Synthesis and Characterization of Diamond and Boron Phosphide

Semiconducting diamond has been synthesized from carbon-metal melts in a 600 ton tetrahedral anvil press at about 60 kbar and 1400°c. The experimental set up, pressure and temperature calibrations, and the growth region in the pressure-temperature regime are indicated. Micrographs of synthesized crystals are shown. The semiconducting properties of diamond, doped with boron, aluminum and titanium have been interpreted using the log R versus l/T curves. In boron-doped diamond ''high concentration" type impurity conduction occurs and the activation energies vary from 0.15 eV to 0.30 eV. The activation energies for aluminum and titanium-doped samples are found to be 0.31 eV and 0.40 eV respectively. The results are consistent with those obtained from optical methods where data were available. Properties of diamond irradiated with 17-18.3 MeV protons have been investigated. Raman· spectral measurements indicate an increase in the lattice constant of 0.0032 ~. The infrared absorption spectrum of proton irradiated diamond shows a characteristic absorption at 6.92 microns. The optical absorption edge of diamond does not seem to be affected by the irradiation. Boron phosphide has been synthesized from the elements at preso sures and temperatures above 20 kbar and 1200 c. The crys tal growth rate has been determined as a function of temperature and press ure from which an activation energy for the process is derived. Optical and Scanning Electron Micrographs of the crystals synthesized revealed a poorly developed morphology with voids present under all conditions of pressure and temperature. The effect of thermal neutron irradiation on the electrical conductivity of boron phosphide, hexagonal boron nitride and boron oxide (B2o3) has been observed by studying the current-voltage characteristics before and during irradiation with a neutron flux of 8 2 about 10 n/cm .sec. In all these compounds, currents were higher, for the same voltage setting, during irradiation. The differences observed during irradiation in current values for the three boron compounds have been explained as being due to the "boron to anion ratio" being different in them. The Appendix to this thesis includes an introduction to the foregoing investigations and describes the equipment used with recommendations for future work.


TABLE OF CONTENTS
The thermodynamic calculation of the equilibrium line between graphite and diamond 1 has been a major contribution towards diamond synthesis. It is in good agreement with the experimental results ob-2 tained by other workers.
Contributions to the carbon phase diagram 2-4 have been made by several investigators.
The part of the diamondgraphite equilibrium line from 0 to 1200°K is based on moderately accurate thermodynamic data. From 1400°K to 2800°K the data is based 5 upon experiments on the growth and graphitization of diamond. It has been suggested that at pressures above 600-700 kbar diamond reduces to a dense metallic state, 15-20 percent denser than diamond. 4 Pressure equipment of various types have been designed and used 3 6 for a variety of purposes. ' It is now possible to assess the merits and demerits of each and select the required type depending upon whether the main consideration is pressure, temperature or sample volume. The tetrahedral anvil press used in this investigation lends itself particularly well for laboratory experiments as needed in this investigation. Experimental data is available regarding the pressure and temperature at which diamond growth occurs. It varies for dif-3 ferent catalyst systems used. The direct conversion of graphite to diamond (without catalyst) is also possible except that the required pressures are above 125 kbar. 4 Different types of reaction cells have been devised. They can either be directly heated or indirectly heated, 2 ' 3 ' 6 (with a heater tube enclosing the sample). The actual transformation from carbon to diamond, when using a solvent, occurs 3 across a very thin solvent film separating the carbon and diamond.
The growth region in the pressure-temperature regime is the area bounded by the melting point line of the solvent and the graphite diamond equilibrium line. 7 5 Semiconductivity in synthetic diamond is induced by introduction of specific impurities like Be, B or Al to suitable mixtures of carbon and solvents like the transition metals iron, nickel and 8,9 cobalt.
The impurities may be introduced by direct addition to 10 11 the growth mixture, diffusion techniques or ion bombardment. ' The electrical resistivity of the crystal decreases with increasing concentration of the impurity atoms. The crystals may have resistivities as low as 10 3 ohm-cm with activation energies for conduct i on It has been observed that the electrical properties can be markedly altered by subjecting the crystals to 7 heat treatment. Table I shows the electrical characteristics of natural semiconducting crystals.
The effect of pressure and temperature is important not only in the actual production of semiconducting crystals but also in deter-. 15 16 mining the color, size and shape of the crystals synthesized. ' At low pressures where nucleation is slow, it is found that cubic development is favored at the lower temperature, gradually changing to cubo-octahedra and eventually octohedra at higher temperatures.
At higher pressure s where nucleation is rapid, fine sized crystals are normally encountered.
Doping, apart from changing the electrical and optical properties 8 ' 9 also influences the morphology of the crysta l s. For example, in boron doped diamond crystals the cubic habit is pro- press. The details of construction and operation may be found elsewhere. Heating of the samples was accomplished by so-called direct heating whereby a transformer provides a low voltage, high current (600-800 amp e res) which is passed via the press anvils through the sample itself. Runs were made in which the total wattage passed through the cell was monitored and the temperature read from a previously obtained calibration curve on the specific sample design used.
Temperature calibrations were made using a Pt-Pt, 10% Rh thermocouple.
A typical calibration curve is shown in Figure 1. Pressure c alibration is done by obtaining the relationship between ram pressure and working pressure. Known pressure values for polymorphic transition in bismuth and thallium were used. These transitions are accompanied by sharp resistivity changes in the metals. Cells containing the respective metal s , which were extruded into wires from bulk form in our laboratory, are monitored for this resistivity change and the appropriate ram pre ssure noted. A typical calibrati on curve is shown in Figure 2. All runs were s ubsequently made by monitoring the r am pressures . No attempts were made to individually determine press ure and/or temperature in each cell. The cell design for actual crystal growth i s shown in Fi gure 3.
A typical growth sequence consisted of pressurizing the sample followed by gradual heating of the sample by increasing the wattage in about three minutes after which growth ensues at essentially stable conditions for about ten to fifteen minutes. At the end of the run the current is rapidly decreased, effectively quenching the sample, after which the pressure is gradually decreased. After removal of the sample and the surrounding pyrophyllite pressurizing medium, the cell is "cleaned" to yield the desired end product.
Cleaning is done by dissolution and oxidation processes in a variety of baths. The sequence of events is as follows: The cell is first broken up mechanically and observed under * the microscope. The sample holder material (pyrophyllite) is carefully removed as much as possible without losing any of the diamond crystals.
The material is then transferred to a clean 250 cc beaker.
To this is added 150 cc of a saturated solution of sodium or potassium dichromate in concentrated sulfuric acid. This solution is now reddish brown in color. It is carefully heated at about 60°c for half an hour. Stirring is done by means of a glass coated magnetic stirrer.
When the solution turns bluish green in color, no further reaction takes place and the heating is stopped. The mixture is allowed to settle. The clean solution is carefully decanted and the material washed with water and decanted again. This step is repeated until all acid is fully diluted. If graphite is still left, * See page 115 of Appendix. the above procedure is repeated. Aqua regia (3 volumes of cone. HCl and 1 volume of cone. HNO ) is used to remove the metal. The solution and the sample is 3 allowed to boil for about an hour or so using the magnetic stirrer, until all metal has dissolved. It is allowed to settle, decanted and washed repeatedly.
Some of the pyrophyllite which may still remain is dissolved by adding hydrofluoric acid to the residue in a teflon beaker. The solution is heated gently for about one half hour, allowed to cool and then decanted and washed with water as often as necessary. The only residue is now pure diamond.
Results and Discussion:   Figure 5 shows the morphology of the synthesized crystals. Change of morphology as a function of boron dopant level in diamond in the pressure-temperature regime is described in reference 15.
The addition of dopants generally suppresses the equilibrium line, as shown in Figure 6, to lower pressures. The extent of the pressure reduction is dependent on the type and concentration of dopant.
It can therefore not be stressed enough that a variety of parameters can influence the characteristics of the end product.
Subsequently, our growth experiments have been aimed at determing the best conditions for growth of a light-element, fine-grain-size, high-temperature stable, semiconducting product. Such material, in the form of boron doped semiconducting diamond has been delivered to the sponsor.       Experimental results in the diamond growth system.
Micrograph of synthesized diamond (44X).    tion with regard to growth of the material itself, but also on an understanding or interpretation of the conduction mechanism(s) of variously doped diamond. This paper presents the dependence log R = f(l/T) for variously doped diamonds and advances an empirical formula which fits the experimental results fairly well. 2 An inspection of the log R vs l/T curves (e.g. see Fig. 1) of synthesized diamond reveals that they are different from those usually observed in semiconductors at normal doping levels. For these "ordinary" semiconductors, the plot of log R vs l/T normally shows a minimum, which occurs at higher tempera tu res as the impurity concentration increases. To the low temperature side of the minimum, i.e.
in the extrinsic range, the curve is almost linear. Most of the charge carriers come from an impurity level by thermal ionization with the consequence that carrier concentration varies roughly as exp (Ei /nkT), where E is the energy level of the impurity, mp imp T the temperature, k the Boltzmann constant, and n is equal to 1 or 2 depending on whether the semiconductor is "strongly" compensated or not. The low temperature asymptotic slope of the plot gives the activation energy of the impurity level. The variation of these curves for different dopant levels can be explained satisfactorily and quantitatively by the usual theory for the conduction mechanism ll-Boltzmann statistics.
a11d Maxwe Our curves differ, however, from They don't show any tendency of reaching a the usual in two ways. 24 0 a t temperatures as high as 800 K, compared to much lowldnimum even at which most of the normal curves have already er temperatures h d their minima, and secondly, there is a "kink" in the slope.
reac e The temperature at which this occurs is dependent upon majority impurity concentration (Fig. 1). explanation.
This kind of variation needs further Semiconductivity may be induced in diamond in a variety of ways. The most common method is that of the introduction of specific impurities like Be, B, Al to suitable growth mixtures of carbon and solvents or by diffusion of these elements into diamond crystals at 3 4 high temperatures and pressures. ' Introduction of active impurities by means of ion implantation has lately received considerable 5-7 attention.
The electrical resistivity of the crystals decreases with increasing concentration of the impurity atoms ( Fig. 1). Doping during growth or by diffusion has so far produced only p-type conductivity whereas ion doping (with Li, C and . P) can produce n-type conductivity. 6    19 It is found from optical and magnetic resonance measurements that in synthetic boron-doped diamond, nitrogen is the primary donor impurity, with its excess electron located in an antibond orbital where it does not contribute to conduction but may be available for compensation. In our samples nitrogen is not added intentionally.
For convenience of calculation it is assumed that the nitrogen concentration is low and approximately constant for different boron concentration, which is also borne out by infrared measurements.
The boron content is lowest in curve 1 and highest in 4 ( Fig. 1), and the compensation ratio in the reverse order. The low temperature slopes, i.e. low temperature activation energies, are .0304 eV, • 0219 eV, and 0.007 eV for curves 1, 2, 3, and 4 respec-.026 eV, tively.
Miller 2 0 studying the hopping mechanism in the low concenra nge was able to deduce the concentration and compensation tration dependence of E, which has been confirmed experimentally at least be 2xl0 /cm , 5xl0 /cm , lxlO /cm , and 2xl0 /cm for curves 1, 2, 3, and 4 of Figure 1 respectively. High impurity concentrations for all four cases are, however, still indicated. The solid solubility of B in Si and Ge is of the order 5xlo 21 /cm 3 at 1000°K. This is considerably higher than that of other elements and i 22 s primarily due to the small radius of boron.
Since the atomic radius of diamond i s smaller than Si and Ge, we expect a lower solubility of B in diamond. But it is quite unlikely to be lower than 19 3 lO /cm because boron can still go into diamond as a substitional impurity. It is the re for e reas onable to expect that semiconduc t ing diamond has impurity concentrations high enough for band conduction. yr.

ANALYSIS OF EXPERIMENTAL DATA
chosen.

The analysis of the data depends on what physical model is
Since the high temperature range of the curve (i.e. before 29 the change of slope) corresponds to the extrinsic range of a regular i du ctor a "two-impurity" model with boron as majority impurity sem con ' and nitrogen, presumably as minority impurity, seems adequate. If a single acceptor energy level is also assumed, the hole concentration by 23 is given where p = hole concentration, Nd = donor concentration, k =Boltzmann constant, h =Planck's constant, E activation energy and T = temperature.
The resistance curves can normally be fitted as a function of an m as parameters. 1s tee nique is used when Hall coefficient data or equivalently the carrier concentration data are available, with no complication due to mobility.
The Resistance R is defined as where L, A, p, e, p are conventional terms and the mobility is The mobility due to lattice scattering is given by 25  With µ,.., µ in Eq. (2) the only quantity needed to compute -t the resistance is P• where K = Nd/Na = compensation ratio. 31 It is found that the resistance curves (1) and (2) of Figure (1) cannot be fitted using case (i). This probably reflects that assumption p << Nd is not fully satisfied.

Case (ii)
If the other extreme i's true, i' e N << P < N Eq (1) -3/4 the mobility is proportional to (kT) for the high which means region of the curves (1) and (2) of Figure 1, , assuming temperature r e strictly linear in that region and the activation energy they a becomes twice as large as that of case (i). This larger activation energy will reduce the hole concentration (p) of Eq. (1). This makes the assumption Nd << p < N less likely to be valid. a On the other hand, when the mobility ~(kT)-3 / 2 was chosen and the corresponding activation energy used in Eq. (1), the assumption p << Nd < Na seemed less valid. Thus it is safe to conclude that the correct activation energy lies between case (i) and case (ii) and has been found to be 0.15 and 0.30 eV respectively as indicated in Table 1. Po (Nmaj) = re s istivity at a fixe d t emperature.
This expression is derived for the low temperature resistivity of the hopping type of impurity conduction. No theory on the majority centration dependence of resistivity, at temperatures :llDpuri ty con transition from normal to impurity conduction occurs, has where However, guided by Miller's relation, we found that been given. For comparison, the curve for natural diamond is also included in -3/2 Figure 4. If the constant preceding the T term is set equal to the numerical value given by Eq. (2'), then the two curves can be fitted with E, N . , K as listed in Table 2 to within a constant maJ factor of about 10 in both cas es. This cons tant factor could only come in due to the uncertainty of the crystal dimensions. In the above fittings, we assume d the diamond crystals to be cubic with a 32 250 micron edge. All the data are contained in Table s 1 and 2 .

!V• DISCUSSION AND CONCLUSIONS
Among the data g iven in Table 1 and Table 2      Average resistance as a function of percent boron in growth system compared to fit as shown.
Resistance of aluminum doped diamond as a function of temperature. Best fit as shown.
Resistance of natural diamond (curve 1, inner scale) 37 and titanium doped diamond as a function of temperature.
Best fit for titanium doped sample is as shown.

2.
It is approximately the dimensions of the diamonds from which the experimental data were obtained. Gamma-radiation induces fluorescence in the UV region. 5 It has also been found that optical transmission can be 6 increased by annealing neutron irradiated type I diamond.

E. c. Lightowlers
However, apparently no information has been published on the determination of changes in proton-irradiated diamond.
The irradiations were carried out in the proton beam f rom a      Largest crystals are reported to be obtained at process 0 temperatures between 1200 and 1300 C; slow cooling had no effect on the crystal growth but prolonged reaction times produced larger 1 It is further claimed 11 that the crystallization was crysta s. due to the temperature gradient within the sample chamber. The reaction between boron and phosphorus achieves its top rate at 0 12 temperatures above 1200 C and at pressures above 10,000 kbar.
It is interesting to note however, that so far the kinetics of BP growth under high pressure has not yet been studied. It was the purpose of this investigation to study the rate of growth of BP as a function of pressure, temperature and time.

EXPERIMENTAL PROCEDURES
Details of the 600 ton tetrahedral anvil press used in these 13 experiments may be found elsewhere.
The cell in which the boron phosphide crystals have been grown is illus' trated in Figure 1. It consists of a graphite heater tube containing the charge which is 65 201. boron and 80% phosphorus by weight. The ends of the tube are plugged with graphite and tantalum discs with tantalum tabs for electrical conduction. The entire cell is placed in a pyrophyllite tetrahedron. The sample is directly heated by a transformer which provides a low voltage, high current via the press anvils. In all runs the total wattage passed through the cell was monitored and the temperature determined from a standard calibration curve which had ly been constructed for this cell geometry using a Pt-Pt 10% previous Rh thermocouple. A pressure calibration using standard reference pointsl4 provided the relation between the ram pressure and actual cell pressure.
For each individual run, the sample was first pressurized to the desired level and subsequently heated to the required temperature and held for the appropriate length of time.
The cleaning of the crystals involved removing the unreacted phosphorus as the first step, using nitric acid. This was followed by treating the charge in chromic acid and hydrofluoric acid to remove graphite and pyrophyllite particles respectively. The cleaned crystals were identified by X-ray diffractometry initially and later by light microscopy, size permitting.
Particle size determination was accomplished with the use of a polarizing microscope equipped with a filar micrometer. The "size" of a crystal was established as the average of the largest dimensions "a" and "b" measured in mutually perpendicular directions. The average crystal size of each run was obtained by . dividing the sum total of "size" by the number of crystals measured in each case and this was taken as representative of the run. Standard deviation calculations have borne out the validity of this approach. Approximately fifty crystals were measured per run to arrive at the average crystal size. Table 1 shows the average crystal size for different growth times, temperatures and pressures. The average size has been plotted as a function of time for various temperatures in Figure 2 and for P ressure levels in Pressure appears to have an effect on size similar to that of temperature, though not to the same extent (see Figure 3). An increase in pressure increases average crystal size. The data shows that the rate is, however, decreased as a result of higher nucleati on.
The values of the positive slopes of the curves shown in  Figure 5. From the slope of this plot, the activation energy was determined to be 49.27 kcal/mole. For reasons previously given, no attempt was made to determine activation energies separately for nucleation and growth. · n energy reported here can only represent the activation The activatio energy for the total process. Although no data of this nature apto have been reported before, it should be kept in mind that pears i t ion energies measured in this way are of ten much larger than act va 15 those predicted by Eyring's theory.
An explanation for such a devi- 16 17 is g iven by Turnbull and Mott. ation One aspect which is not evident from optical microscopic examination, has been brought out with the aid of a scanning electron microscope; virtually all crystals have a vesicular structure.   After deposition the target was carefully removed and dried. A thin film of aluminum was evaporated on to the target in order to establish uniform contact,

RESULTS AND CONCLUSION
Using the data given in Table 1, a semilog plot of current versus voltage is made for BN pellet as indicated in Figure 3. For electrophoretically deposited BN . Figure 4 i~ drawn using Table 2 data. The dark current is two orders of magnitude smaller than that of the BN pellet. This could be due to the smaller thickness of the deposit as compared to the pellet. The current seems to saturate at relatively low voltages. Some of the other advantages of the electrophoretically deposited target are that the dark currents observed here are clos e to those normally encountered in neutron detection. Secondly, the material thickness plays an important part in the relaxation time and since electrophoretic deposits can I I be made in the range of microns they may prove superior to other forms of target making where evaporation is not quite so straightforward. 92 The current voltage plot for boron oxide is shown in Figure 5.
The trend seems to be similar to that of the BN targets except that the increase in current is much less. The increase in conductivity is governed to a large extent by the neutron interaction with the target material and this interaction increases with increasing boron concentrations in the host crystal since neutrons interact mostly only with the boron atoms present in these compounds. Hence it may be said that the boron concentration being less in s 2 o 3 than in BN is the cause for the lowered increase in current. Figure 6 for BP shows quite a high dark current as compared to that of BN and B 2 o 3 and this may be due to its smaller band gap.
The interesting feature of this plot is the fact that the increase in current during irradiation is more pronounced at higher voltages and this material may therefore be more suited for those targets which have to operate at high fields.
From this preliminary investigation it appears as though these materials may be promising from the point of view of thermal neutron detectors. However, more precise quantitative measurements are needed in order to make any conclusive statement.   Devices such as cathodoluminescent imaging tubes, based on optoelectronic principles, also make for significant interest in new semiconductor materials. 106 The work in this thesis has been concerned with the synthesis, characterization, and applications of some of the important members of the family discussed above, specifically diamond and boron phosphide. This appendix gives an overview of the relevant literature and background on these materials in the two separate sections following.

A. Diamond and Semiconducting Diamond
The graphite-diamond equilibrium line, calculated from thermo-3 dynamic concepts, has been a major contribution for all diamond synthesis work. 4 Subsequently a phase diagram has been reported for 0 carbon extending up to 5000 K and 800 kbar. Phase equilibria for "indirect" formation of diamond (via a metal "catalyst" or solvent), from graphite have also been investigated 5 along with nucleation and growth characteristics. For a metal solvent like nickel, the nickel- 6 carbon phase diagram has been established.
A method for predicting a priori the conditions for indirect diamond synthesis has been 7 developed and calculations have been made for the nickel-carbon and manganese-carbon phase equilibria. The lower pressure-temperature limits for indirect synthesis seems to be controlled by the eutectic temperature at pressure and the diamond-graphite equilibrium line. 7 The direct conversion of carbon into diamond requires pressures, static or dynamic, in excess of 130 kbar and hence is seldom used.
There are two points of view on the graphite-diamond transition. According to one, it is due to rupture of the crystal lattice of graphite and transfer of free carbon atoms, in a melt of "solvent 8-10 catalyst", to the growth surface of the diamond by diffusion.
Others consider the transition to involve reconstruction of the graphite lattice without breakup into individual atoms (solid-state .
The diffusion mechanism seems to be the more 13 prominent one though it does not exclude the possibility of the solid-state mechanism.
The introduction of semiconductivity in diamond is a relatively easy task and can be accomplished in a variety of ways. The most common method is that of the introduction of specific dopants like Be, B, Al either through addition to growth mixtures of carbon and solvents or through diffusion of these elements into diamond crystals 14-16 at high temperatures and pressures.
Introduction of active impurities by means of ion implantation has lately received consider-.

17-19 able attention.
This investigation has been concerned with the semiconducting properties and mechanism of conduction in diamond brought about by impur i ties or dopants. For most semiconductors, the plot of log R versus l/T normally shows a minimum, which occurs at increasingly higher temperatures as the impurity concentration increases.
To the low temperature side of the minimum, that is in the extrinsic range, the curve is almost linear. The low temperature asymptotic s lope of the plot gives the activation energy of the impurity level. The curves obtained in our investigation do not reach a minimum even at 800°K and secondly there is a "kink" in the sl ope . 20  to specimen and appears therefore to be an impurity or structure 28 sensitive property.
The absorption in the 8 to 13 micro~ region was found to be temperature independent whereas the bands in the 2 to 6 micron region showed a temperature dependence which was 29 similar to that of corresponding bands in silicon and germanium.

110
This would indicate that the 2 to 6 micron absorption is related to the characteristic lattice frequencies of diamond while the other band is truly impurity related. It has been shown that the intensity of the latter band is a function of the nitrogen concentration. 30 The absorption characteristics can also be changed by neutron and Largest crystals are reported to be obtained at process temperatures 0 between 1200 and 1300 C; slow cooling had no effect on the crystal growth but prolonged reaction times produced larger crystals. The growth rate of boron phosphide has been reported 45 as a function of temperature, pressure and time. Temperature seems to enhance the growth rate and there appears to be an optimum duration for synthesis of larger crystals. ,

II. METHODS AND EQUIPMENT USED
All high pressure synthesis experiments were carried out in the 600 ton tetrahedral anvil press shown in Figure 1. The theory and design of the original tetrahedral press is described by 11 50,51 Ha • The tetrahedral press is an extension of the "twodimensional" Bridgman anvil concept to three dimensions. A "threedimensional" device is necessary to overcome the problem of the small sample size in the Bridgman anvils. In the te t rahedral press the principle of massive support is still at work but not to the same extent as in Bridgman anvils. This is so because the solid angle subtended by each anvil must decrease as the number of anvils used is increased.
In the tetrahedral press four anvils with triangular faces are driven toward a central point by hydraulic rams whose axes lie along  Figure 3 shows the sample and sample holder as used in this calibration. Two strips of copper (0.015" thick and 3/8" wide) coming from the opposite sides of the tetra-hedron make contact with the ends of the bismuth wire via tantalum taps which serve as the contact between the test specimen and the copper strips. These copper strips are bent against the surface of the tetrahedron and serve as leads for resistance measurement.
Changes of resistance of the bismuth wire during high pressure * compression were measured by a milliohmmeter and electrometer. In a similar fashion, thallium and barium were used for the determination of higher pressure points. The calibration curve is shown in Some of the properties of diamond and boron phosphide have also been studied as part of this investigation. For diamond, the electrical conductivity of heavily doped specimens has been worked out and the effect of proton irradiation investigated.

122
The mechanism of electrical conduction in diamond heavily doped with boron, aluminum and titanium appears to be due to the formation of impurity bands as opposed to the hopping of charge carriers. The activation energies of the impurity levels have been determined from an analysis of the log R versus l/T curves. For boron doped diamond, due to the 'kink' in the log R versus l/T curves, which is unusual, the activation energy has been determined by a s suming a "two-impurity" model with boron as majority impurity and nitrogen as minority impurity. For aluminum and titanium doped samples, which do not show any anomalies up to 200°K, the analytical data could be fitted very well with experimental data.
Type I diamond irradiated with high energy protons shows an increase in the lattice constant which has been detected from Raman spectra l measurements. The infrared absorption spectra before and after irradiation r eveal a characteristic absorption at 6.92 microns in the irradiated sample. This could be · a manifestation of defects induced as a result of irradiation or some possible transmutation of car bon to boron a s explained earlier in the text. However, the band e dge r emaine d the same even after irradiation.
Boron phosphide has been synthesized. The morphology and the e ff ect of thermal neutron irradiation have been investigated.
Op t ical absorption measurements have been made and the fun damental absorption edge has been found.
The rmal neutron irradiation appears to i ncrease the current at fixed voltages in detectors made of boron phosphide, boron oxide and hexagonal boron nitride. Boron phosphide seems to be more suited than the other compounds for detector applications at high voltages due to its tendency to be more sensitive in this region.
The increase in current during irradiation has been explained as resulting from defects (trapping levels) induced as a result of neutron irradiation. In the area of boron phosphide, efforts are being made to synthesize "voidless" single crystals using new growth techniques.
The generation of well-developed homogeneous single crystals will enable more reliable measurements on the fundamental absorption edge and facilitate conductivity studies which are meager at this point for this compound. For thermal neutron detection boron phosphide seems to be a definite possibility. More quantitative experiments are being planned to resolve the actual neutron flux striking the target (by accounting for contribution due to y radiation) in order to calculate the neutron absorption efficiency and