 3,the disproportionately strong growth of some features is observed when the pressure is increased to 300Torr,as seen by the marked peaks in Figure 4c.Thus,similar to TMAl +NH 3,we observe for TMGa +NH 3the presence of a second,distinct product which increases in relative concentration as the pressure is increased.Following the same reasoning as that for TMAl +NH 3,this second product is assumed to involve

the

Figure 3.(a)Isolated experimental difference IR spectrum of 2:1(CH 3)3Al:(NH 3)2and (b)calculated IR spectrum of hydrogen-bonded (CH 3)3Al:NH 3???NH 3(frequency scale factor )0.96for ν(N -H)and ν(C -H)modes,0.95all other modes),and (c)calculated IR spectrum of H 3N:(CH 3)3Al:NH 3(frequency scale factor )0.96for ν(N -H)and ν(C -H)modes,0.97all other modes).Asterisks represent residual NH 3peaks.

TABLE 2:Experimental and Selected Calculated

Frequencies (cm -1)and Relative Intensities (arb units)for the Hydrogen-Bonded (CH 3)3Al:N (1)H 3···N (2)H 3Adduct

exp

calc a I assignment 3803ν(Al -N)4837νs (Al -C 3)56716νas (Al -C 3)60657821νas (Al -C 3)698688130δ(CH 3)rock

722184δ(N (1)H 3)rock +δ(CH 3)rock 751758174δ(N (1)H 3)rock +δ(CH 3)rock 10541060164δs (N (2)H 3)umbrella 116850δs (CH 3)117451δs (CH 3)

12451254168δs (N (1)H 3)umbrella 1627162316δas (N (1)H 3)2825286043ν(CH 3)2885286346ν(CH 3)289228655ν(CH 3)291436ν(CH 3)2919291837ν(CH 3)291962ν(CH 3)2924292412ν(CH 3)292599ν(CH 3)31263094539νs (N (1)H 3)33274νs (N (2)H 3)3361

336237νas (N (1)H 3)3434

20.8

νas (N (2)H 3)

a

ν(C -H)and ν(N -H)modes scaled by 0.96;all other modes scaled by

0.95.

Figure 4.(a)Calculated IR spectrum of (CH 3)3Ga:NH 3(frequency scale factor )0.96),(b)experimental IR spectrum of (CH 3)3Ga +NH 3at 50Torr,and (c)(CH 3)3Ga +NH 3at 300Torr.

Complex Formation of (CH 3)3M (M )Al,Ga)with NH 3J.Phys.Chem.A,Vol.110,No.3,20061097

bonding of a second NH 3molecule to TMGa,i.e.,a 2:1(CH 3)3-Ga:(NH 3)2adduct.

As with TMAl +NH 3,we isolated the spectrum of the 2:1(CH 3)3Ga:(NH 3)2adduct via subtraction of a multiple (5.6)of the 50Torr spectrum from the 300Torr spectrum,the result of which is shown in Figure 5a.A negative peak at 1137cm -1and nearby artifacts are observed resulting from imperfect subtraction of the 1:1(CH 3)3Ga:NH 3adduct,possibly due in part to peak shifting.The calculated spectrum of the 2:1hydrogen-bonded (CH 3)3Ga:NH 3???NH 3adduct is shown in Figure 5b,along with that of the 2:1H 3N:(CH 3)3Ga:NH 3complex,shown in Figure 5c.Again,similar to the case with TMAl,the calculated hydrogen-bonded (CH 3)3Al:NH 3???NH 3spectrum in Figure 5b predicts the major features of the experimental spectrum quite accurately.The calculated 2:1H 3N:(CH 3)3Ga:NH 3spectrum represents a worse fit overall and notably fails to predict the strong peak at 3167cm -1.From this analysis,combined with the energetics calculations which predict that the hydrogen-bonded (CH 3)3Ga:NH 3???NH 3adduct is more stable than the H 3N:(CH 3)3Ga:NH 3complex by 6.4kcal/mol,we can assign the higher pressure 2:1adduct as the hydrogen-bonded (CH 3)3Ga:NH 3???NH 3adduct.The observed experimental frequencies are listed in Table 3,along with selected mode assignments taken from the corresponding calculated hydrogen-bonded (CH 3)3Ga:NH 3???NH 3frequencies.Conclusions

We have investigated the formation of gas-phase adducts of trimethylaluminum and trimethylgallium with ammonia using room-temperature FTIR experiments and DFT calculations.Our results indicate that at higher pressures,for both TMAl and TMGa,a product distinct from the well-known and observed

(CH 3)3M:NH 3adduct grows https://www.wendangku.net/doc/c210745916.html,parison of the experimental and calculated IR spectra,along with calculations of the energetics,indicates that this second product is the result of hydrogen bonding of a second NH 3molecule to the (CH 3)3M:NH 3adduct,which can be written as (CH 3)3M:NH 3???NH 3.Although we have not investigated the reaction of TMIn +NH 3here,we suspect that the hydrogen-bonded (CH 3)3In:NH 3???NH 3adduct may be observed for that system as well,based on the similar chemistries of TMAl,TMGa,and TMIn with NH 3.Because our experimental parameters fall within the typical range for MOCVD AlGaInN deposition,the formation of the hydrogen-bonded (CH 3)3M:NH 3???NH 3adduct can be expected in MOCVD AlGaInN systems,particularly those operated at higher pressures.These results show that hydrogen bonding plays an important role in the interaction of ammonia with metalorganic precursors involving Al,Ga,and In and may also play a role in higher temperature parasitic chemical reactions during AlGaInN growth.The results also suggest that hydrogen bonding may be possible in systems involving TMAl,TMGa,and TMIn with PH 3and AsH 3,which are relevant to AlGaInAs and AlGaInP growth.However,the substantially weaker ability of PH 3and AsH 3to form hydrogen bonds in comparison to NH 340suggests that formation of the analogous 2:1hydrogen-bonded adduct in these systems may be much less favorable.Acknowledgment.Sandia is a multiprogram laboratory operated by Sandia Corp.,a Lockheed Martin Co.,for the United States Department of Energy’s National Nuclear Security Administration under Contract No.DE-AC04-94AL85000.We especially acknowledge support from the Office of Basic Energy Sciences.

References and Notes

(1)Han,J.;Figiel,J.J.;Crawford,M.H.;Banas,M.A.;Bartram,M.E.;Biefeld,R.M.;Song,Y.K.;Nurmikko,A.V.J.Cryst.Growth 1998,195,291.

(2)Chen,C.H.;Liu,H.;Steigerwald,D.;Imler,W.;Kuo,C.P.;Craford,M.G.;Ludowise,M.;Lester,S.;Amano,J.J.Electron.Mater.1996,25,1004.

(3)Sayyah,K.;Chung,B.C.;Gershenzon,M.J.Cryst.Growth 1986,77,

424.

Figure 5.(a)Isolated experimental difference IR spectrum of 2:1(CH 3)3Ga:(NH 3)2and (b)calculated IR spectrum of hydrogen-bonded (CH 3)3Ga:NH 3???NH 3(frequency scale factor )0.96for ν(N -H)and ν(C -H)modes,0.95all other modes),and (c)calculated IR spectrum of H 3N:(CH 3)3Ga:NH 3(frequency scale factor )0.96).Asterisks represent residual NH 3peaks.

TABLE 3:Experimental and Selected Calculated

Frequencies (cm -1)and Relative Intensities (arb units)for the Hydrogen-Bonded (CH 3)3Ga:N (1)H 3···N (2)H 3Adduct

exp

calc I assignment

28051ν(Ga -N)

31124δ(N (1)H 3)rock +δ(N (2)H 3)rock 34325δ(N (1)H 3)rock +δ(N (2)H 3)rock 4714νs (Ga -C 3)50334νas (Ga -C 3)50737νas (Ga -C 3))

55253932δ(N (1)H 3)rock +δ(CH 3)rock 72068δ(CH 3)rock

73972697δ(N (1)H 3)rock +δ(CH 3)rock 10401051169δs (N (2)H 3)umbrella 116125δs (CH 3)117123δs (CH 3)

11861202146δs (N (1)H 3)umbrella 1619162212δas (N (1)H 3)2860287651ν(CH 3)288052ν(CH 3)2945294543ν(CH 3)3167

3143449νs (N (1)H 3)33284νs (N (2)H 3)337544νas (N (1)H 3)3437

19

νas (N (2)H 3)

a

ν(C -H)and ν(N -H)modes scaled by 0.96;all other modes scaled by 0.95.

1098J.Phys.Chem.A,Vol.110,No.3,2006Wang and Creighton

(4)Nakamura,F.;Hashimoto,S.;Hara,M.;Imanaga,S.;Ikeda,M.; Kawai,H.J.Cryst.Growth1998,195,280.

(5)Safvi,S.A.;Redwing,J.M.;Tischler,M.A.;Kuech,T.F.J. Electrochem.Soc.1997,144,1789.

(6)Mihopoulos,T.G.;Gupta,V.;Jensen,K.F.J.Cryst.Growth1998, 195,733.

(7)Thon,A.;Kuech,T.F.Appl.Phys.Lett.1996,69,55.

(8)Creighton,J.R.;Breiland,W.G.;Coltrin,M.E.Electrochem.Soc. Series2002,28.

(9)Matsumoto,K.;Tachibana,A.J.Cryst.Growth2004,272,360.

(10)Watari,F.;Shimizu,S.;Aida,K.;Takayama,E.Bull.Chem.Soc. Jpn.1978,51,1602.

(11)Ault,B.S.J.Phys.Chem.1992,96,7908.

(12)Mu¨ller,J.;Ruschewitz,U.;Indris,O.;Hartwig,H.;Stahl,W.J. Am.Chem.Soc.1999,121,4647.

(13)Sywe,B.S.;Schlup,J.R.;Edgar,J.H.Chem.Mater.1991,3, 737.

(14)Almond,M.J.;Jenkins,C.E.;Rice,D.A.;Hagen,https://www.wendangku.net/doc/c210745916.html,anomet. Chem.1992,439,251.

(15)Kim,S.H.;Kim,H.S.;Hwang,J.S.;Choi,J.G.;Chong,P.J.; Sywe,B.S.;Schlup,J.R.;Edgar,J.H.;Piocos,E.A.;Ault,B.S.;Bertolet, D.C.;Rogers,J.W.;Mazzarese,D.;Jones,K.A.;Conner,W.C.Chem. Mater.1994,278.

(16)Durig,J.R.;Bradley,C.B.;Odom,J.D.Inorg.Chem.1982,21, 1466.

(17)Piocos,E.A.;Ault,B.S.J.Mol.Struct.1999,476,283.

(18)Mu¨ller,J.J.Am.Chem.Soc.1996,118,6370.

(19)Nakamura,K.;Makino,O.;Tachibana,A.;Matsumoto,K.J. Organomet.Chem.2000,611,514.

(20)Makino,O.;Nakamura,K.;Tachibana,A.;Tokunaga,H.;Akutsu, N.;Matsumoto,K.Appl.Surf.Sci.2000,159/160,374.

(21)Simka,H.;Willis,B.G.;Lengyel,I.;Jensen,K.F.Progress in Crystal Growth and Characterization of Materials7th European Workshop on MOVPE(EWMOVPE),June8-11,1997;Haus Wissenshcaft& Kultur:BERLIN,Germany,1997;Vol.35,p117.

(22)Timoshkin,A.Y.;Bettinger,H.F.;Schaefer,H.F.J.Phys.Chem. A2001,105,3240.

(23)Ikenaga,M.;Nakamura,K.;Tachibana,A.;Matsumoto,K.J.Cryst. Growth2002,237,936.

(24)Watwe,R.M.;Dumesic,J.A.;Kuech,T.F.J.Cryst.Growth2000, 221,751.

(25)Creighton,J.R.;Breiland,W.G.;Coltrin,M.E.;Pawlowski,R. P.Appl.Phys.Lett.2002,81,2626.

(26)Creighton,J.R.;Wang,G.T.;Breiland,W.G.;Coltrin,M.E.J. Cryst.Growth2004,261,204.

(27)Becke,A.D.J.Chem.Phys.1993,98,5648.

(28)Frisch,M.J.;Trucks,G.W.;Schlegel,H.B.;Scuseria,G.E.;Robb, M.A.;Cheeseman,J.R.;Montgomery,J.A.,Jr.;Vreven,T.;Kudin,K. N.;Burant,J.C.;Millam,J.M.;Iyengar,S.S.;Tomasi,J.;Barone,V.; Mennucci, B.;Cossi,M.;Scalmani,G.;Rega,N.;Petersson,G. A.; Nakatsuji,H.;Hada,M.;Ehara,M.;Toyota,K.;Fukuda,R.;Hasegawa,J.; Ishida,M.;Nakajima,T.;Honda,Y.;Kitao,O.;Nakai,H.;Klene,M.;Li, X.;Knox,J.E.;Hratchian,H.P.;Cross,J.B.;Adamo,C.;Jaramillo,J.; Gomperts,R.;Stratmann,,R.E.;Yazyev,O.;Austin,A.J.;Cammi,R.; Pomelli,C.;Ochterski,J.W.;Ayala,P.Y.;Morokuma,K.;Voth,G.A.; Salvador,P.;Dannenberg,J.J.;Zakrzewski,V.G.;Dapprich,S.;Daniels, A. D.;Strain,M. C.;Farkas,O.;Malick, D.K.;Rabuck, A. D.; Raghavachari,K.;Foresman,J.B.;Ortiz,J.V.;Cui,Q.;Baboul,A.G.; Clifford,S.;Cioslowski,J.;Stefanov,B.B.;Liu,G.;Liashenko,A.;Piskorz, P.;Komaromi,I.;Martin,R.L.;Fox,D.J.;Keith,T.;Al-Laham,M.A.; Peng,C.Y.;Nanayakkara,A.;Challacombe,M.;Gill,P.M.W.;Johnson, B.;Chen,W.;Wong,M.W.;Gonzalez,C.;Pople,J.A.Gaussian03, Revision A.5ed.;Gaussian,Inc.:Pittsburgh,PA,2003.

(29)Scott,A.P.;Radom,L.J.Phys.Chem.1996,100,16502.

(30)Creighton,J.R.J.Electron.Mater.2002,31,1337.

(31)Using TMAl vapor pressure curve from:Stringfellow,G.B. Organometallic Vapor-Phase Epitaxy,2nd ed.;Academic Press:San Diego, CA,1999.

(32)Using TMGa vapor pressure curve from:Plass,C.;Heinecke,H.; Kayser,O.;Luth,H.;Balk,P.J.Cryst.Growth1988,88,455.

(33)Pelekh,A.;Carr,R.W.J.Phys.Chem.A2001,105,4697.

(34)Creighton,J.R.;Wang,G.T.J.Phys.Chem.A2005,109,133.

(35)Wang,G.T.;Creighton,J.R.J.Phys.Chem.A2004,108,4873.

(36)Abu-Awwad,F.M.THEOCHEM2004,683,57.

(37)Altmann,J.A.;Govender,M.G.;Ford,T.A.Mol.Phys.2005, 103,949.

(38)Huang,N.;MacKerell,A.D.J.Phys.Chem.A2002,106,7820.

(39)Boese,A.D.;Chandra,A.;Martin,J.M.L.;Marx,D.J.Chem. Phys.2003,119,5965.

(40)Sennikov,P.G.J.Phys.Chem.1994,98,4973.

Complex Formation of(CH3)3M(M)Al,Ga)with NH3J.Phys.Chem.A,Vol.110,No.3,20061099