ANDREWS, Judge.

Julian Niles, a doctoral student at Georgia Tech, suffered severe injuries in a laboratory accident. Chemicals he mixed inside a metal canister reacted violently, causing an explosion which sent fragments of the metal into his leg and lower abdomen. He sued Georgia Tech and the Board of Regents, claiming his professor and the University's administration should have provided him with laboratory safety training and should have warned him of the dangers of mixing acetone, ethanol, and nitric acid inside a metal container. Niles presented his case to a jury, but at the close of his liability evidence the trial court directed a verdict in the University's favor. After carefully reviewing the evidence, including the evidence Niles claims was improperly excluded, we affirm the trial court's judgment.

A trial court may direct a verdict in a party's favor only "[i]f there is no conflict in the evidence as to any material issue and the evidence introduced, with all reasonable deductions therefrom, shall demand a particular verdict . . . ." OCGA § 9-11-50 (a). We view that evidence in a light most favorable to Niles, Sorrells Constr. Co. v. Chandler c., P.C., 214 Ga. App. 193, n. 1 ( 447 S.E.2d 101) (1994), and consider only those grounds raised in the University's motion for directed verdict. Nelson v. Polk Co. Historical Society, 216 Ga. App. 756 (1) ( 456 S.E.2d 93) (1995).

In pursuit of a doctorate in physics, Niles was enrolled in a class called "special problems" under Dr. Erbil and was experimenting with methods of producing superconducting crystals. Doctoral students such as Niles have supervision but also spend much time working independently on experiments. This explosion occurred when Niles was cleaning out a metal canister involved in the experiment, using a procedure outlined for him by a former Ph.D. student who had, the previous quarter, been his "mentor" in the lab. The experts who testified believed the explosion was caused by the combination of acetone and ethanol, which are "organic" chemicals, with inorganic nitric acid, inside a metal canister containing residue of titanium isopropoxide.

Before entering this laboratory course, Niles had graduated summa cum laude from the University of the Virgin Islands with a degree in chemistry. He had also obtained a masters' degree in physics from Clark Atlanta University, maintaining a 4.0 grade point average there. He passed his oral comprehensive examinations approximately 10 months before the accident. During his scholastic career, Niles had spent "hundreds" of hours in laboratories. He had previously worked with acetone, ethanol, and nitric acid and knew many of the properties of these chemicals.

Niles' studies had also taught him to use reference materials. He was aware there are "a million" organic compounds, and he had learned to look up the nature of any compounds he did not understand. He had previously used a Merck's index to discover the properties of various chemicals, but he did not use that book — or any other reference — to investigate these chemicals before combining them. Neither did Niles ask any questions of Dr. Erbil, although he was familiar with the professor's "open door" policy. Instead, Niles relied on the process described to him by the former student.

1. Neither Georgia Tech nor Dr. Erbil was required to warn Niles of the dangers of mixing these chemicals. "Whether a duty to warn exists depends upon foreseeability of the [danger], the type of danger involved, and the foreseeability of the user's knowledge of the danger." (Citations and punctuation omitted). Exxon Corp. v. Jones, 209 Ga. App. 373, 375 ( 433 S.E.2d 350) (1993) (discussing duty to warn in products liability context). Although a university student is an invitee to whom the university owes a duty of reasonable care, see Rawlings v. Angelo State Univ., 648 S.W.2d 430, 432 (Tex.App. 1983), college administrators do not stand in loco parentis to adult college students. Bradshaw v. Rawlings, 612 F.2d 135 (3rd Cir. 1979). See Walker v. Daniels, 200 Ga. App. 150, 154-155 ( 407 S.E.2d 70) (1991).

Dr. Erbil had the right to assume that a physics doctoral student, who had graduated with highest honors in chemistry, would either know the dangers of mixing these chemicals or would perform the research necessary to determine those dangers and take the necessary precautions. Niles' own expert, a chemist, testified acetone, ethanol, and nitric acid are "common" chemicals in labs and that "in all probability," mixing them would produce a reaction like the one resulting in Niles' injury. "Ordinarily, there is no duty to give warning to the members of a profession against generally known risks. There need be no warning to one in a particular trade or profession against a danger generally known to that trade or profession." (Citations and punctuation omitted). Brown v. Apollo Indus., 199 Ga. App. 260, 263 (2) (b) ( 404 S.E.2d 447) (1991); Eyster v. Borg-Warner Corp., 131 Ga. App. 702, 705 (2) ( 206 S.E.2d 668) (1974). Under these circumstances, neither Dr. Erbil nor Georgia Tech had any duty to warn a student with a degree in chemistry of the dangers of mixing these common chemicals.

2. Even if Georgia Tech and Dr. Erbil had a duty to warn Niles of this danger by hand-delivering him copies of "material safety data sheets," by reducing the cleaning procedure to writing, or by giving him a laboratory safety course, as he suggests, no evidence supports a reasonable inference that this action would have prevented the accident. "An inference cannot be based upon evidence which is too uncertain or speculative or which raises merely a conjecture or possibility." (Citation and punctuation omitted.) Derry v. Clements, 197 Ga. App. 173, 174 ( 397 S.E.2d 594) (1990). "The plaintiff must introduce evidence which affords a reasonable basis for the conclusion that it is more likely than not that the conduct of the defendant was a cause in fact of the result. A mere possibility of such causation is not enough; and when the matter remains one of pure speculation or conjecture, or the probabilities are at best evenly balanced, it becomes the duty of the court to direct a verdict for the defendant." (Citations and punctuation omitted). Nelson, supra at 757 (2).

With the "million" organic compounds and untold numbers of inorganic compounds which might be combined, nothing in the record supports an inference that any laboratory safety course would have covered the mixing of these chemicals. Niles himself testified he simply relied on his former colleague's oral "outline" of the cleaning procedure and did not read any safety materials which might have warned him of danger because he did not think it worth his time to investigate. Unless we were to hold Georgia Tech responsible for standing at Niles' shoulder and overseeing every action this doctoral student takes in the lab, which we will not do, any claim that his injury was proximately caused by the University's failure to provide him with detailed safety information is too speculative as a matter of law.

3. The court also properly directed a verdict in favor of Georgia Tech and the Board of Regents because Niles is deemed, as a matter of law, to have equal knowledge of the dangers of mixing these chemicals. As in Evans v. Johns Hopkins Univ., 167 A.2d 591, 593 (Md.App. 1961), this doctoral student was "far from being newly initiated in laboratory work," was familiar with the chemicals he mixed, knew ethanol was flammable and that acid on metal would produce hydrogen gas, and had access to information which would tell him more about these chemicals. The bottle of nitric acid he used contained a label encouraging him to read a material safety data sheet about the chemical before using it, but Niles did not read the label. Instead, he "preferred convenience to safety and knowingly took the chance." Evans, supra at 594.

"`[E]ven assuming [Dr. Erbil] had knowledge of the peril, it was [Niles'] burden to come forward with specific evidence that [Erbil's] knowledge of the peril was superior, for that is the true ground of the proprietor's liability.' [Cit.]" Cook v. Home Depot, 214 Ga. App. 133, 135 (1) ( 447 S.E.2d 35) (1994). The evidence shows without dispute, however, that the danger of mixing these chemicals should have been as obvious to Niles as the danger of a rapidly moving stream is to the average citizen. See Harmon v. City of College Park, 218 Ga. App. 136, 137-138 ( 460 S.E.2d 554) (1995); Brown, supra at 263.

4. Because the evidentiary errors which Niles enumerates do not change our decision, we need not consider those enumerations. See MNM 5, Inc. v. Anderson/6438 Northeast Partners, 215 Ga. App. 407, 410 (3) ( 451 S.E.2d 788) (1994).

Judgment affirmed. Pope, P.J., and Smith, J., concur.

ANDREWS, Judge.

DECIDED JUNE 7, 1996 — RECONSIDERATION DENIED JULY 2, 1996 — CERT. APPLIED FOR.